diff --git "a/sft/hyperrouter/trainer_state.json" "b/sft/hyperrouter/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/sft/hyperrouter/trainer_state.json"
@@ -0,0 +1,66571 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999398785546805,
+  "eval_steps": 500,
+  "global_step": 8316,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00012024289063909097,
+      "grad_norm": 16.327665936250575,
+      "learning_rate": 1.6e-08,
+      "loss": 1.4003,
+      "num_input_tokens_seen": 20095,
+      "step": 1
+    },
+    {
+      "epoch": 0.00024048578127818193,
+      "grad_norm": 21.93773318199164,
+      "learning_rate": 3.2e-08,
+      "loss": 1.466,
+      "num_input_tokens_seen": 38475,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003607286719172729,
+      "grad_norm": 16.203408261343995,
+      "learning_rate": 4.8e-08,
+      "loss": 1.309,
+      "num_input_tokens_seen": 56760,
+      "step": 3
+    },
+    {
+      "epoch": 0.00048097156255636386,
+      "grad_norm": 16.462381655575808,
+      "learning_rate": 6.4e-08,
+      "loss": 1.3787,
+      "num_input_tokens_seen": 75345,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006012144531954548,
+      "grad_norm": 18.602705072838127,
+      "learning_rate": 8e-08,
+      "loss": 1.4706,
+      "num_input_tokens_seen": 92950,
+      "step": 5
+    },
+    {
+      "epoch": 0.0007214573438345458,
+      "grad_norm": 18.879179648098496,
+      "learning_rate": 9.6e-08,
+      "loss": 1.3408,
+      "num_input_tokens_seen": 112915,
+      "step": 6
+    },
+    {
+      "epoch": 0.0008417002344736367,
+      "grad_norm": 3.0044250937303145,
+      "learning_rate": 1.12e-07,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 177630,
+      "step": 7
+    },
+    {
+      "epoch": 0.0009619431251127277,
+      "grad_norm": 28.316157789661276,
+      "learning_rate": 1.28e-07,
+      "loss": 1.5846,
+      "num_input_tokens_seen": 192850,
+      "step": 8
+    },
+    {
+      "epoch": 0.0010821860157518186,
+      "grad_norm": 15.87731397869836,
+      "learning_rate": 1.44e-07,
+      "loss": 1.5224,
+      "num_input_tokens_seen": 209115,
+      "step": 9
+    },
+    {
+      "epoch": 0.0012024289063909096,
+      "grad_norm": 15.283545522446454,
+      "learning_rate": 1.6e-07,
+      "loss": 1.5422,
+      "num_input_tokens_seen": 221905,
+      "step": 10
+    },
+    {
+      "epoch": 0.0013226717970300007,
+      "grad_norm": 14.726159898300184,
+      "learning_rate": 1.76e-07,
+      "loss": 1.3525,
+      "num_input_tokens_seen": 241555,
+      "step": 11
+    },
+    {
+      "epoch": 0.0014429146876690916,
+      "grad_norm": 12.070506814161272,
+      "learning_rate": 1.92e-07,
+      "loss": 1.2949,
+      "num_input_tokens_seen": 262405,
+      "step": 12
+    },
+    {
+      "epoch": 0.0015631575783081825,
+      "grad_norm": 15.037569175387873,
+      "learning_rate": 2.0799999999999998e-07,
+      "loss": 1.4472,
+      "num_input_tokens_seen": 279860,
+      "step": 13
+    },
+    {
+      "epoch": 0.0016834004689472734,
+      "grad_norm": 18.056233630077188,
+      "learning_rate": 2.24e-07,
+      "loss": 1.4063,
+      "num_input_tokens_seen": 301765,
+      "step": 14
+    },
+    {
+      "epoch": 0.0018036433595863645,
+      "grad_norm": 19.440810891818074,
+      "learning_rate": 2.4e-07,
+      "loss": 1.4412,
+      "num_input_tokens_seen": 323140,
+      "step": 15
+    },
+    {
+      "epoch": 0.0019238862502254555,
+      "grad_norm": 3.16247087930905,
+      "learning_rate": 2.56e-07,
+      "loss": 0.6797,
+      "num_input_tokens_seen": 378230,
+      "step": 16
+    },
+    {
+      "epoch": 0.0020441291408645466,
+      "grad_norm": 12.790671908296467,
+      "learning_rate": 2.72e-07,
+      "loss": 1.3939,
+      "num_input_tokens_seen": 396130,
+      "step": 17
+    },
+    {
+      "epoch": 0.0021643720315036373,
+      "grad_norm": 3.030801715669472,
+      "learning_rate": 2.88e-07,
+      "loss": 0.7646,
+      "num_input_tokens_seen": 457565,
+      "step": 18
+    },
+    {
+      "epoch": 0.0022846149221427284,
+      "grad_norm": 17.8083550486306,
+      "learning_rate": 3.0399999999999997e-07,
+      "loss": 1.3484,
+      "num_input_tokens_seen": 477960,
+      "step": 19
+    },
+    {
+      "epoch": 0.002404857812781819,
+      "grad_norm": 3.3889049949298844,
+      "learning_rate": 3.2e-07,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 532020,
+      "step": 20
+    },
+    {
+      "epoch": 0.0025251007034209102,
+      "grad_norm": 15.467730560738998,
+      "learning_rate": 3.36e-07,
+      "loss": 1.3661,
+      "num_input_tokens_seen": 549880,
+      "step": 21
+    },
+    {
+      "epoch": 0.0026453435940600014,
+      "grad_norm": 21.911544960122527,
+      "learning_rate": 3.52e-07,
+      "loss": 1.3293,
+      "num_input_tokens_seen": 572290,
+      "step": 22
+    },
+    {
+      "epoch": 0.002765586484699092,
+      "grad_norm": 20.69427988265371,
+      "learning_rate": 3.6799999999999996e-07,
+      "loss": 1.4861,
+      "num_input_tokens_seen": 589705,
+      "step": 23
+    },
+    {
+      "epoch": 0.002885829375338183,
+      "grad_norm": 19.429603528947897,
+      "learning_rate": 3.84e-07,
+      "loss": 1.4827,
+      "num_input_tokens_seen": 608200,
+      "step": 24
+    },
+    {
+      "epoch": 0.003006072265977274,
+      "grad_norm": 17.7170545526128,
+      "learning_rate": 4e-07,
+      "loss": 1.3445,
+      "num_input_tokens_seen": 629060,
+      "step": 25
+    },
+    {
+      "epoch": 0.003126315156616365,
+      "grad_norm": 11.960001671848111,
+      "learning_rate": 4.1599999999999997e-07,
+      "loss": 1.2165,
+      "num_input_tokens_seen": 648295,
+      "step": 26
+    },
+    {
+      "epoch": 0.003246558047255456,
+      "grad_norm": 11.74622932889789,
+      "learning_rate": 4.3199999999999995e-07,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 668170,
+      "step": 27
+    },
+    {
+      "epoch": 0.003366800937894547,
+      "grad_norm": 14.845658893922659,
+      "learning_rate": 4.48e-07,
+      "loss": 1.245,
+      "num_input_tokens_seen": 686765,
+      "step": 28
+    },
+    {
+      "epoch": 0.003487043828533638,
+      "grad_norm": 13.403104423982521,
+      "learning_rate": 4.64e-07,
+      "loss": 1.3309,
+      "num_input_tokens_seen": 706220,
+      "step": 29
+    },
+    {
+      "epoch": 0.003607286719172729,
+      "grad_norm": 11.242430165804379,
+      "learning_rate": 4.8e-07,
+      "loss": 1.258,
+      "num_input_tokens_seen": 726070,
+      "step": 30
+    },
+    {
+      "epoch": 0.00372752960981182,
+      "grad_norm": 14.977088957468919,
+      "learning_rate": 4.96e-07,
+      "loss": 1.279,
+      "num_input_tokens_seen": 744700,
+      "step": 31
+    },
+    {
+      "epoch": 0.003847772500450911,
+      "grad_norm": 14.752931309798909,
+      "learning_rate": 5.12e-07,
+      "loss": 1.354,
+      "num_input_tokens_seen": 763145,
+      "step": 32
+    },
+    {
+      "epoch": 0.003968015391090002,
+      "grad_norm": 13.515572929734024,
+      "learning_rate": 5.28e-07,
+      "loss": 1.331,
+      "num_input_tokens_seen": 779715,
+      "step": 33
+    },
+    {
+      "epoch": 0.004088258281729093,
+      "grad_norm": 11.770082249868278,
+      "learning_rate": 5.44e-07,
+      "loss": 1.1584,
+      "num_input_tokens_seen": 800710,
+      "step": 34
+    },
+    {
+      "epoch": 0.004208501172368184,
+      "grad_norm": 8.89325054688529,
+      "learning_rate": 5.6e-07,
+      "loss": 1.2711,
+      "num_input_tokens_seen": 819980,
+      "step": 35
+    },
+    {
+      "epoch": 0.0043287440630072746,
+      "grad_norm": 8.22879362274236,
+      "learning_rate": 5.76e-07,
+      "loss": 1.1883,
+      "num_input_tokens_seen": 838615,
+      "step": 36
+    },
+    {
+      "epoch": 0.004448986953646365,
+      "grad_norm": 2.6453796695003007,
+      "learning_rate": 5.919999999999999e-07,
+      "loss": 0.7104,
+      "num_input_tokens_seen": 896370,
+      "step": 37
+    },
+    {
+      "epoch": 0.004569229844285457,
+      "grad_norm": 7.717980340101501,
+      "learning_rate": 6.079999999999999e-07,
+      "loss": 1.2097,
+      "num_input_tokens_seen": 914485,
+      "step": 38
+    },
+    {
+      "epoch": 0.0046894727349245475,
+      "grad_norm": 9.793112299185252,
+      "learning_rate": 6.24e-07,
+      "loss": 1.1771,
+      "num_input_tokens_seen": 929050,
+      "step": 39
+    },
+    {
+      "epoch": 0.004809715625563638,
+      "grad_norm": 9.499707645033508,
+      "learning_rate": 6.4e-07,
+      "loss": 1.1611,
+      "num_input_tokens_seen": 946270,
+      "step": 40
+    },
+    {
+      "epoch": 0.00492995851620273,
+      "grad_norm": 7.216169069521672,
+      "learning_rate": 6.56e-07,
+      "loss": 1.1282,
+      "num_input_tokens_seen": 964925,
+      "step": 41
+    },
+    {
+      "epoch": 0.0050502014068418205,
+      "grad_norm": 6.939434270515214,
+      "learning_rate": 6.72e-07,
+      "loss": 1.2078,
+      "num_input_tokens_seen": 982930,
+      "step": 42
+    },
+    {
+      "epoch": 0.005170444297480911,
+      "grad_norm": 22.099305204217522,
+      "learning_rate": 6.879999999999999e-07,
+      "loss": 1.1456,
+      "num_input_tokens_seen": 1003575,
+      "step": 43
+    },
+    {
+      "epoch": 0.005290687188120003,
+      "grad_norm": 8.346215739815168,
+      "learning_rate": 7.04e-07,
+      "loss": 1.0522,
+      "num_input_tokens_seen": 1018935,
+      "step": 44
+    },
+    {
+      "epoch": 0.005410930078759093,
+      "grad_norm": 10.480855956190826,
+      "learning_rate": 7.2e-07,
+      "loss": 1.0691,
+      "num_input_tokens_seen": 1035695,
+      "step": 45
+    },
+    {
+      "epoch": 0.005531172969398184,
+      "grad_norm": 5.395219573536445,
+      "learning_rate": 7.359999999999999e-07,
+      "loss": 1.2168,
+      "num_input_tokens_seen": 1055045,
+      "step": 46
+    },
+    {
+      "epoch": 0.005651415860037276,
+      "grad_norm": 5.775332818730067,
+      "learning_rate": 7.52e-07,
+      "loss": 1.2571,
+      "num_input_tokens_seen": 1074900,
+      "step": 47
+    },
+    {
+      "epoch": 0.005771658750676366,
+      "grad_norm": 7.2837732367086865,
+      "learning_rate": 7.68e-07,
+      "loss": 1.2879,
+      "num_input_tokens_seen": 1090690,
+      "step": 48
+    },
+    {
+      "epoch": 0.005891901641315457,
+      "grad_norm": 5.24541908377467,
+      "learning_rate": 7.84e-07,
+      "loss": 1.1913,
+      "num_input_tokens_seen": 1109180,
+      "step": 49
+    },
+    {
+      "epoch": 0.006012144531954548,
+      "grad_norm": 5.952518299615795,
+      "learning_rate": 8e-07,
+      "loss": 1.0614,
+      "num_input_tokens_seen": 1126250,
+      "step": 50
+    },
+    {
+      "epoch": 0.006132387422593639,
+      "grad_norm": 3.896135622580414,
+      "learning_rate": 8.159999999999999e-07,
+      "loss": 1.0546,
+      "num_input_tokens_seen": 1146190,
+      "step": 51
+    },
+    {
+      "epoch": 0.00625263031323273,
+      "grad_norm": 41.40049844228182,
+      "learning_rate": 8.319999999999999e-07,
+      "loss": 1.1177,
+      "num_input_tokens_seen": 1163045,
+      "step": 52
+    },
+    {
+      "epoch": 0.006372873203871821,
+      "grad_norm": 3.8000165962503587,
+      "learning_rate": 8.48e-07,
+      "loss": 1.0098,
+      "num_input_tokens_seen": 1182895,
+      "step": 53
+    },
+    {
+      "epoch": 0.006493116094510912,
+      "grad_norm": 5.752494032489688,
+      "learning_rate": 8.639999999999999e-07,
+      "loss": 1.0687,
+      "num_input_tokens_seen": 1201215,
+      "step": 54
+    },
+    {
+      "epoch": 0.006613358985150003,
+      "grad_norm": 4.256298698460216,
+      "learning_rate": 8.799999999999999e-07,
+      "loss": 0.9817,
+      "num_input_tokens_seen": 1219080,
+      "step": 55
+    },
+    {
+      "epoch": 0.006733601875789094,
+      "grad_norm": 5.819299297378959,
+      "learning_rate": 8.96e-07,
+      "loss": 1.0443,
+      "num_input_tokens_seen": 1238190,
+      "step": 56
+    },
+    {
+      "epoch": 0.006853844766428185,
+      "grad_norm": 3.690848196547338,
+      "learning_rate": 9.12e-07,
+      "loss": 0.9853,
+      "num_input_tokens_seen": 1254185,
+      "step": 57
+    },
+    {
+      "epoch": 0.006974087657067276,
+      "grad_norm": 3.0774127965544498,
+      "learning_rate": 9.28e-07,
+      "loss": 1.1089,
+      "num_input_tokens_seen": 1275125,
+      "step": 58
+    },
+    {
+      "epoch": 0.007094330547706367,
+      "grad_norm": 2.9513806577729804,
+      "learning_rate": 9.439999999999999e-07,
+      "loss": 1.026,
+      "num_input_tokens_seen": 1295015,
+      "step": 59
+    },
+    {
+      "epoch": 0.007214573438345458,
+      "grad_norm": 3.654693188012229,
+      "learning_rate": 9.6e-07,
+      "loss": 1.0673,
+      "num_input_tokens_seen": 1312620,
+      "step": 60
+    },
+    {
+      "epoch": 0.007334816328984549,
+      "grad_norm": 3.573621302192546,
+      "learning_rate": 9.759999999999998e-07,
+      "loss": 1.068,
+      "num_input_tokens_seen": 1332885,
+      "step": 61
+    },
+    {
+      "epoch": 0.00745505921962364,
+      "grad_norm": 3.7230555195724224,
+      "learning_rate": 9.92e-07,
+      "loss": 0.8656,
+      "num_input_tokens_seen": 1351555,
+      "step": 62
+    },
+    {
+      "epoch": 0.007575302110262731,
+      "grad_norm": 4.149777613467972,
+      "learning_rate": 1.008e-06,
+      "loss": 1.0267,
+      "num_input_tokens_seen": 1369165,
+      "step": 63
+    },
+    {
+      "epoch": 0.007695545000901822,
+      "grad_norm": 6.0850187219007985,
+      "learning_rate": 1.024e-06,
+      "loss": 1.0908,
+      "num_input_tokens_seen": 1388270,
+      "step": 64
+    },
+    {
+      "epoch": 0.007815787891540913,
+      "grad_norm": 3.340632782893104,
+      "learning_rate": 1.04e-06,
+      "loss": 0.9867,
+      "num_input_tokens_seen": 1407300,
+      "step": 65
+    },
+    {
+      "epoch": 0.007936030782180003,
+      "grad_norm": 3.582789825081253,
+      "learning_rate": 1.056e-06,
+      "loss": 0.9946,
+      "num_input_tokens_seen": 1424470,
+      "step": 66
+    },
+    {
+      "epoch": 0.008056273672819095,
+      "grad_norm": 3.1541221437131344,
+      "learning_rate": 1.072e-06,
+      "loss": 0.98,
+      "num_input_tokens_seen": 1442155,
+      "step": 67
+    },
+    {
+      "epoch": 0.008176516563458186,
+      "grad_norm": 3.6902406400222256,
+      "learning_rate": 1.088e-06,
+      "loss": 1.1719,
+      "num_input_tokens_seen": 1459345,
+      "step": 68
+    },
+    {
+      "epoch": 0.008296759454097276,
+      "grad_norm": 3.4214569083423814,
+      "learning_rate": 1.1040000000000001e-06,
+      "loss": 1.0531,
+      "num_input_tokens_seen": 1478285,
+      "step": 69
+    },
+    {
+      "epoch": 0.008417002344736368,
+      "grad_norm": 4.228524063604063,
+      "learning_rate": 1.12e-06,
+      "loss": 0.7166,
+      "num_input_tokens_seen": 1541550,
+      "step": 70
+    },
+    {
+      "epoch": 0.008537245235375458,
+      "grad_norm": 2.6260912402022254,
+      "learning_rate": 1.1359999999999998e-06,
+      "loss": 1.0047,
+      "num_input_tokens_seen": 1560725,
+      "step": 71
+    },
+    {
+      "epoch": 0.008657488126014549,
+      "grad_norm": 2.5369467447112273,
+      "learning_rate": 1.152e-06,
+      "loss": 0.6681,
+      "num_input_tokens_seen": 1627375,
+      "step": 72
+    },
+    {
+      "epoch": 0.00877773101665364,
+      "grad_norm": 3.1859449969188502,
+      "learning_rate": 1.1679999999999999e-06,
+      "loss": 1.0805,
+      "num_input_tokens_seen": 1646330,
+      "step": 73
+    },
+    {
+      "epoch": 0.00889797390729273,
+      "grad_norm": 2.8657630964965906,
+      "learning_rate": 1.1839999999999998e-06,
+      "loss": 1.0692,
+      "num_input_tokens_seen": 1664480,
+      "step": 74
+    },
+    {
+      "epoch": 0.009018216797931822,
+      "grad_norm": 10.61256176415637,
+      "learning_rate": 1.2e-06,
+      "loss": 0.9123,
+      "num_input_tokens_seen": 1680835,
+      "step": 75
+    },
+    {
+      "epoch": 0.009138459688570914,
+      "grad_norm": 2.331018335054291,
+      "learning_rate": 1.2159999999999999e-06,
+      "loss": 1.0267,
+      "num_input_tokens_seen": 1697135,
+      "step": 76
+    },
+    {
+      "epoch": 0.009258702579210003,
+      "grad_norm": 2.5651943665540737,
+      "learning_rate": 1.232e-06,
+      "loss": 0.8549,
+      "num_input_tokens_seen": 1717210,
+      "step": 77
+    },
+    {
+      "epoch": 0.009378945469849095,
+      "grad_norm": 2.765305676893068,
+      "learning_rate": 1.248e-06,
+      "loss": 0.9875,
+      "num_input_tokens_seen": 1736200,
+      "step": 78
+    },
+    {
+      "epoch": 0.009499188360488187,
+      "grad_norm": 3.820065755361661,
+      "learning_rate": 1.2639999999999999e-06,
+      "loss": 0.875,
+      "num_input_tokens_seen": 1754585,
+      "step": 79
+    },
+    {
+      "epoch": 0.009619431251127276,
+      "grad_norm": 2.9095283464670327,
+      "learning_rate": 1.28e-06,
+      "loss": 0.835,
+      "num_input_tokens_seen": 1773515,
+      "step": 80
+    },
+    {
+      "epoch": 0.009739674141766368,
+      "grad_norm": 2.7823775247501747,
+      "learning_rate": 1.296e-06,
+      "loss": 0.9545,
+      "num_input_tokens_seen": 1791375,
+      "step": 81
+    },
+    {
+      "epoch": 0.00985991703240546,
+      "grad_norm": 2.640773573543164,
+      "learning_rate": 1.312e-06,
+      "loss": 0.9682,
+      "num_input_tokens_seen": 1811575,
+      "step": 82
+    },
+    {
+      "epoch": 0.00998015992304455,
+      "grad_norm": 3.7630226558544817,
+      "learning_rate": 1.328e-06,
+      "loss": 0.9128,
+      "num_input_tokens_seen": 1828625,
+      "step": 83
+    },
+    {
+      "epoch": 0.010100402813683641,
+      "grad_norm": 2.856679212156785,
+      "learning_rate": 1.344e-06,
+      "loss": 1.047,
+      "num_input_tokens_seen": 1845020,
+      "step": 84
+    },
+    {
+      "epoch": 0.010220645704322733,
+      "grad_norm": 2.359491250173105,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 0.9729,
+      "num_input_tokens_seen": 1863380,
+      "step": 85
+    },
+    {
+      "epoch": 0.010340888594961822,
+      "grad_norm": 2.468919473260775,
+      "learning_rate": 1.3759999999999998e-06,
+      "loss": 1.0167,
+      "num_input_tokens_seen": 1880685,
+      "step": 86
+    },
+    {
+      "epoch": 0.010461131485600914,
+      "grad_norm": 2.3568124171488276,
+      "learning_rate": 1.3919999999999998e-06,
+      "loss": 1.0316,
+      "num_input_tokens_seen": 1898240,
+      "step": 87
+    },
+    {
+      "epoch": 0.010581374376240005,
+      "grad_norm": 2.2460480078903706,
+      "learning_rate": 1.408e-06,
+      "loss": 0.8916,
+      "num_input_tokens_seen": 1919310,
+      "step": 88
+    },
+    {
+      "epoch": 0.010701617266879095,
+      "grad_norm": 3.4220106666807295,
+      "learning_rate": 1.4239999999999998e-06,
+      "loss": 1.005,
+      "num_input_tokens_seen": 1937895,
+      "step": 89
+    },
+    {
+      "epoch": 0.010821860157518187,
+      "grad_norm": 3.3077822718427985,
+      "learning_rate": 1.44e-06,
+      "loss": 0.8353,
+      "num_input_tokens_seen": 1955380,
+      "step": 90
+    },
+    {
+      "epoch": 0.010942103048157278,
+      "grad_norm": 2.2062997048467285,
+      "learning_rate": 1.456e-06,
+      "loss": 0.9095,
+      "num_input_tokens_seen": 1974835,
+      "step": 91
+    },
+    {
+      "epoch": 0.011062345938796368,
+      "grad_norm": 2.057796965170375,
+      "learning_rate": 1.4719999999999998e-06,
+      "loss": 0.9946,
+      "num_input_tokens_seen": 1998000,
+      "step": 92
+    },
+    {
+      "epoch": 0.01118258882943546,
+      "grad_norm": 2.270378394421132,
+      "learning_rate": 1.488e-06,
+      "loss": 1.0119,
+      "num_input_tokens_seen": 2015855,
+      "step": 93
+    },
+    {
+      "epoch": 0.011302831720074551,
+      "grad_norm": 2.4562066154538016,
+      "learning_rate": 1.504e-06,
+      "loss": 0.8891,
+      "num_input_tokens_seen": 2035320,
+      "step": 94
+    },
+    {
+      "epoch": 0.011423074610713641,
+      "grad_norm": 3.495932623287209,
+      "learning_rate": 1.5199999999999998e-06,
+      "loss": 0.9402,
+      "num_input_tokens_seen": 2051655,
+      "step": 95
+    },
+    {
+      "epoch": 0.011543317501352733,
+      "grad_norm": 2.7479802049500868,
+      "learning_rate": 1.536e-06,
+      "loss": 0.868,
+      "num_input_tokens_seen": 2072610,
+      "step": 96
+    },
+    {
+      "epoch": 0.011663560391991824,
+      "grad_norm": 2.559521939773065,
+      "learning_rate": 1.552e-06,
+      "loss": 0.9809,
+      "num_input_tokens_seen": 2088965,
+      "step": 97
+    },
+    {
+      "epoch": 0.011783803282630914,
+      "grad_norm": 2.0619497793104133,
+      "learning_rate": 1.568e-06,
+      "loss": 0.9235,
+      "num_input_tokens_seen": 2107395,
+      "step": 98
+    },
+    {
+      "epoch": 0.011904046173270006,
+      "grad_norm": 3.228180736810277,
+      "learning_rate": 1.584e-06,
+      "loss": 1.0747,
+      "num_input_tokens_seen": 2124690,
+      "step": 99
+    },
+    {
+      "epoch": 0.012024289063909096,
+      "grad_norm": 4.501754251205946,
+      "learning_rate": 1.6e-06,
+      "loss": 1.0256,
+      "num_input_tokens_seen": 2144090,
+      "step": 100
+    },
+    {
+      "epoch": 0.012144531954548187,
+      "grad_norm": 3.564414236264299,
+      "learning_rate": 1.616e-06,
+      "loss": 1.017,
+      "num_input_tokens_seen": 2160740,
+      "step": 101
+    },
+    {
+      "epoch": 0.012264774845187279,
+      "grad_norm": 2.7003108999019814,
+      "learning_rate": 1.6319999999999998e-06,
+      "loss": 0.8935,
+      "num_input_tokens_seen": 2177060,
+      "step": 102
+    },
+    {
+      "epoch": 0.012385017735826368,
+      "grad_norm": 3.8518331275990394,
+      "learning_rate": 1.648e-06,
+      "loss": 0.9754,
+      "num_input_tokens_seen": 2190160,
+      "step": 103
+    },
+    {
+      "epoch": 0.01250526062646546,
+      "grad_norm": 2.542467424896015,
+      "learning_rate": 1.6639999999999999e-06,
+      "loss": 0.9297,
+      "num_input_tokens_seen": 2208670,
+      "step": 104
+    },
+    {
+      "epoch": 0.012625503517104552,
+      "grad_norm": 1.9717915662426795,
+      "learning_rate": 1.6799999999999998e-06,
+      "loss": 0.9041,
+      "num_input_tokens_seen": 2230565,
+      "step": 105
+    },
+    {
+      "epoch": 0.012745746407743641,
+      "grad_norm": 3.3319317561629416,
+      "learning_rate": 1.696e-06,
+      "loss": 0.981,
+      "num_input_tokens_seen": 2248930,
+      "step": 106
+    },
+    {
+      "epoch": 0.012865989298382733,
+      "grad_norm": 4.152197578092614,
+      "learning_rate": 1.7119999999999999e-06,
+      "loss": 0.9556,
+      "num_input_tokens_seen": 2267770,
+      "step": 107
+    },
+    {
+      "epoch": 0.012986232189021825,
+      "grad_norm": 2.4287397190036013,
+      "learning_rate": 1.7279999999999998e-06,
+      "loss": 1.0385,
+      "num_input_tokens_seen": 2285500,
+      "step": 108
+    },
+    {
+      "epoch": 0.013106475079660914,
+      "grad_norm": 9.633842873328389,
+      "learning_rate": 1.744e-06,
+      "loss": 0.9105,
+      "num_input_tokens_seen": 2303695,
+      "step": 109
+    },
+    {
+      "epoch": 0.013226717970300006,
+      "grad_norm": 3.3400616670912298,
+      "learning_rate": 1.7599999999999999e-06,
+      "loss": 1.0078,
+      "num_input_tokens_seen": 2322330,
+      "step": 110
+    },
+    {
+      "epoch": 0.013346960860939098,
+      "grad_norm": 2.070687517532242,
+      "learning_rate": 1.776e-06,
+      "loss": 0.8576,
+      "num_input_tokens_seen": 2342930,
+      "step": 111
+    },
+    {
+      "epoch": 0.013467203751578187,
+      "grad_norm": 2.174410827059307,
+      "learning_rate": 1.792e-06,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 2363860,
+      "step": 112
+    },
+    {
+      "epoch": 0.013587446642217279,
+      "grad_norm": 2.7768033722885535,
+      "learning_rate": 1.8079999999999999e-06,
+      "loss": 1.0026,
+      "num_input_tokens_seen": 2381385,
+      "step": 113
+    },
+    {
+      "epoch": 0.01370768953285637,
+      "grad_norm": 2.1856467022399917,
+      "learning_rate": 1.824e-06,
+      "loss": 0.9446,
+      "num_input_tokens_seen": 2400780,
+      "step": 114
+    },
+    {
+      "epoch": 0.01382793242349546,
+      "grad_norm": 1.8558621114372431,
+      "learning_rate": 1.84e-06,
+      "loss": 0.8301,
+      "num_input_tokens_seen": 2422820,
+      "step": 115
+    },
+    {
+      "epoch": 0.013948175314134552,
+      "grad_norm": 2.9502438004652007,
+      "learning_rate": 1.856e-06,
+      "loss": 0.9762,
+      "num_input_tokens_seen": 2443295,
+      "step": 116
+    },
+    {
+      "epoch": 0.014068418204773643,
+      "grad_norm": 2.608620903998619,
+      "learning_rate": 1.872e-06,
+      "loss": 0.9259,
+      "num_input_tokens_seen": 2463610,
+      "step": 117
+    },
+    {
+      "epoch": 0.014188661095412733,
+      "grad_norm": 1.7840962844132313,
+      "learning_rate": 1.8879999999999998e-06,
+      "loss": 0.9212,
+      "num_input_tokens_seen": 2484605,
+      "step": 118
+    },
+    {
+      "epoch": 0.014308903986051825,
+      "grad_norm": 1.8457455302803532,
+      "learning_rate": 1.904e-06,
+      "loss": 0.9669,
+      "num_input_tokens_seen": 2504505,
+      "step": 119
+    },
+    {
+      "epoch": 0.014429146876690916,
+      "grad_norm": 5.241660970805831,
+      "learning_rate": 1.92e-06,
+      "loss": 0.9142,
+      "num_input_tokens_seen": 2522885,
+      "step": 120
+    },
+    {
+      "epoch": 0.014549389767330006,
+      "grad_norm": 1.9430717357879224,
+      "learning_rate": 1.9359999999999998e-06,
+      "loss": 0.8675,
+      "num_input_tokens_seen": 2540065,
+      "step": 121
+    },
+    {
+      "epoch": 0.014669632657969098,
+      "grad_norm": 2.294012710983687,
+      "learning_rate": 1.9519999999999997e-06,
+      "loss": 0.9447,
+      "num_input_tokens_seen": 2558535,
+      "step": 122
+    },
+    {
+      "epoch": 0.01478987554860819,
+      "grad_norm": 2.5445207274775177,
+      "learning_rate": 1.968e-06,
+      "loss": 0.9382,
+      "num_input_tokens_seen": 2576485,
+      "step": 123
+    },
+    {
+      "epoch": 0.01491011843924728,
+      "grad_norm": 3.440578373906021,
+      "learning_rate": 1.984e-06,
+      "loss": 0.8937,
+      "num_input_tokens_seen": 2594775,
+      "step": 124
+    },
+    {
+      "epoch": 0.01503036132988637,
+      "grad_norm": 4.179068113596611,
+      "learning_rate": 2e-06,
+      "loss": 0.9264,
+      "num_input_tokens_seen": 2610070,
+      "step": 125
+    },
+    {
+      "epoch": 0.015150604220525462,
+      "grad_norm": 2.0109128094553106,
+      "learning_rate": 2.016e-06,
+      "loss": 1.0141,
+      "num_input_tokens_seen": 2628545,
+      "step": 126
+    },
+    {
+      "epoch": 0.015270847111164552,
+      "grad_norm": 2.4370645100205577,
+      "learning_rate": 2.0319999999999998e-06,
+      "loss": 0.8733,
+      "num_input_tokens_seen": 2649150,
+      "step": 127
+    },
+    {
+      "epoch": 0.015391090001803644,
+      "grad_norm": 2.924916290529411,
+      "learning_rate": 2.048e-06,
+      "loss": 0.8337,
+      "num_input_tokens_seen": 2671155,
+      "step": 128
+    },
+    {
+      "epoch": 0.015511332892442733,
+      "grad_norm": 2.1589013481779813,
+      "learning_rate": 2.064e-06,
+      "loss": 0.8824,
+      "num_input_tokens_seen": 2690930,
+      "step": 129
+    },
+    {
+      "epoch": 0.015631575783081827,
+      "grad_norm": 5.078251625531773,
+      "learning_rate": 2.08e-06,
+      "loss": 0.9196,
+      "num_input_tokens_seen": 2708950,
+      "step": 130
+    },
+    {
+      "epoch": 0.015751818673720917,
+      "grad_norm": 2.5727333362203932,
+      "learning_rate": 2.096e-06,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 2777515,
+      "step": 131
+    },
+    {
+      "epoch": 0.015872061564360006,
+      "grad_norm": 1.9660847199014126,
+      "learning_rate": 2.112e-06,
+      "loss": 0.8605,
+      "num_input_tokens_seen": 2797685,
+      "step": 132
+    },
+    {
+      "epoch": 0.0159923044549991,
+      "grad_norm": 8.993657795989597,
+      "learning_rate": 2.128e-06,
+      "loss": 0.9505,
+      "num_input_tokens_seen": 2816880,
+      "step": 133
+    },
+    {
+      "epoch": 0.01611254734563819,
+      "grad_norm": 2.6035258687730667,
+      "learning_rate": 2.144e-06,
+      "loss": 0.9127,
+      "num_input_tokens_seen": 2833810,
+      "step": 134
+    },
+    {
+      "epoch": 0.01623279023627728,
+      "grad_norm": 2.3695092892084557,
+      "learning_rate": 2.16e-06,
+      "loss": 0.8116,
+      "num_input_tokens_seen": 2850270,
+      "step": 135
+    },
+    {
+      "epoch": 0.016353033126916373,
+      "grad_norm": 2.1015483088611235,
+      "learning_rate": 2.176e-06,
+      "loss": 0.8841,
+      "num_input_tokens_seen": 2868385,
+      "step": 136
+    },
+    {
+      "epoch": 0.016473276017555463,
+      "grad_norm": 2.033342147665128,
+      "learning_rate": 2.192e-06,
+      "loss": 0.8097,
+      "num_input_tokens_seen": 2886555,
+      "step": 137
+    },
+    {
+      "epoch": 0.016593518908194552,
+      "grad_norm": 5.374117582207986,
+      "learning_rate": 2.2080000000000003e-06,
+      "loss": 0.872,
+      "num_input_tokens_seen": 2903490,
+      "step": 138
+    },
+    {
+      "epoch": 0.016713761798833646,
+      "grad_norm": 1.7660412787860578,
+      "learning_rate": 2.2240000000000002e-06,
+      "loss": 0.9242,
+      "num_input_tokens_seen": 2923825,
+      "step": 139
+    },
+    {
+      "epoch": 0.016834004689472735,
+      "grad_norm": 2.8295424624413834,
+      "learning_rate": 2.24e-06,
+      "loss": 0.9521,
+      "num_input_tokens_seen": 2942625,
+      "step": 140
+    },
+    {
+      "epoch": 0.016954247580111825,
+      "grad_norm": 2.632831787081112,
+      "learning_rate": 2.2559999999999997e-06,
+      "loss": 0.9146,
+      "num_input_tokens_seen": 2962145,
+      "step": 141
+    },
+    {
+      "epoch": 0.017074490470750915,
+      "grad_norm": 1.9878878790787027,
+      "learning_rate": 2.2719999999999996e-06,
+      "loss": 0.8116,
+      "num_input_tokens_seen": 2982295,
+      "step": 142
+    },
+    {
+      "epoch": 0.01719473336139001,
+      "grad_norm": 2.662017608521343,
+      "learning_rate": 2.2879999999999995e-06,
+      "loss": 0.9065,
+      "num_input_tokens_seen": 2999565,
+      "step": 143
+    },
+    {
+      "epoch": 0.017314976252029098,
+      "grad_norm": 3.3444672360880463,
+      "learning_rate": 2.304e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 3019085,
+      "step": 144
+    },
+    {
+      "epoch": 0.017435219142668188,
+      "grad_norm": 2.5321708153526066,
+      "learning_rate": 2.32e-06,
+      "loss": 1.0048,
+      "num_input_tokens_seen": 3037890,
+      "step": 145
+    },
+    {
+      "epoch": 0.01755546203330728,
+      "grad_norm": 2.9943159986207117,
+      "learning_rate": 2.3359999999999997e-06,
+      "loss": 0.6683,
+      "num_input_tokens_seen": 3090875,
+      "step": 146
+    },
+    {
+      "epoch": 0.01767570492394637,
+      "grad_norm": 21.870716491930853,
+      "learning_rate": 2.3519999999999997e-06,
+      "loss": 0.8908,
+      "num_input_tokens_seen": 3110875,
+      "step": 147
+    },
+    {
+      "epoch": 0.01779594781458546,
+      "grad_norm": 2.1260132216716134,
+      "learning_rate": 2.3679999999999996e-06,
+      "loss": 0.9163,
+      "num_input_tokens_seen": 3129185,
+      "step": 148
+    },
+    {
+      "epoch": 0.017916190705224554,
+      "grad_norm": 1.8728693818910558,
+      "learning_rate": 2.384e-06,
+      "loss": 0.9325,
+      "num_input_tokens_seen": 3146435,
+      "step": 149
+    },
+    {
+      "epoch": 0.018036433595863644,
+      "grad_norm": 2.0437753837505306,
+      "learning_rate": 2.4e-06,
+      "loss": 0.8984,
+      "num_input_tokens_seen": 3167015,
+      "step": 150
+    },
+    {
+      "epoch": 0.018156676486502734,
+      "grad_norm": 2.291205503660783,
+      "learning_rate": 2.416e-06,
+      "loss": 0.9711,
+      "num_input_tokens_seen": 3182675,
+      "step": 151
+    },
+    {
+      "epoch": 0.018276919377141827,
+      "grad_norm": 2.0189930539034338,
+      "learning_rate": 2.4319999999999998e-06,
+      "loss": 0.8963,
+      "num_input_tokens_seen": 3203770,
+      "step": 152
+    },
+    {
+      "epoch": 0.018397162267780917,
+      "grad_norm": 3.091421591798236,
+      "learning_rate": 2.4479999999999997e-06,
+      "loss": 0.9463,
+      "num_input_tokens_seen": 3221450,
+      "step": 153
+    },
+    {
+      "epoch": 0.018517405158420007,
+      "grad_norm": 3.2727675877354874,
+      "learning_rate": 2.464e-06,
+      "loss": 0.9849,
+      "num_input_tokens_seen": 3243945,
+      "step": 154
+    },
+    {
+      "epoch": 0.0186376480490591,
+      "grad_norm": 1.7930897339711367,
+      "learning_rate": 2.48e-06,
+      "loss": 0.9505,
+      "num_input_tokens_seen": 3262195,
+      "step": 155
+    },
+    {
+      "epoch": 0.01875789093969819,
+      "grad_norm": 2.641276078851638,
+      "learning_rate": 2.496e-06,
+      "loss": 0.7935,
+      "num_input_tokens_seen": 3282455,
+      "step": 156
+    },
+    {
+      "epoch": 0.01887813383033728,
+      "grad_norm": 2.33853955250005,
+      "learning_rate": 2.512e-06,
+      "loss": 1.0863,
+      "num_input_tokens_seen": 3299550,
+      "step": 157
+    },
+    {
+      "epoch": 0.018998376720976373,
+      "grad_norm": 3.0425475329245555,
+      "learning_rate": 2.5279999999999998e-06,
+      "loss": 0.874,
+      "num_input_tokens_seen": 3317085,
+      "step": 158
+    },
+    {
+      "epoch": 0.019118619611615463,
+      "grad_norm": 2.533432673460887,
+      "learning_rate": 2.544e-06,
+      "loss": 0.9242,
+      "num_input_tokens_seen": 3333405,
+      "step": 159
+    },
+    {
+      "epoch": 0.019238862502254553,
+      "grad_norm": 2.068036193078605,
+      "learning_rate": 2.56e-06,
+      "loss": 0.9028,
+      "num_input_tokens_seen": 3351535,
+      "step": 160
+    },
+    {
+      "epoch": 0.019359105392893646,
+      "grad_norm": 8.547892839133752,
+      "learning_rate": 2.576e-06,
+      "loss": 0.928,
+      "num_input_tokens_seen": 3368525,
+      "step": 161
+    },
+    {
+      "epoch": 0.019479348283532736,
+      "grad_norm": 3.0155686367278887,
+      "learning_rate": 2.592e-06,
+      "loss": 0.8641,
+      "num_input_tokens_seen": 3384280,
+      "step": 162
+    },
+    {
+      "epoch": 0.019599591174171826,
+      "grad_norm": 3.029801884905472,
+      "learning_rate": 2.608e-06,
+      "loss": 0.9857,
+      "num_input_tokens_seen": 3402485,
+      "step": 163
+    },
+    {
+      "epoch": 0.01971983406481092,
+      "grad_norm": 2.5536934830139133,
+      "learning_rate": 2.624e-06,
+      "loss": 0.9003,
+      "num_input_tokens_seen": 3420615,
+      "step": 164
+    },
+    {
+      "epoch": 0.01984007695545001,
+      "grad_norm": 3.315173991714741,
+      "learning_rate": 2.64e-06,
+      "loss": 0.7828,
+      "num_input_tokens_seen": 3437530,
+      "step": 165
+    },
+    {
+      "epoch": 0.0199603198460891,
+      "grad_norm": 4.120677509621033,
+      "learning_rate": 2.656e-06,
+      "loss": 0.8391,
+      "num_input_tokens_seen": 3457160,
+      "step": 166
+    },
+    {
+      "epoch": 0.020080562736728192,
+      "grad_norm": 2.0188829573316327,
+      "learning_rate": 2.672e-06,
+      "loss": 0.8717,
+      "num_input_tokens_seen": 3476250,
+      "step": 167
+    },
+    {
+      "epoch": 0.020200805627367282,
+      "grad_norm": 2.163243368932872,
+      "learning_rate": 2.688e-06,
+      "loss": 1.0022,
+      "num_input_tokens_seen": 3494085,
+      "step": 168
+    },
+    {
+      "epoch": 0.02032104851800637,
+      "grad_norm": 2.305938588044211,
+      "learning_rate": 2.704e-06,
+      "loss": 0.8716,
+      "num_input_tokens_seen": 3515380,
+      "step": 169
+    },
+    {
+      "epoch": 0.020441291408645465,
+      "grad_norm": 2.5376505925090993,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 0.8751,
+      "num_input_tokens_seen": 3535190,
+      "step": 170
+    },
+    {
+      "epoch": 0.020561534299284555,
+      "grad_norm": 2.6634625991119214,
+      "learning_rate": 2.736e-06,
+      "loss": 0.8974,
+      "num_input_tokens_seen": 3553835,
+      "step": 171
+    },
+    {
+      "epoch": 0.020681777189923645,
+      "grad_norm": 2.0605741545982994,
+      "learning_rate": 2.7519999999999997e-06,
+      "loss": 1.0094,
+      "num_input_tokens_seen": 3572085,
+      "step": 172
+    },
+    {
+      "epoch": 0.020802020080562738,
+      "grad_norm": 1.7783133406681224,
+      "learning_rate": 2.7679999999999996e-06,
+      "loss": 0.8672,
+      "num_input_tokens_seen": 3593105,
+      "step": 173
+    },
+    {
+      "epoch": 0.020922262971201828,
+      "grad_norm": 2.7612219369538056,
+      "learning_rate": 2.7839999999999995e-06,
+      "loss": 0.9931,
+      "num_input_tokens_seen": 3611790,
+      "step": 174
+    },
+    {
+      "epoch": 0.021042505861840918,
+      "grad_norm": 1.7932511617528923,
+      "learning_rate": 2.8e-06,
+      "loss": 0.9382,
+      "num_input_tokens_seen": 3628505,
+      "step": 175
+    },
+    {
+      "epoch": 0.02116274875248001,
+      "grad_norm": 2.2964324675622207,
+      "learning_rate": 2.816e-06,
+      "loss": 0.9283,
+      "num_input_tokens_seen": 3647480,
+      "step": 176
+    },
+    {
+      "epoch": 0.0212829916431191,
+      "grad_norm": 8.618171898463725,
+      "learning_rate": 2.8319999999999997e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 3667905,
+      "step": 177
+    },
+    {
+      "epoch": 0.02140323453375819,
+      "grad_norm": 2.9829833791235507,
+      "learning_rate": 2.8479999999999997e-06,
+      "loss": 1.0223,
+      "num_input_tokens_seen": 3681600,
+      "step": 178
+    },
+    {
+      "epoch": 0.021523477424397284,
+      "grad_norm": 3.2757078309522942,
+      "learning_rate": 2.8639999999999996e-06,
+      "loss": 0.964,
+      "num_input_tokens_seen": 3694815,
+      "step": 179
+    },
+    {
+      "epoch": 0.021643720315036374,
+      "grad_norm": 5.576198688542104,
+      "learning_rate": 2.88e-06,
+      "loss": 0.9364,
+      "num_input_tokens_seen": 3712635,
+      "step": 180
+    },
+    {
+      "epoch": 0.021763963205675464,
+      "grad_norm": 1.8422395700543908,
+      "learning_rate": 2.896e-06,
+      "loss": 0.9657,
+      "num_input_tokens_seen": 3731985,
+      "step": 181
+    },
+    {
+      "epoch": 0.021884206096314557,
+      "grad_norm": 2.260902259036363,
+      "learning_rate": 2.912e-06,
+      "loss": 0.9113,
+      "num_input_tokens_seen": 3751475,
+      "step": 182
+    },
+    {
+      "epoch": 0.022004448986953647,
+      "grad_norm": 1.7065035724930062,
+      "learning_rate": 2.9279999999999997e-06,
+      "loss": 0.8515,
+      "num_input_tokens_seen": 3771250,
+      "step": 183
+    },
+    {
+      "epoch": 0.022124691877592736,
+      "grad_norm": 2.1608266208153526,
+      "learning_rate": 2.9439999999999997e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 3789620,
+      "step": 184
+    },
+    {
+      "epoch": 0.02224493476823183,
+      "grad_norm": 2.151624959169906,
+      "learning_rate": 2.96e-06,
+      "loss": 0.884,
+      "num_input_tokens_seen": 3810290,
+      "step": 185
+    },
+    {
+      "epoch": 0.02236517765887092,
+      "grad_norm": 2.7629721655525215,
+      "learning_rate": 2.976e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 3828945,
+      "step": 186
+    },
+    {
+      "epoch": 0.02248542054951001,
+      "grad_norm": 2.3923748982885664,
+      "learning_rate": 2.992e-06,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 3852950,
+      "step": 187
+    },
+    {
+      "epoch": 0.022605663440149103,
+      "grad_norm": 2.0384417946723525,
+      "learning_rate": 3.008e-06,
+      "loss": 0.8736,
+      "num_input_tokens_seen": 3871625,
+      "step": 188
+    },
+    {
+      "epoch": 0.022725906330788193,
+      "grad_norm": 2.0998931061117774,
+      "learning_rate": 3.0239999999999998e-06,
+      "loss": 0.9131,
+      "num_input_tokens_seen": 3891910,
+      "step": 189
+    },
+    {
+      "epoch": 0.022846149221427282,
+      "grad_norm": 2.6885335566194954,
+      "learning_rate": 3.0399999999999997e-06,
+      "loss": 0.806,
+      "num_input_tokens_seen": 3912535,
+      "step": 190
+    },
+    {
+      "epoch": 0.022966392112066376,
+      "grad_norm": 2.528212138727624,
+      "learning_rate": 3.056e-06,
+      "loss": 0.925,
+      "num_input_tokens_seen": 3930495,
+      "step": 191
+    },
+    {
+      "epoch": 0.023086635002705466,
+      "grad_norm": 1.9604148235959722,
+      "learning_rate": 3.072e-06,
+      "loss": 0.6824,
+      "num_input_tokens_seen": 3989680,
+      "step": 192
+    },
+    {
+      "epoch": 0.023206877893344555,
+      "grad_norm": 2.199692256384985,
+      "learning_rate": 3.088e-06,
+      "loss": 0.875,
+      "num_input_tokens_seen": 4007710,
+      "step": 193
+    },
+    {
+      "epoch": 0.02332712078398365,
+      "grad_norm": 2.0800724959674075,
+      "learning_rate": 3.104e-06,
+      "loss": 0.8934,
+      "num_input_tokens_seen": 4028025,
+      "step": 194
+    },
+    {
+      "epoch": 0.02344736367462274,
+      "grad_norm": 2.501253824161815,
+      "learning_rate": 3.1199999999999998e-06,
+      "loss": 0.8741,
+      "num_input_tokens_seen": 4047170,
+      "step": 195
+    },
+    {
+      "epoch": 0.02356760656526183,
+      "grad_norm": 2.737793027306602,
+      "learning_rate": 3.136e-06,
+      "loss": 0.8622,
+      "num_input_tokens_seen": 4063890,
+      "step": 196
+    },
+    {
+      "epoch": 0.02368784945590092,
+      "grad_norm": 2.6920416062813595,
+      "learning_rate": 3.152e-06,
+      "loss": 0.8675,
+      "num_input_tokens_seen": 4085950,
+      "step": 197
+    },
+    {
+      "epoch": 0.02380809234654001,
+      "grad_norm": 2.385608531737704,
+      "learning_rate": 3.168e-06,
+      "loss": 0.8189,
+      "num_input_tokens_seen": 4105990,
+      "step": 198
+    },
+    {
+      "epoch": 0.0239283352371791,
+      "grad_norm": 4.449159379014996,
+      "learning_rate": 3.184e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 4123540,
+      "step": 199
+    },
+    {
+      "epoch": 0.02404857812781819,
+      "grad_norm": 3.036401726967315,
+      "learning_rate": 3.2e-06,
+      "loss": 0.8932,
+      "num_input_tokens_seen": 4144640,
+      "step": 200
+    },
+    {
+      "epoch": 0.024168821018457284,
+      "grad_norm": 4.079062602335863,
+      "learning_rate": 3.216e-06,
+      "loss": 0.8859,
+      "num_input_tokens_seen": 4161890,
+      "step": 201
+    },
+    {
+      "epoch": 0.024289063909096374,
+      "grad_norm": 2.4279335564897044,
+      "learning_rate": 3.232e-06,
+      "loss": 0.9,
+      "num_input_tokens_seen": 4181210,
+      "step": 202
+    },
+    {
+      "epoch": 0.024409306799735464,
+      "grad_norm": 2.1161649705444936,
+      "learning_rate": 3.248e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 4199145,
+      "step": 203
+    },
+    {
+      "epoch": 0.024529549690374557,
+      "grad_norm": 3.388974815324258,
+      "learning_rate": 3.2639999999999996e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 4218915,
+      "step": 204
+    },
+    {
+      "epoch": 0.024649792581013647,
+      "grad_norm": 2.617284163672077,
+      "learning_rate": 3.2799999999999995e-06,
+      "loss": 0.9179,
+      "num_input_tokens_seen": 4238330,
+      "step": 205
+    },
+    {
+      "epoch": 0.024770035471652737,
+      "grad_norm": 2.1494673801549276,
+      "learning_rate": 3.296e-06,
+      "loss": 0.9498,
+      "num_input_tokens_seen": 4260270,
+      "step": 206
+    },
+    {
+      "epoch": 0.02489027836229183,
+      "grad_norm": 3.1290271939230396,
+      "learning_rate": 3.312e-06,
+      "loss": 0.9356,
+      "num_input_tokens_seen": 4279645,
+      "step": 207
+    },
+    {
+      "epoch": 0.02501052125293092,
+      "grad_norm": 1.835330343206158,
+      "learning_rate": 3.3279999999999997e-06,
+      "loss": 0.8781,
+      "num_input_tokens_seen": 4299870,
+      "step": 208
+    },
+    {
+      "epoch": 0.02513076414357001,
+      "grad_norm": 2.7546319081587716,
+      "learning_rate": 3.3439999999999997e-06,
+      "loss": 0.9351,
+      "num_input_tokens_seen": 4316435,
+      "step": 209
+    },
+    {
+      "epoch": 0.025251007034209103,
+      "grad_norm": 1.992628943271235,
+      "learning_rate": 3.3599999999999996e-06,
+      "loss": 0.947,
+      "num_input_tokens_seen": 4332735,
+      "step": 210
+    },
+    {
+      "epoch": 0.025371249924848193,
+      "grad_norm": 2.4268098972266694,
+      "learning_rate": 3.3759999999999995e-06,
+      "loss": 0.942,
+      "num_input_tokens_seen": 4349320,
+      "step": 211
+    },
+    {
+      "epoch": 0.025491492815487283,
+      "grad_norm": 2.1744302593160585,
+      "learning_rate": 3.392e-06,
+      "loss": 0.8166,
+      "num_input_tokens_seen": 4368480,
+      "step": 212
+    },
+    {
+      "epoch": 0.025611735706126376,
+      "grad_norm": 2.8711878058016747,
+      "learning_rate": 3.408e-06,
+      "loss": 0.9028,
+      "num_input_tokens_seen": 4387470,
+      "step": 213
+    },
+    {
+      "epoch": 0.025731978596765466,
+      "grad_norm": 2.269168587715047,
+      "learning_rate": 3.4239999999999997e-06,
+      "loss": 0.931,
+      "num_input_tokens_seen": 4404795,
+      "step": 214
+    },
+    {
+      "epoch": 0.025852221487404556,
+      "grad_norm": 1.6805178706742634,
+      "learning_rate": 3.4399999999999997e-06,
+      "loss": 0.9555,
+      "num_input_tokens_seen": 4424830,
+      "step": 215
+    },
+    {
+      "epoch": 0.02597246437804365,
+      "grad_norm": 2.0318722960568776,
+      "learning_rate": 3.4559999999999996e-06,
+      "loss": 0.8713,
+      "num_input_tokens_seen": 4445455,
+      "step": 216
+    },
+    {
+      "epoch": 0.02609270726868274,
+      "grad_norm": 1.7359027883562923,
+      "learning_rate": 3.472e-06,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 4466285,
+      "step": 217
+    },
+    {
+      "epoch": 0.02621295015932183,
+      "grad_norm": 2.0583098476347024,
+      "learning_rate": 3.488e-06,
+      "loss": 0.8926,
+      "num_input_tokens_seen": 4485320,
+      "step": 218
+    },
+    {
+      "epoch": 0.026333193049960922,
+      "grad_norm": 7.091301056826507,
+      "learning_rate": 3.504e-06,
+      "loss": 0.7684,
+      "num_input_tokens_seen": 4504175,
+      "step": 219
+    },
+    {
+      "epoch": 0.026453435940600012,
+      "grad_norm": 2.115494348025387,
+      "learning_rate": 3.5199999999999998e-06,
+      "loss": 0.84,
+      "num_input_tokens_seen": 4521730,
+      "step": 220
+    },
+    {
+      "epoch": 0.026573678831239102,
+      "grad_norm": 3.0038967169121604,
+      "learning_rate": 3.5359999999999997e-06,
+      "loss": 0.9121,
+      "num_input_tokens_seen": 4538640,
+      "step": 221
+    },
+    {
+      "epoch": 0.026693921721878195,
+      "grad_norm": 2.408591637044203,
+      "learning_rate": 3.552e-06,
+      "loss": 0.869,
+      "num_input_tokens_seen": 4555180,
+      "step": 222
+    },
+    {
+      "epoch": 0.026814164612517285,
+      "grad_norm": 3.1693145932686453,
+      "learning_rate": 3.568e-06,
+      "loss": 0.801,
+      "num_input_tokens_seen": 4568380,
+      "step": 223
+    },
+    {
+      "epoch": 0.026934407503156375,
+      "grad_norm": 1.838139943358324,
+      "learning_rate": 3.584e-06,
+      "loss": 0.9592,
+      "num_input_tokens_seen": 4589265,
+      "step": 224
+    },
+    {
+      "epoch": 0.027054650393795468,
+      "grad_norm": 3.157467581599869,
+      "learning_rate": 3.6e-06,
+      "loss": 0.8215,
+      "num_input_tokens_seen": 4609295,
+      "step": 225
+    },
+    {
+      "epoch": 0.027174893284434558,
+      "grad_norm": 1.9012003366482346,
+      "learning_rate": 3.6159999999999998e-06,
+      "loss": 0.8722,
+      "num_input_tokens_seen": 4632785,
+      "step": 226
+    },
+    {
+      "epoch": 0.027295136175073648,
+      "grad_norm": 3.346106484239129,
+      "learning_rate": 3.632e-06,
+      "loss": 0.7924,
+      "num_input_tokens_seen": 4652505,
+      "step": 227
+    },
+    {
+      "epoch": 0.02741537906571274,
+      "grad_norm": 2.1370629666543968,
+      "learning_rate": 3.648e-06,
+      "loss": 0.6666,
+      "num_input_tokens_seen": 4712020,
+      "step": 228
+    },
+    {
+      "epoch": 0.02753562195635183,
+      "grad_norm": 3.7176260038177706,
+      "learning_rate": 3.664e-06,
+      "loss": 0.8575,
+      "num_input_tokens_seen": 4731715,
+      "step": 229
+    },
+    {
+      "epoch": 0.02765586484699092,
+      "grad_norm": 2.505145173064042,
+      "learning_rate": 3.68e-06,
+      "loss": 0.8268,
+      "num_input_tokens_seen": 4753060,
+      "step": 230
+    },
+    {
+      "epoch": 0.027776107737630014,
+      "grad_norm": 2.192119201297155,
+      "learning_rate": 3.696e-06,
+      "loss": 0.94,
+      "num_input_tokens_seen": 4770795,
+      "step": 231
+    },
+    {
+      "epoch": 0.027896350628269104,
+      "grad_norm": 2.0026943252555833,
+      "learning_rate": 3.712e-06,
+      "loss": 0.8762,
+      "num_input_tokens_seen": 4792800,
+      "step": 232
+    },
+    {
+      "epoch": 0.028016593518908194,
+      "grad_norm": 2.71220011604548,
+      "learning_rate": 3.728e-06,
+      "loss": 0.8741,
+      "num_input_tokens_seen": 4810615,
+      "step": 233
+    },
+    {
+      "epoch": 0.028136836409547287,
+      "grad_norm": 3.1390203467971576,
+      "learning_rate": 3.744e-06,
+      "loss": 0.8368,
+      "num_input_tokens_seen": 4827680,
+      "step": 234
+    },
+    {
+      "epoch": 0.028257079300186377,
+      "grad_norm": 2.3865331315851193,
+      "learning_rate": 3.7599999999999996e-06,
+      "loss": 0.8407,
+      "num_input_tokens_seen": 4847165,
+      "step": 235
+    },
+    {
+      "epoch": 0.028377322190825467,
+      "grad_norm": 2.3064731188423213,
+      "learning_rate": 3.7759999999999995e-06,
+      "loss": 0.8599,
+      "num_input_tokens_seen": 4867025,
+      "step": 236
+    },
+    {
+      "epoch": 0.02849756508146456,
+      "grad_norm": 2.4143007919638677,
+      "learning_rate": 3.7919999999999994e-06,
+      "loss": 0.9558,
+      "num_input_tokens_seen": 4883725,
+      "step": 237
+    },
+    {
+      "epoch": 0.02861780797210365,
+      "grad_norm": 2.160624400580004,
+      "learning_rate": 3.808e-06,
+      "loss": 0.8349,
+      "num_input_tokens_seen": 4901435,
+      "step": 238
+    },
+    {
+      "epoch": 0.02873805086274274,
+      "grad_norm": 2.4326143267020766,
+      "learning_rate": 3.823999999999999e-06,
+      "loss": 0.9177,
+      "num_input_tokens_seen": 4920435,
+      "step": 239
+    },
+    {
+      "epoch": 0.028858293753381833,
+      "grad_norm": 2.358785591466701,
+      "learning_rate": 3.84e-06,
+      "loss": 0.8991,
+      "num_input_tokens_seen": 4937480,
+      "step": 240
+    },
+    {
+      "epoch": 0.028978536644020923,
+      "grad_norm": 2.088270749436349,
+      "learning_rate": 3.856e-06,
+      "loss": 0.8987,
+      "num_input_tokens_seen": 4955485,
+      "step": 241
+    },
+    {
+      "epoch": 0.029098779534660012,
+      "grad_norm": 2.638493191469904,
+      "learning_rate": 3.8719999999999995e-06,
+      "loss": 0.9405,
+      "num_input_tokens_seen": 4974865,
+      "step": 242
+    },
+    {
+      "epoch": 0.029219022425299106,
+      "grad_norm": 3.1103169614571122,
+      "learning_rate": 3.888e-06,
+      "loss": 0.9672,
+      "num_input_tokens_seen": 4992340,
+      "step": 243
+    },
+    {
+      "epoch": 0.029339265315938196,
+      "grad_norm": 3.2183741580010072,
+      "learning_rate": 3.903999999999999e-06,
+      "loss": 0.8611,
+      "num_input_tokens_seen": 5010400,
+      "step": 244
+    },
+    {
+      "epoch": 0.029459508206577285,
+      "grad_norm": 2.0765174747318538,
+      "learning_rate": 3.92e-06,
+      "loss": 0.9179,
+      "num_input_tokens_seen": 5028005,
+      "step": 245
+    },
+    {
+      "epoch": 0.02957975109721638,
+      "grad_norm": 9.100222361512776,
+      "learning_rate": 3.936e-06,
+      "loss": 0.8626,
+      "num_input_tokens_seen": 5047405,
+      "step": 246
+    },
+    {
+      "epoch": 0.02969999398785547,
+      "grad_norm": 1.7199314276763102,
+      "learning_rate": 3.952e-06,
+      "loss": 0.8203,
+      "num_input_tokens_seen": 5067665,
+      "step": 247
+    },
+    {
+      "epoch": 0.02982023687849456,
+      "grad_norm": 2.5520564078883865,
+      "learning_rate": 3.968e-06,
+      "loss": 0.8387,
+      "num_input_tokens_seen": 5089190,
+      "step": 248
+    },
+    {
+      "epoch": 0.02994047976913365,
+      "grad_norm": 2.1312146138094055,
+      "learning_rate": 3.9839999999999995e-06,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 5109790,
+      "step": 249
+    },
+    {
+      "epoch": 0.03006072265977274,
+      "grad_norm": 3.8187510687502186,
+      "learning_rate": 4e-06,
+      "loss": 0.8636,
+      "num_input_tokens_seen": 5129345,
+      "step": 250
+    },
+    {
+      "epoch": 0.03018096555041183,
+      "grad_norm": 3.0551705792419606,
+      "learning_rate": 3.999999848300794e-06,
+      "loss": 0.9252,
+      "num_input_tokens_seen": 5148050,
+      "step": 251
+    },
+    {
+      "epoch": 0.030301208441050925,
+      "grad_norm": 3.342254164173398,
+      "learning_rate": 3.999999393203203e-06,
+      "loss": 0.9036,
+      "num_input_tokens_seen": 5170180,
+      "step": 252
+    },
+    {
+      "epoch": 0.030421451331690014,
+      "grad_norm": 1.801950593877542,
+      "learning_rate": 3.999998634707293e-06,
+      "loss": 0.8607,
+      "num_input_tokens_seen": 5189450,
+      "step": 253
+    },
+    {
+      "epoch": 0.030541694222329104,
+      "grad_norm": 3.050840993158165,
+      "learning_rate": 3.999997572813182e-06,
+      "loss": 0.9817,
+      "num_input_tokens_seen": 5206980,
+      "step": 254
+    },
+    {
+      "epoch": 0.030661937112968194,
+      "grad_norm": 2.241968008255119,
+      "learning_rate": 3.999996207521028e-06,
+      "loss": 0.8901,
+      "num_input_tokens_seen": 5225410,
+      "step": 255
+    },
+    {
+      "epoch": 0.030782180003607287,
+      "grad_norm": 3.246824562679245,
+      "learning_rate": 3.999994538831039e-06,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 5241715,
+      "step": 256
+    },
+    {
+      "epoch": 0.030902422894246377,
+      "grad_norm": 2.600954218048441,
+      "learning_rate": 3.99999256674347e-06,
+      "loss": 0.8627,
+      "num_input_tokens_seen": 5261585,
+      "step": 257
+    },
+    {
+      "epoch": 0.031022665784885467,
+      "grad_norm": 1.6355519479195977,
+      "learning_rate": 3.999990291258618e-06,
+      "loss": 0.5823,
+      "num_input_tokens_seen": 5319995,
+      "step": 258
+    },
+    {
+      "epoch": 0.03114290867552456,
+      "grad_norm": 2.603380132499256,
+      "learning_rate": 3.999987712376829e-06,
+      "loss": 0.8719,
+      "num_input_tokens_seen": 5338035,
+      "step": 259
+    },
+    {
+      "epoch": 0.031263151566163654,
+      "grad_norm": 2.0922252056673805,
+      "learning_rate": 3.999984830098494e-06,
+      "loss": 0.8356,
+      "num_input_tokens_seen": 5357335,
+      "step": 260
+    },
+    {
+      "epoch": 0.03138339445680274,
+      "grad_norm": 2.7727989478822352,
+      "learning_rate": 3.999981644424051e-06,
+      "loss": 0.9976,
+      "num_input_tokens_seen": 5371855,
+      "step": 261
+    },
+    {
+      "epoch": 0.03150363734744183,
+      "grad_norm": 2.262770203824872,
+      "learning_rate": 3.999978155353982e-06,
+      "loss": 0.8793,
+      "num_input_tokens_seen": 5388720,
+      "step": 262
+    },
+    {
+      "epoch": 0.03162388023808092,
+      "grad_norm": 2.279218115073108,
+      "learning_rate": 3.9999743628888186e-06,
+      "loss": 0.8065,
+      "num_input_tokens_seen": 5410230,
+      "step": 263
+    },
+    {
+      "epoch": 0.03174412312872001,
+      "grad_norm": 2.193294935735661,
+      "learning_rate": 3.999970267029133e-06,
+      "loss": 0.8987,
+      "num_input_tokens_seen": 5428910,
+      "step": 264
+    },
+    {
+      "epoch": 0.0318643660193591,
+      "grad_norm": 4.773916362267976,
+      "learning_rate": 3.999965867775548e-06,
+      "loss": 0.8056,
+      "num_input_tokens_seen": 5449025,
+      "step": 265
+    },
+    {
+      "epoch": 0.0319846089099982,
+      "grad_norm": 3.605386742796827,
+      "learning_rate": 3.9999611651287315e-06,
+      "loss": 0.8809,
+      "num_input_tokens_seen": 5466900,
+      "step": 266
+    },
+    {
+      "epoch": 0.03210485180063729,
+      "grad_norm": 2.9377025853461185,
+      "learning_rate": 3.999956159089396e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 5484070,
+      "step": 267
+    },
+    {
+      "epoch": 0.03222509469127638,
+      "grad_norm": 2.71557954022971,
+      "learning_rate": 3.999950849658302e-06,
+      "loss": 0.8119,
+      "num_input_tokens_seen": 5502710,
+      "step": 268
+    },
+    {
+      "epoch": 0.03234533758191547,
+      "grad_norm": 2.6379679571760133,
+      "learning_rate": 3.999945236836254e-06,
+      "loss": 0.8534,
+      "num_input_tokens_seen": 5521395,
+      "step": 269
+    },
+    {
+      "epoch": 0.03246558047255456,
+      "grad_norm": 3.064483366059667,
+      "learning_rate": 3.999939320624103e-06,
+      "loss": 0.9626,
+      "num_input_tokens_seen": 5536265,
+      "step": 270
+    },
+    {
+      "epoch": 0.03258582336319365,
+      "grad_norm": 1.8722943900274187,
+      "learning_rate": 3.999933101022749e-06,
+      "loss": 0.9041,
+      "num_input_tokens_seen": 5556390,
+      "step": 271
+    },
+    {
+      "epoch": 0.032706066253832745,
+      "grad_norm": 2.0240204821360033,
+      "learning_rate": 3.999926578033132e-06,
+      "loss": 0.8716,
+      "num_input_tokens_seen": 5575925,
+      "step": 272
+    },
+    {
+      "epoch": 0.032826309144471835,
+      "grad_norm": 2.985965330324514,
+      "learning_rate": 3.999919751656244e-06,
+      "loss": 0.6461,
+      "num_input_tokens_seen": 5602545,
+      "step": 273
+    },
+    {
+      "epoch": 0.032946552035110925,
+      "grad_norm": 2.428543330489089,
+      "learning_rate": 3.9999126218931195e-06,
+      "loss": 0.7741,
+      "num_input_tokens_seen": 5620300,
+      "step": 274
+    },
+    {
+      "epoch": 0.033066794925750015,
+      "grad_norm": 2.3527258567159324,
+      "learning_rate": 3.99990518874484e-06,
+      "loss": 0.9148,
+      "num_input_tokens_seen": 5636460,
+      "step": 275
+    },
+    {
+      "epoch": 0.033187037816389105,
+      "grad_norm": 2.3683697736305787,
+      "learning_rate": 3.999897452212534e-06,
+      "loss": 0.9356,
+      "num_input_tokens_seen": 5653510,
+      "step": 276
+    },
+    {
+      "epoch": 0.033307280707028195,
+      "grad_norm": 2.2864172720411413,
+      "learning_rate": 3.999889412297374e-06,
+      "loss": 1.005,
+      "num_input_tokens_seen": 5672655,
+      "step": 277
+    },
+    {
+      "epoch": 0.03342752359766729,
+      "grad_norm": 2.112918349242022,
+      "learning_rate": 3.999881069000581e-06,
+      "loss": 0.8048,
+      "num_input_tokens_seen": 5692105,
+      "step": 278
+    },
+    {
+      "epoch": 0.03354776648830638,
+      "grad_norm": 2.577115421343023,
+      "learning_rate": 3.99987242232342e-06,
+      "loss": 0.895,
+      "num_input_tokens_seen": 5706830,
+      "step": 279
+    },
+    {
+      "epoch": 0.03366800937894547,
+      "grad_norm": 3.1148692821528727,
+      "learning_rate": 3.9998634722672026e-06,
+      "loss": 0.8087,
+      "num_input_tokens_seen": 5726605,
+      "step": 280
+    },
+    {
+      "epoch": 0.03378825226958456,
+      "grad_norm": 2.206459668735334,
+      "learning_rate": 3.999854218833286e-06,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 5747145,
+      "step": 281
+    },
+    {
+      "epoch": 0.03390849516022365,
+      "grad_norm": 3.7228930099786677,
+      "learning_rate": 3.999844662023075e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 5766740,
+      "step": 282
+    },
+    {
+      "epoch": 0.03402873805086274,
+      "grad_norm": 1.948517128806398,
+      "learning_rate": 3.999834801838018e-06,
+      "loss": 0.9347,
+      "num_input_tokens_seen": 5785440,
+      "step": 283
+    },
+    {
+      "epoch": 0.03414898094150183,
+      "grad_norm": 2.294143880301539,
+      "learning_rate": 3.9998246382796115e-06,
+      "loss": 0.7589,
+      "num_input_tokens_seen": 5804740,
+      "step": 284
+    },
+    {
+      "epoch": 0.03426922383214093,
+      "grad_norm": 2.391251086777688,
+      "learning_rate": 3.999814171349399e-06,
+      "loss": 0.9201,
+      "num_input_tokens_seen": 5822320,
+      "step": 285
+    },
+    {
+      "epoch": 0.03438946672278002,
+      "grad_norm": 2.1133667002300442,
+      "learning_rate": 3.9998034010489655e-06,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 5845730,
+      "step": 286
+    },
+    {
+      "epoch": 0.03450970961341911,
+      "grad_norm": 3.314309450055164,
+      "learning_rate": 3.999792327379946e-06,
+      "loss": 0.7698,
+      "num_input_tokens_seen": 5864825,
+      "step": 287
+    },
+    {
+      "epoch": 0.034629952504058197,
+      "grad_norm": 2.949741690829669,
+      "learning_rate": 3.999780950344021e-06,
+      "loss": 0.9856,
+      "num_input_tokens_seen": 5882735,
+      "step": 288
+    },
+    {
+      "epoch": 0.034750195394697286,
+      "grad_norm": 1.868232974093652,
+      "learning_rate": 3.999769269942916e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 5902495,
+      "step": 289
+    },
+    {
+      "epoch": 0.034870438285336376,
+      "grad_norm": 1.9171755697676127,
+      "learning_rate": 3.999757286178402e-06,
+      "loss": 0.8277,
+      "num_input_tokens_seen": 5924650,
+      "step": 290
+    },
+    {
+      "epoch": 0.03499068117597547,
+      "grad_norm": 1.91351843615992,
+      "learning_rate": 3.999744999052299e-06,
+      "loss": 0.9143,
+      "num_input_tokens_seen": 5945760,
+      "step": 291
+    },
+    {
+      "epoch": 0.03511092406661456,
+      "grad_norm": 1.4049375044033747,
+      "learning_rate": 3.9997324085664675e-06,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 6005710,
+      "step": 292
+    },
+    {
+      "epoch": 0.03523116695725365,
+      "grad_norm": 2.8311938625552555,
+      "learning_rate": 3.999719514722821e-06,
+      "loss": 0.9253,
+      "num_input_tokens_seen": 6025560,
+      "step": 293
+    },
+    {
+      "epoch": 0.03535140984789274,
+      "grad_norm": 2.177330379393852,
+      "learning_rate": 3.999706317523314e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 6043840,
+      "step": 294
+    },
+    {
+      "epoch": 0.03547165273853183,
+      "grad_norm": 1.9078416401645966,
+      "learning_rate": 3.999692816969948e-06,
+      "loss": 0.8689,
+      "num_input_tokens_seen": 6063095,
+      "step": 295
+    },
+    {
+      "epoch": 0.03559189562917092,
+      "grad_norm": 1.3296035902349783,
+      "learning_rate": 3.999679013064772e-06,
+      "loss": 0.717,
+      "num_input_tokens_seen": 6129560,
+      "step": 296
+    },
+    {
+      "epoch": 0.03571213851981002,
+      "grad_norm": 2.5049054173091,
+      "learning_rate": 3.99966490580988e-06,
+      "loss": 0.8737,
+      "num_input_tokens_seen": 6146640,
+      "step": 297
+    },
+    {
+      "epoch": 0.03583238141044911,
+      "grad_norm": 5.201711075237415,
+      "learning_rate": 3.999650495207411e-06,
+      "loss": 0.6842,
+      "num_input_tokens_seen": 6172385,
+      "step": 298
+    },
+    {
+      "epoch": 0.0359526243010882,
+      "grad_norm": 2.4201328370170963,
+      "learning_rate": 3.999635781259553e-06,
+      "loss": 0.9094,
+      "num_input_tokens_seen": 6187370,
+      "step": 299
+    },
+    {
+      "epoch": 0.03607286719172729,
+      "grad_norm": 1.3336515001859994,
+      "learning_rate": 3.999620763968535e-06,
+      "loss": 0.5733,
+      "num_input_tokens_seen": 6245965,
+      "step": 300
+    },
+    {
+      "epoch": 0.03619311008236638,
+      "grad_norm": 2.287233399432092,
+      "learning_rate": 3.999605443336638e-06,
+      "loss": 0.8769,
+      "num_input_tokens_seen": 6267815,
+      "step": 301
+    },
+    {
+      "epoch": 0.03631335297300547,
+      "grad_norm": 2.972305278115536,
+      "learning_rate": 3.999589819366185e-06,
+      "loss": 0.9133,
+      "num_input_tokens_seen": 6281325,
+      "step": 302
+    },
+    {
+      "epoch": 0.036433595863644565,
+      "grad_norm": 1.9292914270325559,
+      "learning_rate": 3.999573892059547e-06,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 6300175,
+      "step": 303
+    },
+    {
+      "epoch": 0.036553838754283655,
+      "grad_norm": 2.211612608083357,
+      "learning_rate": 3.999557661419138e-06,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 6320045,
+      "step": 304
+    },
+    {
+      "epoch": 0.036674081644922744,
+      "grad_norm": 2.091846611025725,
+      "learning_rate": 3.9995411274474225e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 6339045,
+      "step": 305
+    },
+    {
+      "epoch": 0.036794324535561834,
+      "grad_norm": 2.1567298014040865,
+      "learning_rate": 3.999524290146908e-06,
+      "loss": 0.8335,
+      "num_input_tokens_seen": 6358970,
+      "step": 306
+    },
+    {
+      "epoch": 0.036914567426200924,
+      "grad_norm": 2.545724521004469,
+      "learning_rate": 3.9995071495201485e-06,
+      "loss": 0.9203,
+      "num_input_tokens_seen": 6375795,
+      "step": 307
+    },
+    {
+      "epoch": 0.037034810316840014,
+      "grad_norm": 2.411243479714235,
+      "learning_rate": 3.999489705569744e-06,
+      "loss": 0.9852,
+      "num_input_tokens_seen": 6393215,
+      "step": 308
+    },
+    {
+      "epoch": 0.03715505320747911,
+      "grad_norm": 2.0455696199630515,
+      "learning_rate": 3.999471958298341e-06,
+      "loss": 0.8827,
+      "num_input_tokens_seen": 6411845,
+      "step": 309
+    },
+    {
+      "epoch": 0.0372752960981182,
+      "grad_norm": 1.7629648877215305,
+      "learning_rate": 3.999453907708631e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 6433970,
+      "step": 310
+    },
+    {
+      "epoch": 0.03739553898875729,
+      "grad_norm": 1.7179807281318327,
+      "learning_rate": 3.999435553803353e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 6453090,
+      "step": 311
+    },
+    {
+      "epoch": 0.03751578187939638,
+      "grad_norm": 2.9646680862818555,
+      "learning_rate": 3.999416896585292e-06,
+      "loss": 0.8518,
+      "num_input_tokens_seen": 6469840,
+      "step": 312
+    },
+    {
+      "epoch": 0.03763602477003547,
+      "grad_norm": 4.713428310853323,
+      "learning_rate": 3.9993979360572775e-06,
+      "loss": 0.8707,
+      "num_input_tokens_seen": 6489700,
+      "step": 313
+    },
+    {
+      "epoch": 0.03775626766067456,
+      "grad_norm": 5.195056705690607,
+      "learning_rate": 3.999378672222185e-06,
+      "loss": 0.8437,
+      "num_input_tokens_seen": 6507205,
+      "step": 314
+    },
+    {
+      "epoch": 0.03787651055131366,
+      "grad_norm": 2.2639651165503487,
+      "learning_rate": 3.9993591050829385e-06,
+      "loss": 0.8406,
+      "num_input_tokens_seen": 6524790,
+      "step": 315
+    },
+    {
+      "epoch": 0.037996753441952746,
+      "grad_norm": 1.8829600150311268,
+      "learning_rate": 3.999339234642506e-06,
+      "loss": 0.7988,
+      "num_input_tokens_seen": 6544260,
+      "step": 316
+    },
+    {
+      "epoch": 0.038116996332591836,
+      "grad_norm": 2.091952206463273,
+      "learning_rate": 3.9993190609038994e-06,
+      "loss": 0.8488,
+      "num_input_tokens_seen": 6562745,
+      "step": 317
+    },
+    {
+      "epoch": 0.038237239223230926,
+      "grad_norm": 1.8130077445805735,
+      "learning_rate": 3.999298583870182e-06,
+      "loss": 0.8513,
+      "num_input_tokens_seen": 6582050,
+      "step": 318
+    },
+    {
+      "epoch": 0.038357482113870016,
+      "grad_norm": 2.4000573010845656,
+      "learning_rate": 3.999277803544458e-06,
+      "loss": 0.7909,
+      "num_input_tokens_seen": 6601925,
+      "step": 319
+    },
+    {
+      "epoch": 0.038477725004509106,
+      "grad_norm": 1.044544796224368,
+      "learning_rate": 3.999256719929882e-06,
+      "loss": 0.6733,
+      "num_input_tokens_seen": 6662920,
+      "step": 320
+    },
+    {
+      "epoch": 0.0385979678951482,
+      "grad_norm": 1.4062611867353632,
+      "learning_rate": 3.999235333029651e-06,
+      "loss": 0.7547,
+      "num_input_tokens_seen": 6716580,
+      "step": 321
+    },
+    {
+      "epoch": 0.03871821078578729,
+      "grad_norm": 1.7787580536231962,
+      "learning_rate": 3.999213642847009e-06,
+      "loss": 0.8153,
+      "num_input_tokens_seen": 6736885,
+      "step": 322
+    },
+    {
+      "epoch": 0.03883845367642638,
+      "grad_norm": 2.3954289062976977,
+      "learning_rate": 3.999191649385247e-06,
+      "loss": 0.9194,
+      "num_input_tokens_seen": 6757780,
+      "step": 323
+    },
+    {
+      "epoch": 0.03895869656706547,
+      "grad_norm": 1.0371840965369292,
+      "learning_rate": 3.999169352647702e-06,
+      "loss": 0.6407,
+      "num_input_tokens_seen": 6818680,
+      "step": 324
+    },
+    {
+      "epoch": 0.03907893945770456,
+      "grad_norm": 1.8061138549551585,
+      "learning_rate": 3.999146752637755e-06,
+      "loss": 0.8295,
+      "num_input_tokens_seen": 6839445,
+      "step": 325
+    },
+    {
+      "epoch": 0.03919918234834365,
+      "grad_norm": 2.3364462339647702,
+      "learning_rate": 3.999123849358836e-06,
+      "loss": 0.9128,
+      "num_input_tokens_seen": 6856830,
+      "step": 326
+    },
+    {
+      "epoch": 0.03931942523898275,
+      "grad_norm": 2.173017484126889,
+      "learning_rate": 3.999100642814418e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 6876990,
+      "step": 327
+    },
+    {
+      "epoch": 0.03943966812962184,
+      "grad_norm": 2.2828147302825417,
+      "learning_rate": 3.999077133008022e-06,
+      "loss": 0.9068,
+      "num_input_tokens_seen": 6895295,
+      "step": 328
+    },
+    {
+      "epoch": 0.03955991102026093,
+      "grad_norm": 1.8399729475566025,
+      "learning_rate": 3.9990533199432145e-06,
+      "loss": 0.9204,
+      "num_input_tokens_seen": 6916510,
+      "step": 329
+    },
+    {
+      "epoch": 0.03968015391090002,
+      "grad_norm": 2.2392228130067715,
+      "learning_rate": 3.999029203623608e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 6933950,
+      "step": 330
+    },
+    {
+      "epoch": 0.03980039680153911,
+      "grad_norm": 1.9495961584579014,
+      "learning_rate": 3.99900478405286e-06,
+      "loss": 0.872,
+      "num_input_tokens_seen": 6952980,
+      "step": 331
+    },
+    {
+      "epoch": 0.0399206396921782,
+      "grad_norm": 3.880844829169628,
+      "learning_rate": 3.998980061234676e-06,
+      "loss": 0.8359,
+      "num_input_tokens_seen": 6970615,
+      "step": 332
+    },
+    {
+      "epoch": 0.040040882582817294,
+      "grad_norm": 2.782601678691449,
+      "learning_rate": 3.9989550351728055e-06,
+      "loss": 0.7717,
+      "num_input_tokens_seen": 6987265,
+      "step": 333
+    },
+    {
+      "epoch": 0.040161125473456384,
+      "grad_norm": 2.2663500631062363,
+      "learning_rate": 3.998929705871046e-06,
+      "loss": 0.8401,
+      "num_input_tokens_seen": 7004340,
+      "step": 334
+    },
+    {
+      "epoch": 0.040281368364095474,
+      "grad_norm": 2.498865321604322,
+      "learning_rate": 3.99890407333324e-06,
+      "loss": 0.8988,
+      "num_input_tokens_seen": 7022590,
+      "step": 335
+    },
+    {
+      "epoch": 0.040401611254734564,
+      "grad_norm": 1.6995487788026244,
+      "learning_rate": 3.998878137563275e-06,
+      "loss": 0.8713,
+      "num_input_tokens_seen": 7041860,
+      "step": 336
+    },
+    {
+      "epoch": 0.040521854145373654,
+      "grad_norm": 2.7293390219193214,
+      "learning_rate": 3.998851898565085e-06,
+      "loss": 0.8619,
+      "num_input_tokens_seen": 7061385,
+      "step": 337
+    },
+    {
+      "epoch": 0.04064209703601274,
+      "grad_norm": 3.684342630032326,
+      "learning_rate": 3.998825356342653e-06,
+      "loss": 0.8402,
+      "num_input_tokens_seen": 7081280,
+      "step": 338
+    },
+    {
+      "epoch": 0.04076233992665183,
+      "grad_norm": 3.0892320542520904,
+      "learning_rate": 3.998798510900003e-06,
+      "loss": 0.7385,
+      "num_input_tokens_seen": 7103800,
+      "step": 339
+    },
+    {
+      "epoch": 0.04088258281729093,
+      "grad_norm": 3.4568619704056247,
+      "learning_rate": 3.998771362241207e-06,
+      "loss": 0.8523,
+      "num_input_tokens_seen": 7123925,
+      "step": 340
+    },
+    {
+      "epoch": 0.04100282570793002,
+      "grad_norm": 2.1653040587591437,
+      "learning_rate": 3.998743910370385e-06,
+      "loss": 0.8912,
+      "num_input_tokens_seen": 7142505,
+      "step": 341
+    },
+    {
+      "epoch": 0.04112306859856911,
+      "grad_norm": 2.07299718987642,
+      "learning_rate": 3.998716155291702e-06,
+      "loss": 0.7425,
+      "num_input_tokens_seen": 7160065,
+      "step": 342
+    },
+    {
+      "epoch": 0.0412433114892082,
+      "grad_norm": 3.1630435537184276,
+      "learning_rate": 3.998688097009366e-06,
+      "loss": 0.917,
+      "num_input_tokens_seen": 7180550,
+      "step": 343
+    },
+    {
+      "epoch": 0.04136355437984729,
+      "grad_norm": 5.364945260908992,
+      "learning_rate": 3.998659735527636e-06,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 7199360,
+      "step": 344
+    },
+    {
+      "epoch": 0.04148379727048638,
+      "grad_norm": 1.96164837181541,
+      "learning_rate": 3.998631070850813e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 7219700,
+      "step": 345
+    },
+    {
+      "epoch": 0.041604040161125476,
+      "grad_norm": 2.402471200441801,
+      "learning_rate": 3.9986021029832455e-06,
+      "loss": 0.8411,
+      "num_input_tokens_seen": 7236735,
+      "step": 346
+    },
+    {
+      "epoch": 0.041724283051764566,
+      "grad_norm": 2.4751875439210544,
+      "learning_rate": 3.9985728319293285e-06,
+      "loss": 0.9197,
+      "num_input_tokens_seen": 7250430,
+      "step": 347
+    },
+    {
+      "epoch": 0.041844525942403656,
+      "grad_norm": 2.1605151134772966,
+      "learning_rate": 3.998543257693501e-06,
+      "loss": 0.8586,
+      "num_input_tokens_seen": 7266905,
+      "step": 348
+    },
+    {
+      "epoch": 0.041964768833042745,
+      "grad_norm": 2.5648519501592015,
+      "learning_rate": 3.998513380280251e-06,
+      "loss": 0.8907,
+      "num_input_tokens_seen": 7286905,
+      "step": 349
+    },
+    {
+      "epoch": 0.042085011723681835,
+      "grad_norm": 4.04527688897182,
+      "learning_rate": 3.99848319969411e-06,
+      "loss": 0.9515,
+      "num_input_tokens_seen": 7304225,
+      "step": 350
+    },
+    {
+      "epoch": 0.042205254614320925,
+      "grad_norm": 2.424947683136834,
+      "learning_rate": 3.9984527159396564e-06,
+      "loss": 0.7904,
+      "num_input_tokens_seen": 7322585,
+      "step": 351
+    },
+    {
+      "epoch": 0.04232549750496002,
+      "grad_norm": 2.100428755097484,
+      "learning_rate": 3.9984219290215154e-06,
+      "loss": 0.8378,
+      "num_input_tokens_seen": 7342480,
+      "step": 352
+    },
+    {
+      "epoch": 0.04244574039559911,
+      "grad_norm": 1.6392051947588568,
+      "learning_rate": 3.998390838944356e-06,
+      "loss": 0.8912,
+      "num_input_tokens_seen": 7363705,
+      "step": 353
+    },
+    {
+      "epoch": 0.0425659832862382,
+      "grad_norm": 2.721800287622192,
+      "learning_rate": 3.998359445712895e-06,
+      "loss": 0.9063,
+      "num_input_tokens_seen": 7382530,
+      "step": 354
+    },
+    {
+      "epoch": 0.04268622617687729,
+      "grad_norm": 2.1972396776567025,
+      "learning_rate": 3.9983277493318955e-06,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 7401545,
+      "step": 355
+    },
+    {
+      "epoch": 0.04280646906751638,
+      "grad_norm": 1.7791387808058512,
+      "learning_rate": 3.998295749806165e-06,
+      "loss": 0.8145,
+      "num_input_tokens_seen": 7422490,
+      "step": 356
+    },
+    {
+      "epoch": 0.04292671195815547,
+      "grad_norm": 2.350104202077775,
+      "learning_rate": 3.998263447140558e-06,
+      "loss": 0.8487,
+      "num_input_tokens_seen": 7442410,
+      "step": 357
+    },
+    {
+      "epoch": 0.04304695484879457,
+      "grad_norm": 2.017620277002353,
+      "learning_rate": 3.998230841339976e-06,
+      "loss": 0.8244,
+      "num_input_tokens_seen": 7464140,
+      "step": 358
+    },
+    {
+      "epoch": 0.04316719773943366,
+      "grad_norm": 2.085608740652392,
+      "learning_rate": 3.998197932409363e-06,
+      "loss": 0.8567,
+      "num_input_tokens_seen": 7481870,
+      "step": 359
+    },
+    {
+      "epoch": 0.04328744063007275,
+      "grad_norm": 3.9792600701715806,
+      "learning_rate": 3.9981647203537125e-06,
+      "loss": 0.8705,
+      "num_input_tokens_seen": 7499090,
+      "step": 360
+    },
+    {
+      "epoch": 0.04340768352071184,
+      "grad_norm": 1.9932683344172748,
+      "learning_rate": 3.998131205178063e-06,
+      "loss": 0.9655,
+      "num_input_tokens_seen": 7517280,
+      "step": 361
+    },
+    {
+      "epoch": 0.04352792641135093,
+      "grad_norm": 2.51843094924229,
+      "learning_rate": 3.998097386887498e-06,
+      "loss": 0.7806,
+      "num_input_tokens_seen": 7534075,
+      "step": 362
+    },
+    {
+      "epoch": 0.04364816930199002,
+      "grad_norm": 1.811306405222341,
+      "learning_rate": 3.998063265487148e-06,
+      "loss": 0.8502,
+      "num_input_tokens_seen": 7554845,
+      "step": 363
+    },
+    {
+      "epoch": 0.043768412192629114,
+      "grad_norm": 1.7882383407992861,
+      "learning_rate": 3.99802884098219e-06,
+      "loss": 0.8242,
+      "num_input_tokens_seen": 7572675,
+      "step": 364
+    },
+    {
+      "epoch": 0.043888655083268203,
+      "grad_norm": 2.579248167976788,
+      "learning_rate": 3.997994113377845e-06,
+      "loss": 0.825,
+      "num_input_tokens_seen": 7591295,
+      "step": 365
+    },
+    {
+      "epoch": 0.04400889797390729,
+      "grad_norm": 2.6396282439807726,
+      "learning_rate": 3.9979590826793815e-06,
+      "loss": 0.8469,
+      "num_input_tokens_seen": 7612205,
+      "step": 366
+    },
+    {
+      "epoch": 0.04412914086454638,
+      "grad_norm": 2.2554866451576343,
+      "learning_rate": 3.997923748892113e-06,
+      "loss": 0.8281,
+      "num_input_tokens_seen": 7631245,
+      "step": 367
+    },
+    {
+      "epoch": 0.04424938375518547,
+      "grad_norm": 1.6408659123853069,
+      "learning_rate": 3.9978881120214015e-06,
+      "loss": 0.8907,
+      "num_input_tokens_seen": 7652485,
+      "step": 368
+    },
+    {
+      "epoch": 0.04436962664582456,
+      "grad_norm": 1.9384176610863997,
+      "learning_rate": 3.997852172072652e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 7673420,
+      "step": 369
+    },
+    {
+      "epoch": 0.04448986953646366,
+      "grad_norm": 2.9730436133467797,
+      "learning_rate": 3.9978159290513155e-06,
+      "loss": 0.9066,
+      "num_input_tokens_seen": 7691220,
+      "step": 370
+    },
+    {
+      "epoch": 0.04461011242710275,
+      "grad_norm": 1.7167953630057466,
+      "learning_rate": 3.997779382962892e-06,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 7713825,
+      "step": 371
+    },
+    {
+      "epoch": 0.04473035531774184,
+      "grad_norm": 2.0204905057262006,
+      "learning_rate": 3.997742533812924e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 7736810,
+      "step": 372
+    },
+    {
+      "epoch": 0.04485059820838093,
+      "grad_norm": 2.76087899022977,
+      "learning_rate": 3.997705381607001e-06,
+      "loss": 0.9243,
+      "num_input_tokens_seen": 7753345,
+      "step": 373
+    },
+    {
+      "epoch": 0.04497084109902002,
+      "grad_norm": 1.1474367204063174,
+      "learning_rate": 3.997667926350761e-06,
+      "loss": 0.6379,
+      "num_input_tokens_seen": 7811395,
+      "step": 374
+    },
+    {
+      "epoch": 0.04509108398965911,
+      "grad_norm": 1.1441503455875692,
+      "learning_rate": 3.997630168049886e-06,
+      "loss": 0.6095,
+      "num_input_tokens_seen": 7869480,
+      "step": 375
+    },
+    {
+      "epoch": 0.045211326880298205,
+      "grad_norm": 3.3751714026683666,
+      "learning_rate": 3.997592106710101e-06,
+      "loss": 0.786,
+      "num_input_tokens_seen": 7888660,
+      "step": 376
+    },
+    {
+      "epoch": 0.045331569770937295,
+      "grad_norm": 3.1353310862799533,
+      "learning_rate": 3.997553742337182e-06,
+      "loss": 0.6727,
+      "num_input_tokens_seen": 7907805,
+      "step": 377
+    },
+    {
+      "epoch": 0.045451812661576385,
+      "grad_norm": 1.8457131451840292,
+      "learning_rate": 3.997515074936949e-06,
+      "loss": 0.9199,
+      "num_input_tokens_seen": 7928400,
+      "step": 378
+    },
+    {
+      "epoch": 0.045572055552215475,
+      "grad_norm": 3.0976728941209335,
+      "learning_rate": 3.997476104515268e-06,
+      "loss": 0.8734,
+      "num_input_tokens_seen": 7946310,
+      "step": 379
+    },
+    {
+      "epoch": 0.045692298442854565,
+      "grad_norm": 2.023514488361353,
+      "learning_rate": 3.9974368310780485e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 7963205,
+      "step": 380
+    },
+    {
+      "epoch": 0.045812541333493655,
+      "grad_norm": 2.539372544643226,
+      "learning_rate": 3.997397254631251e-06,
+      "loss": 0.7518,
+      "num_input_tokens_seen": 7983545,
+      "step": 381
+    },
+    {
+      "epoch": 0.04593278422413275,
+      "grad_norm": 0.9979931868716626,
+      "learning_rate": 3.997357375180878e-06,
+      "loss": 0.6555,
+      "num_input_tokens_seen": 8047545,
+      "step": 382
+    },
+    {
+      "epoch": 0.04605302711477184,
+      "grad_norm": 1.8558539879785372,
+      "learning_rate": 3.997317192732979e-06,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 8066045,
+      "step": 383
+    },
+    {
+      "epoch": 0.04617327000541093,
+      "grad_norm": 2.165453528024373,
+      "learning_rate": 3.99727670729365e-06,
+      "loss": 0.8214,
+      "num_input_tokens_seen": 8084325,
+      "step": 384
+    },
+    {
+      "epoch": 0.04629351289605002,
+      "grad_norm": 1.7732804162030202,
+      "learning_rate": 3.997235918869033e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 8105080,
+      "step": 385
+    },
+    {
+      "epoch": 0.04641375578668911,
+      "grad_norm": 1.9076410478989632,
+      "learning_rate": 3.997194827465315e-06,
+      "loss": 0.8391,
+      "num_input_tokens_seen": 8123395,
+      "step": 386
+    },
+    {
+      "epoch": 0.0465339986773282,
+      "grad_norm": 2.843945838370176,
+      "learning_rate": 3.997153433088728e-06,
+      "loss": 0.929,
+      "num_input_tokens_seen": 8140240,
+      "step": 387
+    },
+    {
+      "epoch": 0.0466542415679673,
+      "grad_norm": 2.370632887481483,
+      "learning_rate": 3.997111735745554e-06,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 8162930,
+      "step": 388
+    },
+    {
+      "epoch": 0.04677448445860639,
+      "grad_norm": 1.9790277251201742,
+      "learning_rate": 3.997069735442118e-06,
+      "loss": 0.8385,
+      "num_input_tokens_seen": 8182345,
+      "step": 389
+    },
+    {
+      "epoch": 0.04689472734924548,
+      "grad_norm": 1.4445478443904698,
+      "learning_rate": 3.997027432184792e-06,
+      "loss": 0.7969,
+      "num_input_tokens_seen": 8206725,
+      "step": 390
+    },
+    {
+      "epoch": 0.04701497023988457,
+      "grad_norm": 1.9265438364287897,
+      "learning_rate": 3.99698482597999e-06,
+      "loss": 0.8819,
+      "num_input_tokens_seen": 8224125,
+      "step": 391
+    },
+    {
+      "epoch": 0.04713521313052366,
+      "grad_norm": 0.9855986454469535,
+      "learning_rate": 3.99694191683418e-06,
+      "loss": 0.6634,
+      "num_input_tokens_seen": 8284645,
+      "step": 392
+    },
+    {
+      "epoch": 0.047255456021162746,
+      "grad_norm": 2.0243212319592274,
+      "learning_rate": 3.996898704753867e-06,
+      "loss": 0.833,
+      "num_input_tokens_seen": 8302315,
+      "step": 393
+    },
+    {
+      "epoch": 0.04737569891180184,
+      "grad_norm": 2.4106680007724735,
+      "learning_rate": 3.996855189745609e-06,
+      "loss": 0.8927,
+      "num_input_tokens_seen": 8321300,
+      "step": 394
+    },
+    {
+      "epoch": 0.04749594180244093,
+      "grad_norm": 2.0119202428119425,
+      "learning_rate": 3.996811371816007e-06,
+      "loss": 0.9257,
+      "num_input_tokens_seen": 8343445,
+      "step": 395
+    },
+    {
+      "epoch": 0.04761618469308002,
+      "grad_norm": 1.8566418594632506,
+      "learning_rate": 3.996767250971707e-06,
+      "loss": 0.7908,
+      "num_input_tokens_seen": 8365905,
+      "step": 396
+    },
+    {
+      "epoch": 0.04773642758371911,
+      "grad_norm": 2.000281657982543,
+      "learning_rate": 3.996722827219403e-06,
+      "loss": 0.8725,
+      "num_input_tokens_seen": 8387240,
+      "step": 397
+    },
+    {
+      "epoch": 0.0478566704743582,
+      "grad_norm": 2.545853532271167,
+      "learning_rate": 3.996678100565833e-06,
+      "loss": 0.8245,
+      "num_input_tokens_seen": 8406015,
+      "step": 398
+    },
+    {
+      "epoch": 0.04797691336499729,
+      "grad_norm": 2.762848078763912,
+      "learning_rate": 3.996633071017783e-06,
+      "loss": 0.8767,
+      "num_input_tokens_seen": 8422365,
+      "step": 399
+    },
+    {
+      "epoch": 0.04809715625563638,
+      "grad_norm": 2.342698481896991,
+      "learning_rate": 3.996587738582084e-06,
+      "loss": 0.8178,
+      "num_input_tokens_seen": 8438885,
+      "step": 400
+    },
+    {
+      "epoch": 0.04821739914627548,
+      "grad_norm": 2.2034345914585676,
+      "learning_rate": 3.9965421032656115e-06,
+      "loss": 0.8628,
+      "num_input_tokens_seen": 8458535,
+      "step": 401
+    },
+    {
+      "epoch": 0.04833764203691457,
+      "grad_norm": 2.715378292705908,
+      "learning_rate": 3.99649616507529e-06,
+      "loss": 0.9508,
+      "num_input_tokens_seen": 8477350,
+      "step": 402
+    },
+    {
+      "epoch": 0.04845788492755366,
+      "grad_norm": 1.0300050063232031,
+      "learning_rate": 3.996449924018088e-06,
+      "loss": 0.6779,
+      "num_input_tokens_seen": 8537530,
+      "step": 403
+    },
+    {
+      "epoch": 0.04857812781819275,
+      "grad_norm": 1.9694461933945187,
+      "learning_rate": 3.99640338010102e-06,
+      "loss": 0.8075,
+      "num_input_tokens_seen": 8556355,
+      "step": 404
+    },
+    {
+      "epoch": 0.04869837070883184,
+      "grad_norm": 2.128598282821002,
+      "learning_rate": 3.996356533331146e-06,
+      "loss": 0.792,
+      "num_input_tokens_seen": 8577945,
+      "step": 405
+    },
+    {
+      "epoch": 0.04881861359947093,
+      "grad_norm": 3.753623292226186,
+      "learning_rate": 3.996309383715573e-06,
+      "loss": 0.6196,
+      "num_input_tokens_seen": 8596445,
+      "step": 406
+    },
+    {
+      "epoch": 0.048938856490110025,
+      "grad_norm": 2.1457829206397796,
+      "learning_rate": 3.996261931261454e-06,
+      "loss": 0.7354,
+      "num_input_tokens_seen": 8614745,
+      "step": 407
+    },
+    {
+      "epoch": 0.049059099380749115,
+      "grad_norm": 1.7048124938654663,
+      "learning_rate": 3.996214175975987e-06,
+      "loss": 0.8649,
+      "num_input_tokens_seen": 8634985,
+      "step": 408
+    },
+    {
+      "epoch": 0.049179342271388204,
+      "grad_norm": 2.21201774714938,
+      "learning_rate": 3.996166117866417e-06,
+      "loss": 0.7991,
+      "num_input_tokens_seen": 8656640,
+      "step": 409
+    },
+    {
+      "epoch": 0.049299585162027294,
+      "grad_norm": 3.4935376058018184,
+      "learning_rate": 3.996117756940035e-06,
+      "loss": 0.8726,
+      "num_input_tokens_seen": 8673045,
+      "step": 410
+    },
+    {
+      "epoch": 0.049419828052666384,
+      "grad_norm": 2.106591600290592,
+      "learning_rate": 3.996069093204175e-06,
+      "loss": 0.9743,
+      "num_input_tokens_seen": 8688725,
+      "step": 411
+    },
+    {
+      "epoch": 0.049540070943305474,
+      "grad_norm": 2.481923273742918,
+      "learning_rate": 3.996020126666221e-06,
+      "loss": 0.891,
+      "num_input_tokens_seen": 8705425,
+      "step": 412
+    },
+    {
+      "epoch": 0.04966031383394457,
+      "grad_norm": 2.916968640282379,
+      "learning_rate": 3.995970857333601e-06,
+      "loss": 0.829,
+      "num_input_tokens_seen": 8725555,
+      "step": 413
+    },
+    {
+      "epoch": 0.04978055672458366,
+      "grad_norm": 2.038539611976774,
+      "learning_rate": 3.995921285213789e-06,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 8745535,
+      "step": 414
+    },
+    {
+      "epoch": 0.04990079961522275,
+      "grad_norm": 2.3906471122837547,
+      "learning_rate": 3.995871410314305e-06,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 8763815,
+      "step": 415
+    },
+    {
+      "epoch": 0.05002104250586184,
+      "grad_norm": 1.0956416157862419,
+      "learning_rate": 3.995821232642714e-06,
+      "loss": 0.6606,
+      "num_input_tokens_seen": 8821940,
+      "step": 416
+    },
+    {
+      "epoch": 0.05014128539650093,
+      "grad_norm": 2.7368031799510644,
+      "learning_rate": 3.995770752206629e-06,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 8842735,
+      "step": 417
+    },
+    {
+      "epoch": 0.05026152828714002,
+      "grad_norm": 2.095156556256337,
+      "learning_rate": 3.995719969013709e-06,
+      "loss": 0.9669,
+      "num_input_tokens_seen": 8859635,
+      "step": 418
+    },
+    {
+      "epoch": 0.05038177117777912,
+      "grad_norm": 2.8008922263705243,
+      "learning_rate": 3.995668883071655e-06,
+      "loss": 0.8633,
+      "num_input_tokens_seen": 8875580,
+      "step": 419
+    },
+    {
+      "epoch": 0.050502014068418206,
+      "grad_norm": 2.4970550100186437,
+      "learning_rate": 3.995617494388219e-06,
+      "loss": 0.9164,
+      "num_input_tokens_seen": 8893420,
+      "step": 420
+    },
+    {
+      "epoch": 0.050622256959057296,
+      "grad_norm": 1.9457230550900073,
+      "learning_rate": 3.995565802971196e-06,
+      "loss": 0.8125,
+      "num_input_tokens_seen": 8913115,
+      "step": 421
+    },
+    {
+      "epoch": 0.050742499849696386,
+      "grad_norm": 2.0390095219509674,
+      "learning_rate": 3.995513808828427e-06,
+      "loss": 0.6809,
+      "num_input_tokens_seen": 8935630,
+      "step": 422
+    },
+    {
+      "epoch": 0.050862742740335476,
+      "grad_norm": 1.9830020846860377,
+      "learning_rate": 3.9954615119678e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 8953905,
+      "step": 423
+    },
+    {
+      "epoch": 0.050982985630974566,
+      "grad_norm": 1.9335259142108472,
+      "learning_rate": 3.995408912397248e-06,
+      "loss": 0.796,
+      "num_input_tokens_seen": 8971520,
+      "step": 424
+    },
+    {
+      "epoch": 0.05110322852161366,
+      "grad_norm": 2.46969107853472,
+      "learning_rate": 3.99535601012475e-06,
+      "loss": 0.927,
+      "num_input_tokens_seen": 8986570,
+      "step": 425
+    },
+    {
+      "epoch": 0.05122347141225275,
+      "grad_norm": 1.8121846701150468,
+      "learning_rate": 3.995302805158333e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 9008945,
+      "step": 426
+    },
+    {
+      "epoch": 0.05134371430289184,
+      "grad_norm": 2.1195635279438654,
+      "learning_rate": 3.9952492975060665e-06,
+      "loss": 0.8385,
+      "num_input_tokens_seen": 9028735,
+      "step": 427
+    },
+    {
+      "epoch": 0.05146395719353093,
+      "grad_norm": 2.498068536352519,
+      "learning_rate": 3.995195487176067e-06,
+      "loss": 0.8649,
+      "num_input_tokens_seen": 9048685,
+      "step": 428
+    },
+    {
+      "epoch": 0.05158420008417002,
+      "grad_norm": 1.9557368729466096,
+      "learning_rate": 3.995141374176499e-06,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 9066800,
+      "step": 429
+    },
+    {
+      "epoch": 0.05170444297480911,
+      "grad_norm": 1.0325867068624763,
+      "learning_rate": 3.995086958515572e-06,
+      "loss": 0.6815,
+      "num_input_tokens_seen": 9124540,
+      "step": 430
+    },
+    {
+      "epoch": 0.05182468586544821,
+      "grad_norm": 1.0132728741287735,
+      "learning_rate": 3.995032240201538e-06,
+      "loss": 0.648,
+      "num_input_tokens_seen": 9186655,
+      "step": 431
+    },
+    {
+      "epoch": 0.0519449287560873,
+      "grad_norm": 1.1170326233416954,
+      "learning_rate": 3.9949772192427e-06,
+      "loss": 0.6585,
+      "num_input_tokens_seen": 9233000,
+      "step": 432
+    },
+    {
+      "epoch": 0.05206517164672639,
+      "grad_norm": 2.0169408873886088,
+      "learning_rate": 3.994921895647405e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 9250890,
+      "step": 433
+    },
+    {
+      "epoch": 0.05218541453736548,
+      "grad_norm": 0.8909107788343686,
+      "learning_rate": 3.994866269424043e-06,
+      "loss": 0.5729,
+      "num_input_tokens_seen": 9306980,
+      "step": 434
+    },
+    {
+      "epoch": 0.05230565742800457,
+      "grad_norm": 2.2576947765939206,
+      "learning_rate": 3.9948103405810545e-06,
+      "loss": 0.783,
+      "num_input_tokens_seen": 9325650,
+      "step": 435
+    },
+    {
+      "epoch": 0.05242590031864366,
+      "grad_norm": 2.022624134660986,
+      "learning_rate": 3.994754109126923e-06,
+      "loss": 0.8571,
+      "num_input_tokens_seen": 9346865,
+      "step": 436
+    },
+    {
+      "epoch": 0.052546143209282754,
+      "grad_norm": 1.742196396548821,
+      "learning_rate": 3.994697575070181e-06,
+      "loss": 0.9278,
+      "num_input_tokens_seen": 9366045,
+      "step": 437
+    },
+    {
+      "epoch": 0.052666386099921844,
+      "grad_norm": 2.9849960890835567,
+      "learning_rate": 3.994640738419402e-06,
+      "loss": 0.909,
+      "num_input_tokens_seen": 9385140,
+      "step": 438
+    },
+    {
+      "epoch": 0.052786628990560934,
+      "grad_norm": 1.9772482984084772,
+      "learning_rate": 3.9945835991832075e-06,
+      "loss": 0.8034,
+      "num_input_tokens_seen": 9406745,
+      "step": 439
+    },
+    {
+      "epoch": 0.052906871881200024,
+      "grad_norm": 2.5061167807327234,
+      "learning_rate": 3.994526157370268e-06,
+      "loss": 0.9321,
+      "num_input_tokens_seen": 9425080,
+      "step": 440
+    },
+    {
+      "epoch": 0.053027114771839114,
+      "grad_norm": 0.9495371009024924,
+      "learning_rate": 3.994468412989296e-06,
+      "loss": 0.6172,
+      "num_input_tokens_seen": 9486210,
+      "step": 441
+    },
+    {
+      "epoch": 0.053147357662478203,
+      "grad_norm": 2.0882140774216134,
+      "learning_rate": 3.994410366049052e-06,
+      "loss": 0.9484,
+      "num_input_tokens_seen": 9503790,
+      "step": 442
+    },
+    {
+      "epoch": 0.0532676005531173,
+      "grad_norm": 2.3515828496158893,
+      "learning_rate": 3.994352016558341e-06,
+      "loss": 0.8394,
+      "num_input_tokens_seen": 9520815,
+      "step": 443
+    },
+    {
+      "epoch": 0.05338784344375639,
+      "grad_norm": 2.1256973121918388,
+      "learning_rate": 3.994293364526014e-06,
+      "loss": 0.7441,
+      "num_input_tokens_seen": 9541420,
+      "step": 444
+    },
+    {
+      "epoch": 0.05350808633439548,
+      "grad_norm": 2.153764930048083,
+      "learning_rate": 3.99423440996097e-06,
+      "loss": 0.8435,
+      "num_input_tokens_seen": 9560680,
+      "step": 445
+    },
+    {
+      "epoch": 0.05362832922503457,
+      "grad_norm": 4.727095064478514,
+      "learning_rate": 3.994175152872152e-06,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 9579485,
+      "step": 446
+    },
+    {
+      "epoch": 0.05374857211567366,
+      "grad_norm": 2.2550071022204383,
+      "learning_rate": 3.994115593268548e-06,
+      "loss": 0.7971,
+      "num_input_tokens_seen": 9598985,
+      "step": 447
+    },
+    {
+      "epoch": 0.05386881500631275,
+      "grad_norm": 2.0551462595466172,
+      "learning_rate": 3.994055731159195e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 9616175,
+      "step": 448
+    },
+    {
+      "epoch": 0.053989057896951846,
+      "grad_norm": 1.884791025630162,
+      "learning_rate": 3.993995566553172e-06,
+      "loss": 0.8782,
+      "num_input_tokens_seen": 9634860,
+      "step": 449
+    },
+    {
+      "epoch": 0.054109300787590936,
+      "grad_norm": 1.715468162095248,
+      "learning_rate": 3.993935099459607e-06,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 9656195,
+      "step": 450
+    },
+    {
+      "epoch": 0.054229543678230026,
+      "grad_norm": 2.0018991408391567,
+      "learning_rate": 3.993874329887673e-06,
+      "loss": 0.75,
+      "num_input_tokens_seen": 9674570,
+      "step": 451
+    },
+    {
+      "epoch": 0.054349786568869116,
+      "grad_norm": 2.5245653679969604,
+      "learning_rate": 3.993813257846589e-06,
+      "loss": 0.8651,
+      "num_input_tokens_seen": 9691045,
+      "step": 452
+    },
+    {
+      "epoch": 0.054470029459508205,
+      "grad_norm": 5.073426511016169,
+      "learning_rate": 3.993751883345619e-06,
+      "loss": 0.9362,
+      "num_input_tokens_seen": 9709125,
+      "step": 453
+    },
+    {
+      "epoch": 0.054590272350147295,
+      "grad_norm": 2.738593803944557,
+      "learning_rate": 3.993690206394073e-06,
+      "loss": 0.8783,
+      "num_input_tokens_seen": 9725145,
+      "step": 454
+    },
+    {
+      "epoch": 0.054710515240786385,
+      "grad_norm": 2.2131290663029,
+      "learning_rate": 3.993628227001307e-06,
+      "loss": 0.8844,
+      "num_input_tokens_seen": 9743065,
+      "step": 455
+    },
+    {
+      "epoch": 0.05483075813142548,
+      "grad_norm": 3.708675805457093,
+      "learning_rate": 3.993565945176726e-06,
+      "loss": 0.7133,
+      "num_input_tokens_seen": 9763810,
+      "step": 456
+    },
+    {
+      "epoch": 0.05495100102206457,
+      "grad_norm": 2.5102640724122764,
+      "learning_rate": 3.993503360929776e-06,
+      "loss": 0.8478,
+      "num_input_tokens_seen": 9782415,
+      "step": 457
+    },
+    {
+      "epoch": 0.05507124391270366,
+      "grad_norm": 3.2399637815336764,
+      "learning_rate": 3.99344047426995e-06,
+      "loss": 0.8145,
+      "num_input_tokens_seen": 9803395,
+      "step": 458
+    },
+    {
+      "epoch": 0.05519148680334275,
+      "grad_norm": 2.2112175900612194,
+      "learning_rate": 3.993377285206789e-06,
+      "loss": 0.9372,
+      "num_input_tokens_seen": 9822900,
+      "step": 459
+    },
+    {
+      "epoch": 0.05531172969398184,
+      "grad_norm": 1.6419042211246246,
+      "learning_rate": 3.99331379374988e-06,
+      "loss": 0.8669,
+      "num_input_tokens_seen": 9846225,
+      "step": 460
+    },
+    {
+      "epoch": 0.05543197258462093,
+      "grad_norm": 2.262798783247126,
+      "learning_rate": 3.993249999908852e-06,
+      "loss": 0.8075,
+      "num_input_tokens_seen": 9866095,
+      "step": 461
+    },
+    {
+      "epoch": 0.05555221547526003,
+      "grad_norm": 1.9975957376622713,
+      "learning_rate": 3.993185903693384e-06,
+      "loss": 0.8714,
+      "num_input_tokens_seen": 9882615,
+      "step": 462
+    },
+    {
+      "epoch": 0.05567245836589912,
+      "grad_norm": 2.1838738078948317,
+      "learning_rate": 3.9931215051131995e-06,
+      "loss": 0.8334,
+      "num_input_tokens_seen": 9902980,
+      "step": 463
+    },
+    {
+      "epoch": 0.05579270125653821,
+      "grad_norm": 1.9815331584439932,
+      "learning_rate": 3.993056804178068e-06,
+      "loss": 0.8071,
+      "num_input_tokens_seen": 9924245,
+      "step": 464
+    },
+    {
+      "epoch": 0.0559129441471773,
+      "grad_norm": 2.0167152865221043,
+      "learning_rate": 3.992991800897803e-06,
+      "loss": 0.8468,
+      "num_input_tokens_seen": 9943770,
+      "step": 465
+    },
+    {
+      "epoch": 0.05603318703781639,
+      "grad_norm": 2.577183816569157,
+      "learning_rate": 3.9929264952822665e-06,
+      "loss": 0.8954,
+      "num_input_tokens_seen": 9961025,
+      "step": 466
+    },
+    {
+      "epoch": 0.05615342992845548,
+      "grad_norm": 2.3021964040444125,
+      "learning_rate": 3.992860887341366e-06,
+      "loss": 0.8805,
+      "num_input_tokens_seen": 9978915,
+      "step": 467
+    },
+    {
+      "epoch": 0.056273672819094574,
+      "grad_norm": 1.9914289743116045,
+      "learning_rate": 3.992794977085052e-06,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 9996635,
+      "step": 468
+    },
+    {
+      "epoch": 0.056393915709733664,
+      "grad_norm": 2.3997097017497557,
+      "learning_rate": 3.992728764523326e-06,
+      "loss": 0.8459,
+      "num_input_tokens_seen": 10015300,
+      "step": 469
+    },
+    {
+      "epoch": 0.05651415860037275,
+      "grad_norm": 1.8025508944863202,
+      "learning_rate": 3.99266224966623e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 10035935,
+      "step": 470
+    },
+    {
+      "epoch": 0.05663440149101184,
+      "grad_norm": 1.8615208777759509,
+      "learning_rate": 3.992595432523855e-06,
+      "loss": 0.8732,
+      "num_input_tokens_seen": 10052945,
+      "step": 471
+    },
+    {
+      "epoch": 0.05675464438165093,
+      "grad_norm": 2.290649144336832,
+      "learning_rate": 3.992528313106338e-06,
+      "loss": 0.8604,
+      "num_input_tokens_seen": 10070865,
+      "step": 472
+    },
+    {
+      "epoch": 0.05687488727229002,
+      "grad_norm": 2.2578693713553504,
+      "learning_rate": 3.9924608914238595e-06,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 10085580,
+      "step": 473
+    },
+    {
+      "epoch": 0.05699513016292912,
+      "grad_norm": 5.416889397401193,
+      "learning_rate": 3.992393167486648e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 10104450,
+      "step": 474
+    },
+    {
+      "epoch": 0.05711537305356821,
+      "grad_norm": 2.1768818390965174,
+      "learning_rate": 3.992325141304977e-06,
+      "loss": 0.8043,
+      "num_input_tokens_seen": 10122125,
+      "step": 475
+    },
+    {
+      "epoch": 0.0572356159442073,
+      "grad_norm": 3.5235808281468404,
+      "learning_rate": 3.992256812889166e-06,
+      "loss": 0.8623,
+      "num_input_tokens_seen": 10137950,
+      "step": 476
+    },
+    {
+      "epoch": 0.05735585883484639,
+      "grad_norm": 4.680692353486723,
+      "learning_rate": 3.992188182249582e-06,
+      "loss": 0.7706,
+      "num_input_tokens_seen": 10159565,
+      "step": 477
+    },
+    {
+      "epoch": 0.05747610172548548,
+      "grad_norm": 2.0088306482333,
+      "learning_rate": 3.992119249396633e-06,
+      "loss": 0.9114,
+      "num_input_tokens_seen": 10177970,
+      "step": 478
+    },
+    {
+      "epoch": 0.05759634461612457,
+      "grad_norm": 1.891449319167649,
+      "learning_rate": 3.992050014340778e-06,
+      "loss": 0.821,
+      "num_input_tokens_seen": 10198045,
+      "step": 479
+    },
+    {
+      "epoch": 0.057716587506763666,
+      "grad_norm": 1.8591412030440282,
+      "learning_rate": 3.99198047709252e-06,
+      "loss": 0.5827,
+      "num_input_tokens_seen": 10259285,
+      "step": 480
+    },
+    {
+      "epoch": 0.057836830397402755,
+      "grad_norm": 1.8712158366408569,
+      "learning_rate": 3.991910637662408e-06,
+      "loss": 0.7941,
+      "num_input_tokens_seen": 10279295,
+      "step": 481
+    },
+    {
+      "epoch": 0.057957073288041845,
+      "grad_norm": 1.867321617714949,
+      "learning_rate": 3.9918404960610355e-06,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 10298045,
+      "step": 482
+    },
+    {
+      "epoch": 0.058077316178680935,
+      "grad_norm": 2.5490296345981354,
+      "learning_rate": 3.991770052299043e-06,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 10315995,
+      "step": 483
+    },
+    {
+      "epoch": 0.058197559069320025,
+      "grad_norm": 2.654946995194073,
+      "learning_rate": 3.991699306387118e-06,
+      "loss": 0.8825,
+      "num_input_tokens_seen": 10334185,
+      "step": 484
+    },
+    {
+      "epoch": 0.058317801959959115,
+      "grad_norm": 1.8533565075365972,
+      "learning_rate": 3.991628258335991e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 10356110,
+      "step": 485
+    },
+    {
+      "epoch": 0.05843804485059821,
+      "grad_norm": 3.7283403379959834,
+      "learning_rate": 3.991556908156442e-06,
+      "loss": 0.879,
+      "num_input_tokens_seen": 10372355,
+      "step": 486
+    },
+    {
+      "epoch": 0.0585582877412373,
+      "grad_norm": 1.8900953341548485,
+      "learning_rate": 3.9914852558592914e-06,
+      "loss": 0.8653,
+      "num_input_tokens_seen": 10393125,
+      "step": 487
+    },
+    {
+      "epoch": 0.05867853063187639,
+      "grad_norm": 7.081801536948411,
+      "learning_rate": 3.991413301455413e-06,
+      "loss": 0.8098,
+      "num_input_tokens_seen": 10409295,
+      "step": 488
+    },
+    {
+      "epoch": 0.05879877352251548,
+      "grad_norm": 2.261731731221124,
+      "learning_rate": 3.991341044955719e-06,
+      "loss": 0.7744,
+      "num_input_tokens_seen": 10428770,
+      "step": 489
+    },
+    {
+      "epoch": 0.05891901641315457,
+      "grad_norm": 2.1536266199498333,
+      "learning_rate": 3.991268486371172e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 10447045,
+      "step": 490
+    },
+    {
+      "epoch": 0.05903925930379366,
+      "grad_norm": 2.4097094101941345,
+      "learning_rate": 3.991195625712779e-06,
+      "loss": 0.8746,
+      "num_input_tokens_seen": 10463730,
+      "step": 491
+    },
+    {
+      "epoch": 0.05915950219443276,
+      "grad_norm": 2.0052990238002812,
+      "learning_rate": 3.991122462991592e-06,
+      "loss": 0.8178,
+      "num_input_tokens_seen": 10482970,
+      "step": 492
+    },
+    {
+      "epoch": 0.05927974508507185,
+      "grad_norm": 3.124127433901469,
+      "learning_rate": 3.991048998218712e-06,
+      "loss": 0.8192,
+      "num_input_tokens_seen": 10495995,
+      "step": 493
+    },
+    {
+      "epoch": 0.05939998797571094,
+      "grad_norm": 2.8505665571866277,
+      "learning_rate": 3.990975231405281e-06,
+      "loss": 0.7643,
+      "num_input_tokens_seen": 10514165,
+      "step": 494
+    },
+    {
+      "epoch": 0.05952023086635003,
+      "grad_norm": 2.189328628985515,
+      "learning_rate": 3.990901162562491e-06,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 10534575,
+      "step": 495
+    },
+    {
+      "epoch": 0.05964047375698912,
+      "grad_norm": 1.9509847279846109,
+      "learning_rate": 3.9908267917015765e-06,
+      "loss": 0.9065,
+      "num_input_tokens_seen": 10552355,
+      "step": 496
+    },
+    {
+      "epoch": 0.059760716647628206,
+      "grad_norm": 1.9308492860027955,
+      "learning_rate": 3.990752118833821e-06,
+      "loss": 0.9164,
+      "num_input_tokens_seen": 10569515,
+      "step": 497
+    },
+    {
+      "epoch": 0.0598809595382673,
+      "grad_norm": 1.8686151732403626,
+      "learning_rate": 3.990677143970553e-06,
+      "loss": 0.774,
+      "num_input_tokens_seen": 10590045,
+      "step": 498
+    },
+    {
+      "epoch": 0.06000120242890639,
+      "grad_norm": 3.2748339879636847,
+      "learning_rate": 3.990601867123144e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 10609490,
+      "step": 499
+    },
+    {
+      "epoch": 0.06012144531954548,
+      "grad_norm": 2.4337536752088513,
+      "learning_rate": 3.990526288303014e-06,
+      "loss": 0.8524,
+      "num_input_tokens_seen": 10628000,
+      "step": 500
+    },
+    {
+      "epoch": 0.06024168821018457,
+      "grad_norm": 1.739362748024375,
+      "learning_rate": 3.9904504075216295e-06,
+      "loss": 0.9034,
+      "num_input_tokens_seen": 10648480,
+      "step": 501
+    },
+    {
+      "epoch": 0.06036193110082366,
+      "grad_norm": 2.598974519715081,
+      "learning_rate": 3.990374224790501e-06,
+      "loss": 0.9345,
+      "num_input_tokens_seen": 10666405,
+      "step": 502
+    },
+    {
+      "epoch": 0.06048217399146275,
+      "grad_norm": 1.9783589926431029,
+      "learning_rate": 3.990297740121185e-06,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 10684060,
+      "step": 503
+    },
+    {
+      "epoch": 0.06060241688210185,
+      "grad_norm": 1.8694123166580234,
+      "learning_rate": 3.990220953525284e-06,
+      "loss": 0.7721,
+      "num_input_tokens_seen": 10700890,
+      "step": 504
+    },
+    {
+      "epoch": 0.06072265977274094,
+      "grad_norm": 4.149100831355966,
+      "learning_rate": 3.9901438650144465e-06,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 10716860,
+      "step": 505
+    },
+    {
+      "epoch": 0.06084290266338003,
+      "grad_norm": 2.9211339624220134,
+      "learning_rate": 3.990066474600367e-06,
+      "loss": 0.9181,
+      "num_input_tokens_seen": 10734550,
+      "step": 506
+    },
+    {
+      "epoch": 0.06096314555401912,
+      "grad_norm": 2.3361459304190144,
+      "learning_rate": 3.989988782294786e-06,
+      "loss": 0.6782,
+      "num_input_tokens_seen": 10754360,
+      "step": 507
+    },
+    {
+      "epoch": 0.06108338844465821,
+      "grad_norm": 1.9149143850380357,
+      "learning_rate": 3.989910788109489e-06,
+      "loss": 0.9458,
+      "num_input_tokens_seen": 10770730,
+      "step": 508
+    },
+    {
+      "epoch": 0.0612036313352973,
+      "grad_norm": 2.0719309128283157,
+      "learning_rate": 3.989832492056307e-06,
+      "loss": 0.7555,
+      "num_input_tokens_seen": 10791475,
+      "step": 509
+    },
+    {
+      "epoch": 0.06132387422593639,
+      "grad_norm": 3.213781318485671,
+      "learning_rate": 3.989753894147119e-06,
+      "loss": 0.808,
+      "num_input_tokens_seen": 10811320,
+      "step": 510
+    },
+    {
+      "epoch": 0.061444117116575485,
+      "grad_norm": 1.8575915014479352,
+      "learning_rate": 3.989674994393846e-06,
+      "loss": 0.7979,
+      "num_input_tokens_seen": 10830515,
+      "step": 511
+    },
+    {
+      "epoch": 0.061564360007214575,
+      "grad_norm": 2.2259339012668633,
+      "learning_rate": 3.98959579280846e-06,
+      "loss": 0.944,
+      "num_input_tokens_seen": 10848635,
+      "step": 512
+    },
+    {
+      "epoch": 0.061684602897853665,
+      "grad_norm": 2.9625590471088588,
+      "learning_rate": 3.989516289402973e-06,
+      "loss": 0.8301,
+      "num_input_tokens_seen": 10863985,
+      "step": 513
+    },
+    {
+      "epoch": 0.061804845788492754,
+      "grad_norm": 2.6729341212316524,
+      "learning_rate": 3.989436484189447e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 10881650,
+      "step": 514
+    },
+    {
+      "epoch": 0.061925088679131844,
+      "grad_norm": 2.7756938883123508,
+      "learning_rate": 3.9893563771799885e-06,
+      "loss": 0.8167,
+      "num_input_tokens_seen": 10897845,
+      "step": 515
+    },
+    {
+      "epoch": 0.062045331569770934,
+      "grad_norm": 2.2899433764014985,
+      "learning_rate": 3.989275968386749e-06,
+      "loss": 0.8699,
+      "num_input_tokens_seen": 10915475,
+      "step": 516
+    },
+    {
+      "epoch": 0.06216557446041003,
+      "grad_norm": 2.0005072297173645,
+      "learning_rate": 3.989195257821926e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 10933680,
+      "step": 517
+    },
+    {
+      "epoch": 0.06228581735104912,
+      "grad_norm": 2.2103394415461524,
+      "learning_rate": 3.989114245497765e-06,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 10953200,
+      "step": 518
+    },
+    {
+      "epoch": 0.06240606024168821,
+      "grad_norm": 2.4226692730025374,
+      "learning_rate": 3.989032931426554e-06,
+      "loss": 0.9517,
+      "num_input_tokens_seen": 10970075,
+      "step": 519
+    },
+    {
+      "epoch": 0.06252630313232731,
+      "grad_norm": 1.9818754387454325,
+      "learning_rate": 3.9889513156206295e-06,
+      "loss": 0.8715,
+      "num_input_tokens_seen": 10989235,
+      "step": 520
+    },
+    {
+      "epoch": 0.06264654602296639,
+      "grad_norm": 3.5067303838808677,
+      "learning_rate": 3.988869398092371e-06,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 11008865,
+      "step": 521
+    },
+    {
+      "epoch": 0.06276678891360549,
+      "grad_norm": 2.20054691021171,
+      "learning_rate": 3.988787178854206e-06,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 11028120,
+      "step": 522
+    },
+    {
+      "epoch": 0.06288703180424457,
+      "grad_norm": 2.125064072691492,
+      "learning_rate": 3.988704657918608e-06,
+      "loss": 0.8785,
+      "num_input_tokens_seen": 11046900,
+      "step": 523
+    },
+    {
+      "epoch": 0.06300727469488367,
+      "grad_norm": 2.937737490325927,
+      "learning_rate": 3.988621835298094e-06,
+      "loss": 0.8109,
+      "num_input_tokens_seen": 11063835,
+      "step": 524
+    },
+    {
+      "epoch": 0.06312751758552275,
+      "grad_norm": 1.8389244440830004,
+      "learning_rate": 3.988538711005229e-06,
+      "loss": 0.9203,
+      "num_input_tokens_seen": 11083010,
+      "step": 525
+    },
+    {
+      "epoch": 0.06324776047616185,
+      "grad_norm": 2.4991849503375385,
+      "learning_rate": 3.988455285052622e-06,
+      "loss": 0.8956,
+      "num_input_tokens_seen": 11098910,
+      "step": 526
+    },
+    {
+      "epoch": 0.06336800336680094,
+      "grad_norm": 2.089901340841375,
+      "learning_rate": 3.98837155745293e-06,
+      "loss": 0.8299,
+      "num_input_tokens_seen": 11116670,
+      "step": 527
+    },
+    {
+      "epoch": 0.06348824625744003,
+      "grad_norm": 2.024543591355315,
+      "learning_rate": 3.988287528218854e-06,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 11135175,
+      "step": 528
+    },
+    {
+      "epoch": 0.06360848914807912,
+      "grad_norm": 19.767599156326686,
+      "learning_rate": 3.98820319736314e-06,
+      "loss": 0.8956,
+      "num_input_tokens_seen": 11151510,
+      "step": 529
+    },
+    {
+      "epoch": 0.0637287320387182,
+      "grad_norm": 2.1993387451356927,
+      "learning_rate": 3.988118564898582e-06,
+      "loss": 0.8552,
+      "num_input_tokens_seen": 11170770,
+      "step": 530
+    },
+    {
+      "epoch": 0.0638489749293573,
+      "grad_norm": 2.637295695710938,
+      "learning_rate": 3.988033630838019e-06,
+      "loss": 0.893,
+      "num_input_tokens_seen": 11184530,
+      "step": 531
+    },
+    {
+      "epoch": 0.0639692178199964,
+      "grad_norm": 2.4154880255597475,
+      "learning_rate": 3.987948395194334e-06,
+      "loss": 0.8716,
+      "num_input_tokens_seen": 11206630,
+      "step": 532
+    },
+    {
+      "epoch": 0.06408946071063548,
+      "grad_norm": 3.1376336621433674,
+      "learning_rate": 3.987862857980458e-06,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 11222295,
+      "step": 533
+    },
+    {
+      "epoch": 0.06420970360127458,
+      "grad_norm": 1.962397799320079,
+      "learning_rate": 3.987777019209368e-06,
+      "loss": 0.7658,
+      "num_input_tokens_seen": 11242530,
+      "step": 534
+    },
+    {
+      "epoch": 0.06432994649191366,
+      "grad_norm": 1.7555089945683016,
+      "learning_rate": 3.987690878894084e-06,
+      "loss": 0.8174,
+      "num_input_tokens_seen": 11261965,
+      "step": 535
+    },
+    {
+      "epoch": 0.06445018938255276,
+      "grad_norm": 2.3033790486068138,
+      "learning_rate": 3.987604437047673e-06,
+      "loss": 0.8471,
+      "num_input_tokens_seen": 11281485,
+      "step": 536
+    },
+    {
+      "epoch": 0.06457043227319184,
+      "grad_norm": 2.152247193422331,
+      "learning_rate": 3.987517693683251e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 11299780,
+      "step": 537
+    },
+    {
+      "epoch": 0.06469067516383094,
+      "grad_norm": 2.3594580487430146,
+      "learning_rate": 3.9874306488139745e-06,
+      "loss": 0.9531,
+      "num_input_tokens_seen": 11314760,
+      "step": 538
+    },
+    {
+      "epoch": 0.06481091805447003,
+      "grad_norm": 1.8652951661676307,
+      "learning_rate": 3.987343302453049e-06,
+      "loss": 0.8687,
+      "num_input_tokens_seen": 11335755,
+      "step": 539
+    },
+    {
+      "epoch": 0.06493116094510912,
+      "grad_norm": 1.974887311788785,
+      "learning_rate": 3.987255654613724e-06,
+      "loss": 0.8265,
+      "num_input_tokens_seen": 11359240,
+      "step": 540
+    },
+    {
+      "epoch": 0.06505140383574821,
+      "grad_norm": 2.578322020091757,
+      "learning_rate": 3.987167705309296e-06,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 11378235,
+      "step": 541
+    },
+    {
+      "epoch": 0.0651716467263873,
+      "grad_norm": 2.171382508485462,
+      "learning_rate": 3.987079454553108e-06,
+      "loss": 0.9487,
+      "num_input_tokens_seen": 11395905,
+      "step": 542
+    },
+    {
+      "epoch": 0.0652918896170264,
+      "grad_norm": 2.031591960511134,
+      "learning_rate": 3.986990902358546e-06,
+      "loss": 0.914,
+      "num_input_tokens_seen": 11412565,
+      "step": 543
+    },
+    {
+      "epoch": 0.06541213250766549,
+      "grad_norm": 2.364654999890037,
+      "learning_rate": 3.986902048739045e-06,
+      "loss": 0.9279,
+      "num_input_tokens_seen": 11432230,
+      "step": 544
+    },
+    {
+      "epoch": 0.06553237539830457,
+      "grad_norm": 2.711368073516525,
+      "learning_rate": 3.986812893708082e-06,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 11448140,
+      "step": 545
+    },
+    {
+      "epoch": 0.06565261828894367,
+      "grad_norm": 2.357659962486148,
+      "learning_rate": 3.9867234372791826e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 11465815,
+      "step": 546
+    },
+    {
+      "epoch": 0.06577286117958275,
+      "grad_norm": 1.6216400663048482,
+      "learning_rate": 3.986633679465918e-06,
+      "loss": 0.8736,
+      "num_input_tokens_seen": 11485690,
+      "step": 547
+    },
+    {
+      "epoch": 0.06589310407022185,
+      "grad_norm": 2.2378904325170477,
+      "learning_rate": 3.986543620281904e-06,
+      "loss": 0.8114,
+      "num_input_tokens_seen": 11505060,
+      "step": 548
+    },
+    {
+      "epoch": 0.06601334696086093,
+      "grad_norm": 1.7709272087727046,
+      "learning_rate": 3.986453259740802e-06,
+      "loss": 0.9079,
+      "num_input_tokens_seen": 11522950,
+      "step": 549
+    },
+    {
+      "epoch": 0.06613358985150003,
+      "grad_norm": 3.0325094513822615,
+      "learning_rate": 3.986362597856319e-06,
+      "loss": 0.7896,
+      "num_input_tokens_seen": 11539170,
+      "step": 550
+    },
+    {
+      "epoch": 0.06625383274213913,
+      "grad_norm": 2.629614271337295,
+      "learning_rate": 3.986271634642211e-06,
+      "loss": 0.8135,
+      "num_input_tokens_seen": 11555870,
+      "step": 551
+    },
+    {
+      "epoch": 0.06637407563277821,
+      "grad_norm": 2.4093144249983816,
+      "learning_rate": 3.986180370112274e-06,
+      "loss": 0.8199,
+      "num_input_tokens_seen": 11572110,
+      "step": 552
+    },
+    {
+      "epoch": 0.0664943185234173,
+      "grad_norm": 3.207173330731257,
+      "learning_rate": 3.986088804280354e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 11591560,
+      "step": 553
+    },
+    {
+      "epoch": 0.06661456141405639,
+      "grad_norm": 2.4998134604066484,
+      "learning_rate": 3.985996937160342e-06,
+      "loss": 0.935,
+      "num_input_tokens_seen": 11610470,
+      "step": 554
+    },
+    {
+      "epoch": 0.06673480430469549,
+      "grad_norm": 2.3329724698044307,
+      "learning_rate": 3.985904768766173e-06,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 11632965,
+      "step": 555
+    },
+    {
+      "epoch": 0.06685504719533458,
+      "grad_norm": 3.3615753032798015,
+      "learning_rate": 3.98581229911183e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 11651605,
+      "step": 556
+    },
+    {
+      "epoch": 0.06697529008597367,
+      "grad_norm": 1.6775108627143491,
+      "learning_rate": 3.985719528211341e-06,
+      "loss": 0.9134,
+      "num_input_tokens_seen": 11670695,
+      "step": 557
+    },
+    {
+      "epoch": 0.06709553297661276,
+      "grad_norm": 0.9879548240861006,
+      "learning_rate": 3.985626456078777e-06,
+      "loss": 0.6735,
+      "num_input_tokens_seen": 11735070,
+      "step": 558
+    },
+    {
+      "epoch": 0.06721577586725185,
+      "grad_norm": 2.5028444587692573,
+      "learning_rate": 3.985533082728259e-06,
+      "loss": 0.8653,
+      "num_input_tokens_seen": 11750445,
+      "step": 559
+    },
+    {
+      "epoch": 0.06733601875789094,
+      "grad_norm": 3.0584516804082695,
+      "learning_rate": 3.985439408173951e-06,
+      "loss": 0.7492,
+      "num_input_tokens_seen": 11770390,
+      "step": 560
+    },
+    {
+      "epoch": 0.06745626164853002,
+      "grad_norm": 3.0339846946346958,
+      "learning_rate": 3.9853454324300634e-06,
+      "loss": 0.7143,
+      "num_input_tokens_seen": 11789320,
+      "step": 561
+    },
+    {
+      "epoch": 0.06757650453916912,
+      "grad_norm": 2.411848715752077,
+      "learning_rate": 3.985251155510852e-06,
+      "loss": 0.7864,
+      "num_input_tokens_seen": 11808070,
+      "step": 562
+    },
+    {
+      "epoch": 0.06769674742980822,
+      "grad_norm": 1.809311906223256,
+      "learning_rate": 3.98515657743062e-06,
+      "loss": 0.8087,
+      "num_input_tokens_seen": 11827255,
+      "step": 563
+    },
+    {
+      "epoch": 0.0678169903204473,
+      "grad_norm": 2.3796174930654455,
+      "learning_rate": 3.985061698203711e-06,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 11844090,
+      "step": 564
+    },
+    {
+      "epoch": 0.0679372332110864,
+      "grad_norm": 0.9471033307838439,
+      "learning_rate": 3.984966517844523e-06,
+      "loss": 0.6705,
+      "num_input_tokens_seen": 11899055,
+      "step": 565
+    },
+    {
+      "epoch": 0.06805747610172548,
+      "grad_norm": 2.394016397515654,
+      "learning_rate": 3.984871036367492e-06,
+      "loss": 0.813,
+      "num_input_tokens_seen": 11918800,
+      "step": 566
+    },
+    {
+      "epoch": 0.06817771899236458,
+      "grad_norm": 2.7025524872217273,
+      "learning_rate": 3.984775253787102e-06,
+      "loss": 0.8334,
+      "num_input_tokens_seen": 11936810,
+      "step": 567
+    },
+    {
+      "epoch": 0.06829796188300366,
+      "grad_norm": 3.181706360496871,
+      "learning_rate": 3.984679170117885e-06,
+      "loss": 0.882,
+      "num_input_tokens_seen": 11952735,
+      "step": 568
+    },
+    {
+      "epoch": 0.06841820477364276,
+      "grad_norm": 4.039934765051067,
+      "learning_rate": 3.984582785374415e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 11969895,
+      "step": 569
+    },
+    {
+      "epoch": 0.06853844766428185,
+      "grad_norm": 2.0723409886287025,
+      "learning_rate": 3.9844860995713155e-06,
+      "loss": 0.8134,
+      "num_input_tokens_seen": 11989155,
+      "step": 570
+    },
+    {
+      "epoch": 0.06865869055492094,
+      "grad_norm": 2.498957001948544,
+      "learning_rate": 3.9843891127232524e-06,
+      "loss": 0.8195,
+      "num_input_tokens_seen": 12006410,
+      "step": 571
+    },
+    {
+      "epoch": 0.06877893344556003,
+      "grad_norm": 2.6896757999777052,
+      "learning_rate": 3.984291824844938e-06,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 12021225,
+      "step": 572
+    },
+    {
+      "epoch": 0.06889917633619912,
+      "grad_norm": 2.484003095385279,
+      "learning_rate": 3.984194235951132e-06,
+      "loss": 0.8459,
+      "num_input_tokens_seen": 12037090,
+      "step": 573
+    },
+    {
+      "epoch": 0.06901941922683821,
+      "grad_norm": 2.9793247518218493,
+      "learning_rate": 3.9840963460566375e-06,
+      "loss": 0.8451,
+      "num_input_tokens_seen": 12055590,
+      "step": 574
+    },
+    {
+      "epoch": 0.06913966211747731,
+      "grad_norm": 1.5502389198021764,
+      "learning_rate": 3.983998155176305e-06,
+      "loss": 0.8925,
+      "num_input_tokens_seen": 12075670,
+      "step": 575
+    },
+    {
+      "epoch": 0.06925990500811639,
+      "grad_norm": 0.9987332730721558,
+      "learning_rate": 3.9838996633250305e-06,
+      "loss": 0.5921,
+      "num_input_tokens_seen": 12135905,
+      "step": 576
+    },
+    {
+      "epoch": 0.06938014789875549,
+      "grad_norm": 2.426078566379721,
+      "learning_rate": 3.983800870517753e-06,
+      "loss": 0.8815,
+      "num_input_tokens_seen": 12152415,
+      "step": 577
+    },
+    {
+      "epoch": 0.06950039078939457,
+      "grad_norm": 3.500097985887705,
+      "learning_rate": 3.983701776769463e-06,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 12169545,
+      "step": 578
+    },
+    {
+      "epoch": 0.06962063368003367,
+      "grad_norm": 2.1469261042102312,
+      "learning_rate": 3.9836023820951885e-06,
+      "loss": 0.8543,
+      "num_input_tokens_seen": 12188480,
+      "step": 579
+    },
+    {
+      "epoch": 0.06974087657067275,
+      "grad_norm": 2.1056036573009393,
+      "learning_rate": 3.983502686510011e-06,
+      "loss": 0.6852,
+      "num_input_tokens_seen": 12209030,
+      "step": 580
+    },
+    {
+      "epoch": 0.06986111946131185,
+      "grad_norm": 1.9860317082522783,
+      "learning_rate": 3.9834026900290525e-06,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 12228145,
+      "step": 581
+    },
+    {
+      "epoch": 0.06998136235195095,
+      "grad_norm": 1.9386870743027589,
+      "learning_rate": 3.983302392667483e-06,
+      "loss": 0.9968,
+      "num_input_tokens_seen": 12248710,
+      "step": 582
+    },
+    {
+      "epoch": 0.07010160524259003,
+      "grad_norm": 1.7863563520140477,
+      "learning_rate": 3.983201794440517e-06,
+      "loss": 0.936,
+      "num_input_tokens_seen": 12268005,
+      "step": 583
+    },
+    {
+      "epoch": 0.07022184813322913,
+      "grad_norm": 1.7565383272743134,
+      "learning_rate": 3.9831008953634165e-06,
+      "loss": 0.6831,
+      "num_input_tokens_seen": 12287015,
+      "step": 584
+    },
+    {
+      "epoch": 0.07034209102386821,
+      "grad_norm": 4.215097231793874,
+      "learning_rate": 3.9829996954514864e-06,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 12305875,
+      "step": 585
+    },
+    {
+      "epoch": 0.0704623339145073,
+      "grad_norm": 1.9596951197989274,
+      "learning_rate": 3.982898194720079e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 12326325,
+      "step": 586
+    },
+    {
+      "epoch": 0.0705825768051464,
+      "grad_norm": 2.8021330753838245,
+      "learning_rate": 3.982796393184592e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 12345125,
+      "step": 587
+    },
+    {
+      "epoch": 0.07070281969578548,
+      "grad_norm": 0.8474575318761562,
+      "learning_rate": 3.98269429086047e-06,
+      "loss": 0.6509,
+      "num_input_tokens_seen": 12402685,
+      "step": 588
+    },
+    {
+      "epoch": 0.07082306258642458,
+      "grad_norm": 2.8861483170817683,
+      "learning_rate": 3.982591887763199e-06,
+      "loss": 0.8649,
+      "num_input_tokens_seen": 12419865,
+      "step": 589
+    },
+    {
+      "epoch": 0.07094330547706366,
+      "grad_norm": 2.453987272615334,
+      "learning_rate": 3.982489183908316e-06,
+      "loss": 0.8157,
+      "num_input_tokens_seen": 12436005,
+      "step": 590
+    },
+    {
+      "epoch": 0.07106354836770276,
+      "grad_norm": 1.7210524960477205,
+      "learning_rate": 3.982386179311399e-06,
+      "loss": 0.84,
+      "num_input_tokens_seen": 12456245,
+      "step": 591
+    },
+    {
+      "epoch": 0.07118379125834184,
+      "grad_norm": 5.366920478723499,
+      "learning_rate": 3.982282873988075e-06,
+      "loss": 0.8739,
+      "num_input_tokens_seen": 12473840,
+      "step": 592
+    },
+    {
+      "epoch": 0.07130403414898094,
+      "grad_norm": 1.6965908445137974,
+      "learning_rate": 3.982179267954016e-06,
+      "loss": 0.8689,
+      "num_input_tokens_seen": 12493990,
+      "step": 593
+    },
+    {
+      "epoch": 0.07142427703962004,
+      "grad_norm": 2.346572727571377,
+      "learning_rate": 3.982075361224937e-06,
+      "loss": 0.9714,
+      "num_input_tokens_seen": 12512075,
+      "step": 594
+    },
+    {
+      "epoch": 0.07154451993025912,
+      "grad_norm": 2.5202770051089964,
+      "learning_rate": 3.981971153816602e-06,
+      "loss": 0.8871,
+      "num_input_tokens_seen": 12529400,
+      "step": 595
+    },
+    {
+      "epoch": 0.07166476282089822,
+      "grad_norm": 1.5841916320621197,
+      "learning_rate": 3.981866645744819e-06,
+      "loss": 0.9521,
+      "num_input_tokens_seen": 12549835,
+      "step": 596
+    },
+    {
+      "epoch": 0.0717850057115373,
+      "grad_norm": 2.3125913889850755,
+      "learning_rate": 3.9817618370254416e-06,
+      "loss": 0.8122,
+      "num_input_tokens_seen": 12566210,
+      "step": 597
+    },
+    {
+      "epoch": 0.0719052486021764,
+      "grad_norm": 2.9059936280908767,
+      "learning_rate": 3.9816567276743684e-06,
+      "loss": 0.8656,
+      "num_input_tokens_seen": 12585795,
+      "step": 598
+    },
+    {
+      "epoch": 0.0720254914928155,
+      "grad_norm": 2.3164204614290758,
+      "learning_rate": 3.9815513177075466e-06,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 12604300,
+      "step": 599
+    },
+    {
+      "epoch": 0.07214573438345458,
+      "grad_norm": 1.6556249744709417,
+      "learning_rate": 3.9814456071409646e-06,
+      "loss": 0.7075,
+      "num_input_tokens_seen": 12624555,
+      "step": 600
+    },
+    {
+      "epoch": 0.07226597727409367,
+      "grad_norm": 2.458557845784842,
+      "learning_rate": 3.981339595990659e-06,
+      "loss": 0.8634,
+      "num_input_tokens_seen": 12642805,
+      "step": 601
+    },
+    {
+      "epoch": 0.07238622016473276,
+      "grad_norm": 2.154435173306513,
+      "learning_rate": 3.981233284272713e-06,
+      "loss": 0.8113,
+      "num_input_tokens_seen": 12662270,
+      "step": 602
+    },
+    {
+      "epoch": 0.07250646305537185,
+      "grad_norm": 1.5308595601567334,
+      "learning_rate": 3.981126672003253e-06,
+      "loss": 0.8908,
+      "num_input_tokens_seen": 12684665,
+      "step": 603
+    },
+    {
+      "epoch": 0.07262670594601094,
+      "grad_norm": 3.135527009694162,
+      "learning_rate": 3.981019759198451e-06,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 12703335,
+      "step": 604
+    },
+    {
+      "epoch": 0.07274694883665003,
+      "grad_norm": 2.079649669982467,
+      "learning_rate": 3.980912545874528e-06,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 12723220,
+      "step": 605
+    },
+    {
+      "epoch": 0.07286719172728913,
+      "grad_norm": 2.2494378998527247,
+      "learning_rate": 3.980805032047746e-06,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 12744410,
+      "step": 606
+    },
+    {
+      "epoch": 0.07298743461792821,
+      "grad_norm": 2.3893920902041406,
+      "learning_rate": 3.980697217734415e-06,
+      "loss": 0.8056,
+      "num_input_tokens_seen": 12761870,
+      "step": 607
+    },
+    {
+      "epoch": 0.07310767750856731,
+      "grad_norm": 1.9937039003108203,
+      "learning_rate": 3.980589102950891e-06,
+      "loss": 0.9136,
+      "num_input_tokens_seen": 12779755,
+      "step": 608
+    },
+    {
+      "epoch": 0.07322792039920639,
+      "grad_norm": 2.523353089931453,
+      "learning_rate": 3.9804806877135755e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 12797520,
+      "step": 609
+    },
+    {
+      "epoch": 0.07334816328984549,
+      "grad_norm": 2.2228327105234498,
+      "learning_rate": 3.980371972038915e-06,
+      "loss": 0.8655,
+      "num_input_tokens_seen": 12817730,
+      "step": 610
+    },
+    {
+      "epoch": 0.07346840618048459,
+      "grad_norm": 1.7637150102749324,
+      "learning_rate": 3.980262955943399e-06,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 12837115,
+      "step": 611
+    },
+    {
+      "epoch": 0.07358864907112367,
+      "grad_norm": 2.711290502876977,
+      "learning_rate": 3.980153639443569e-06,
+      "loss": 0.8821,
+      "num_input_tokens_seen": 12852820,
+      "step": 612
+    },
+    {
+      "epoch": 0.07370889196176277,
+      "grad_norm": 2.702649968237358,
+      "learning_rate": 3.980044022556005e-06,
+      "loss": 0.8009,
+      "num_input_tokens_seen": 12872225,
+      "step": 613
+    },
+    {
+      "epoch": 0.07382913485240185,
+      "grad_norm": 2.2520685274667596,
+      "learning_rate": 3.9799341052973375e-06,
+      "loss": 0.7245,
+      "num_input_tokens_seen": 12891780,
+      "step": 614
+    },
+    {
+      "epoch": 0.07394937774304094,
+      "grad_norm": 2.4478281279471057,
+      "learning_rate": 3.979823887684241e-06,
+      "loss": 0.7559,
+      "num_input_tokens_seen": 12910440,
+      "step": 615
+    },
+    {
+      "epoch": 0.07406962063368003,
+      "grad_norm": 2.356110600878282,
+      "learning_rate": 3.979713369733434e-06,
+      "loss": 0.8559,
+      "num_input_tokens_seen": 12928025,
+      "step": 616
+    },
+    {
+      "epoch": 0.07418986352431912,
+      "grad_norm": 2.026122785561901,
+      "learning_rate": 3.979602551461683e-06,
+      "loss": 0.841,
+      "num_input_tokens_seen": 12948525,
+      "step": 617
+    },
+    {
+      "epoch": 0.07431010641495822,
+      "grad_norm": 2.2039280579521723,
+      "learning_rate": 3.979491432885799e-06,
+      "loss": 0.9235,
+      "num_input_tokens_seen": 12964510,
+      "step": 618
+    },
+    {
+      "epoch": 0.0744303493055973,
+      "grad_norm": 2.009084833115377,
+      "learning_rate": 3.97938001402264e-06,
+      "loss": 0.8279,
+      "num_input_tokens_seen": 12983355,
+      "step": 619
+    },
+    {
+      "epoch": 0.0745505921962364,
+      "grad_norm": 3.41525478576904,
+      "learning_rate": 3.979268294889105e-06,
+      "loss": 0.812,
+      "num_input_tokens_seen": 12998625,
+      "step": 620
+    },
+    {
+      "epoch": 0.07467083508687548,
+      "grad_norm": 4.618832816567739,
+      "learning_rate": 3.979156275502143e-06,
+      "loss": 0.7523,
+      "num_input_tokens_seen": 13022005,
+      "step": 621
+    },
+    {
+      "epoch": 0.07479107797751458,
+      "grad_norm": 2.2994009478839437,
+      "learning_rate": 3.979043955878749e-06,
+      "loss": 0.9241,
+      "num_input_tokens_seen": 13039570,
+      "step": 622
+    },
+    {
+      "epoch": 0.07491132086815366,
+      "grad_norm": 2.667507308198865,
+      "learning_rate": 3.978931336035959e-06,
+      "loss": 0.8308,
+      "num_input_tokens_seen": 13058100,
+      "step": 623
+    },
+    {
+      "epoch": 0.07503156375879276,
+      "grad_norm": 2.4720470950896956,
+      "learning_rate": 3.9788184159908595e-06,
+      "loss": 0.8211,
+      "num_input_tokens_seen": 13074950,
+      "step": 624
+    },
+    {
+      "epoch": 0.07515180664943186,
+      "grad_norm": 3.2995386523457624,
+      "learning_rate": 3.97870519576058e-06,
+      "loss": 0.8234,
+      "num_input_tokens_seen": 13091095,
+      "step": 625
+    },
+    {
+      "epoch": 0.07527204954007094,
+      "grad_norm": 2.431466590709614,
+      "learning_rate": 3.978591675362295e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 13109530,
+      "step": 626
+    },
+    {
+      "epoch": 0.07539229243071004,
+      "grad_norm": 2.186595191657186,
+      "learning_rate": 3.978477854813226e-06,
+      "loss": 0.8785,
+      "num_input_tokens_seen": 13128590,
+      "step": 627
+    },
+    {
+      "epoch": 0.07551253532134912,
+      "grad_norm": 1.8913845495403419,
+      "learning_rate": 3.97836373413064e-06,
+      "loss": 0.8281,
+      "num_input_tokens_seen": 13146365,
+      "step": 628
+    },
+    {
+      "epoch": 0.07563277821198822,
+      "grad_norm": 1.899667587905074,
+      "learning_rate": 3.978249313331848e-06,
+      "loss": 0.75,
+      "num_input_tokens_seen": 13164315,
+      "step": 629
+    },
+    {
+      "epoch": 0.07575302110262731,
+      "grad_norm": 5.47117097086709,
+      "learning_rate": 3.978134592434208e-06,
+      "loss": 0.6289,
+      "num_input_tokens_seen": 13181785,
+      "step": 630
+    },
+    {
+      "epoch": 0.0758732639932664,
+      "grad_norm": 1.1013894868521719,
+      "learning_rate": 3.978019571455123e-06,
+      "loss": 0.6476,
+      "num_input_tokens_seen": 13233450,
+      "step": 631
+    },
+    {
+      "epoch": 0.07599350688390549,
+      "grad_norm": 2.1038314624286376,
+      "learning_rate": 3.977904250412042e-06,
+      "loss": 0.848,
+      "num_input_tokens_seen": 13252125,
+      "step": 632
+    },
+    {
+      "epoch": 0.07611374977454458,
+      "grad_norm": 2.6170011284608776,
+      "learning_rate": 3.97778862932246e-06,
+      "loss": 0.8636,
+      "num_input_tokens_seen": 13269010,
+      "step": 633
+    },
+    {
+      "epoch": 0.07623399266518367,
+      "grad_norm": 2.001324452203999,
+      "learning_rate": 3.9776727082039144e-06,
+      "loss": 0.9345,
+      "num_input_tokens_seen": 13285700,
+      "step": 634
+    },
+    {
+      "epoch": 0.07635423555582276,
+      "grad_norm": 0.9320747082599746,
+      "learning_rate": 3.977556487073991e-06,
+      "loss": 0.5866,
+      "num_input_tokens_seen": 13339975,
+      "step": 635
+    },
+    {
+      "epoch": 0.07647447844646185,
+      "grad_norm": 1.9892059577342283,
+      "learning_rate": 3.97743996595032e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 13359735,
+      "step": 636
+    },
+    {
+      "epoch": 0.07659472133710095,
+      "grad_norm": 1.664456914361808,
+      "learning_rate": 3.9773231448505804e-06,
+      "loss": 0.8174,
+      "num_input_tokens_seen": 13381245,
+      "step": 637
+    },
+    {
+      "epoch": 0.07671496422774003,
+      "grad_norm": 1.9811590021932886,
+      "learning_rate": 3.977206023792491e-06,
+      "loss": 0.7667,
+      "num_input_tokens_seen": 13400855,
+      "step": 638
+    },
+    {
+      "epoch": 0.07683520711837913,
+      "grad_norm": 2.905793672851696,
+      "learning_rate": 3.97708860279382e-06,
+      "loss": 0.8166,
+      "num_input_tokens_seen": 13418685,
+      "step": 639
+    },
+    {
+      "epoch": 0.07695545000901821,
+      "grad_norm": 1.7325124029352574,
+      "learning_rate": 3.97697088187238e-06,
+      "loss": 0.7888,
+      "num_input_tokens_seen": 13438920,
+      "step": 640
+    },
+    {
+      "epoch": 0.07707569289965731,
+      "grad_norm": 2.00514909654152,
+      "learning_rate": 3.976852861046029e-06,
+      "loss": 0.9079,
+      "num_input_tokens_seen": 13455255,
+      "step": 641
+    },
+    {
+      "epoch": 0.0771959357902964,
+      "grad_norm": 1.545971169790335,
+      "learning_rate": 3.97673454033267e-06,
+      "loss": 0.7941,
+      "num_input_tokens_seen": 13477075,
+      "step": 642
+    },
+    {
+      "epoch": 0.07731617868093549,
+      "grad_norm": 2.134137000040373,
+      "learning_rate": 3.976615919750254e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 13494495,
+      "step": 643
+    },
+    {
+      "epoch": 0.07743642157157458,
+      "grad_norm": 1.9974233890778168,
+      "learning_rate": 3.976496999316775e-06,
+      "loss": 0.8633,
+      "num_input_tokens_seen": 13512970,
+      "step": 644
+    },
+    {
+      "epoch": 0.07755666446221367,
+      "grad_norm": 2.819701164318358,
+      "learning_rate": 3.976377779050271e-06,
+      "loss": 0.8406,
+      "num_input_tokens_seen": 13530820,
+      "step": 645
+    },
+    {
+      "epoch": 0.07767690735285276,
+      "grad_norm": 2.3351755050819896,
+      "learning_rate": 3.976258258968831e-06,
+      "loss": 0.8353,
+      "num_input_tokens_seen": 13549085,
+      "step": 646
+    },
+    {
+      "epoch": 0.07779715024349185,
+      "grad_norm": 2.858789616542731,
+      "learning_rate": 3.976138439090583e-06,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 13566885,
+      "step": 647
+    },
+    {
+      "epoch": 0.07791739313413094,
+      "grad_norm": 2.376160404782646,
+      "learning_rate": 3.976018319433706e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 13584150,
+      "step": 648
+    },
+    {
+      "epoch": 0.07803763602477004,
+      "grad_norm": 5.744370796872292,
+      "learning_rate": 3.9758979000164205e-06,
+      "loss": 0.9185,
+      "num_input_tokens_seen": 13600690,
+      "step": 649
+    },
+    {
+      "epoch": 0.07815787891540912,
+      "grad_norm": 2.751184724187577,
+      "learning_rate": 3.975777180856995e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 13619530,
+      "step": 650
+    },
+    {
+      "epoch": 0.07827812180604822,
+      "grad_norm": 2.471317199751255,
+      "learning_rate": 3.975656161973742e-06,
+      "loss": 0.8673,
+      "num_input_tokens_seen": 13638335,
+      "step": 651
+    },
+    {
+      "epoch": 0.0783983646966873,
+      "grad_norm": 2.5016421566209517,
+      "learning_rate": 3.9755348433850194e-06,
+      "loss": 0.8896,
+      "num_input_tokens_seen": 13653395,
+      "step": 652
+    },
+    {
+      "epoch": 0.0785186075873264,
+      "grad_norm": 1.1513181630416465,
+      "learning_rate": 3.975413225109232e-06,
+      "loss": 0.7172,
+      "num_input_tokens_seen": 13713665,
+      "step": 653
+    },
+    {
+      "epoch": 0.0786388504779655,
+      "grad_norm": 3.6696349499046605,
+      "learning_rate": 3.975291307164829e-06,
+      "loss": 0.9334,
+      "num_input_tokens_seen": 13732030,
+      "step": 654
+    },
+    {
+      "epoch": 0.07875909336860458,
+      "grad_norm": 2.140405776309736,
+      "learning_rate": 3.975169089570306e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 13750125,
+      "step": 655
+    },
+    {
+      "epoch": 0.07887933625924368,
+      "grad_norm": 3.300815460154189,
+      "learning_rate": 3.975046572344202e-06,
+      "loss": 0.9103,
+      "num_input_tokens_seen": 13766305,
+      "step": 656
+    },
+    {
+      "epoch": 0.07899957914988276,
+      "grad_norm": 2.1186246298588185,
+      "learning_rate": 3.974923755505103e-06,
+      "loss": 0.7223,
+      "num_input_tokens_seen": 13785255,
+      "step": 657
+    },
+    {
+      "epoch": 0.07911982204052186,
+      "grad_norm": 1.6538269995736954,
+      "learning_rate": 3.974800639071641e-06,
+      "loss": 0.9001,
+      "num_input_tokens_seen": 13805695,
+      "step": 658
+    },
+    {
+      "epoch": 0.07924006493116094,
+      "grad_norm": 2.356686144343739,
+      "learning_rate": 3.974677223062492e-06,
+      "loss": 1.0068,
+      "num_input_tokens_seen": 13822630,
+      "step": 659
+    },
+    {
+      "epoch": 0.07936030782180004,
+      "grad_norm": 8.20502877875863,
+      "learning_rate": 3.974553507496378e-06,
+      "loss": 0.7419,
+      "num_input_tokens_seen": 13840925,
+      "step": 660
+    },
+    {
+      "epoch": 0.07948055071243913,
+      "grad_norm": 2.133690918767308,
+      "learning_rate": 3.974429492392068e-06,
+      "loss": 0.8789,
+      "num_input_tokens_seen": 13860670,
+      "step": 661
+    },
+    {
+      "epoch": 0.07960079360307822,
+      "grad_norm": 2.516218640713848,
+      "learning_rate": 3.974305177768373e-06,
+      "loss": 0.9021,
+      "num_input_tokens_seen": 13878600,
+      "step": 662
+    },
+    {
+      "epoch": 0.07972103649371731,
+      "grad_norm": 2.4852367365637718,
+      "learning_rate": 3.974180563644152e-06,
+      "loss": 0.8472,
+      "num_input_tokens_seen": 13896885,
+      "step": 663
+    },
+    {
+      "epoch": 0.0798412793843564,
+      "grad_norm": 2.3134242343203257,
+      "learning_rate": 3.97405565003831e-06,
+      "loss": 0.8874,
+      "num_input_tokens_seen": 13912690,
+      "step": 664
+    },
+    {
+      "epoch": 0.07996152227499549,
+      "grad_norm": 2.5881119570541826,
+      "learning_rate": 3.973930436969794e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 13930865,
+      "step": 665
+    },
+    {
+      "epoch": 0.08008176516563459,
+      "grad_norm": 2.02909855328078,
+      "learning_rate": 3.973804924457602e-06,
+      "loss": 0.8578,
+      "num_input_tokens_seen": 13948665,
+      "step": 666
+    },
+    {
+      "epoch": 0.08020200805627367,
+      "grad_norm": 1.8463505585984528,
+      "learning_rate": 3.973679112520771e-06,
+      "loss": 0.8503,
+      "num_input_tokens_seen": 13970100,
+      "step": 667
+    },
+    {
+      "epoch": 0.08032225094691277,
+      "grad_norm": 1.9982400017381965,
+      "learning_rate": 3.973553001178389e-06,
+      "loss": 0.987,
+      "num_input_tokens_seen": 13987325,
+      "step": 668
+    },
+    {
+      "epoch": 0.08044249383755185,
+      "grad_norm": 2.1371043066152686,
+      "learning_rate": 3.973426590449585e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 14005000,
+      "step": 669
+    },
+    {
+      "epoch": 0.08056273672819095,
+      "grad_norm": 2.1604572631445897,
+      "learning_rate": 3.9732998803535364e-06,
+      "loss": 0.7592,
+      "num_input_tokens_seen": 14022780,
+      "step": 670
+    },
+    {
+      "epoch": 0.08068297961883003,
+      "grad_norm": 2.361880208337802,
+      "learning_rate": 3.973172870909465e-06,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 14037265,
+      "step": 671
+    },
+    {
+      "epoch": 0.08080322250946913,
+      "grad_norm": 2.4549367924767385,
+      "learning_rate": 3.973045562136638e-06,
+      "loss": 0.8122,
+      "num_input_tokens_seen": 14053800,
+      "step": 672
+    },
+    {
+      "epoch": 0.08092346540010822,
+      "grad_norm": 2.2493861764505767,
+      "learning_rate": 3.972917954054368e-06,
+      "loss": 0.9059,
+      "num_input_tokens_seen": 14072075,
+      "step": 673
+    },
+    {
+      "epoch": 0.08104370829074731,
+      "grad_norm": 2.3594301558737754,
+      "learning_rate": 3.972790046682013e-06,
+      "loss": 0.8122,
+      "num_input_tokens_seen": 14090470,
+      "step": 674
+    },
+    {
+      "epoch": 0.0811639511813864,
+      "grad_norm": 3.098041091611056,
+      "learning_rate": 3.972661840038977e-06,
+      "loss": 0.796,
+      "num_input_tokens_seen": 14110480,
+      "step": 675
+    },
+    {
+      "epoch": 0.08128419407202549,
+      "grad_norm": 2.362030104678018,
+      "learning_rate": 3.972533334144707e-06,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 14127125,
+      "step": 676
+    },
+    {
+      "epoch": 0.08140443696266458,
+      "grad_norm": 2.6912333012220215,
+      "learning_rate": 3.972404529018699e-06,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 14146705,
+      "step": 677
+    },
+    {
+      "epoch": 0.08152467985330367,
+      "grad_norm": 1.8724244685298441,
+      "learning_rate": 3.972275424680493e-06,
+      "loss": 0.8552,
+      "num_input_tokens_seen": 14166535,
+      "step": 678
+    },
+    {
+      "epoch": 0.08164492274394276,
+      "grad_norm": 2.5566213312496706,
+      "learning_rate": 3.972146021149673e-06,
+      "loss": 0.9084,
+      "num_input_tokens_seen": 14184530,
+      "step": 679
+    },
+    {
+      "epoch": 0.08176516563458186,
+      "grad_norm": 2.6648812216977773,
+      "learning_rate": 3.972016318445868e-06,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 14202250,
+      "step": 680
+    },
+    {
+      "epoch": 0.08188540852522094,
+      "grad_norm": 2.1998959665602746,
+      "learning_rate": 3.971886316588757e-06,
+      "loss": 0.9064,
+      "num_input_tokens_seen": 14222475,
+      "step": 681
+    },
+    {
+      "epoch": 0.08200565141586004,
+      "grad_norm": 3.0602546884036093,
+      "learning_rate": 3.9717560155980595e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 14237845,
+      "step": 682
+    },
+    {
+      "epoch": 0.08212589430649912,
+      "grad_norm": 1.9067787928465365,
+      "learning_rate": 3.971625415493542e-06,
+      "loss": 0.9266,
+      "num_input_tokens_seen": 14255885,
+      "step": 683
+    },
+    {
+      "epoch": 0.08224613719713822,
+      "grad_norm": 1.9935229796720129,
+      "learning_rate": 3.971494516295017e-06,
+      "loss": 0.8697,
+      "num_input_tokens_seen": 14275055,
+      "step": 684
+    },
+    {
+      "epoch": 0.08236638008777732,
+      "grad_norm": 2.034360920203988,
+      "learning_rate": 3.971363318022341e-06,
+      "loss": 0.8461,
+      "num_input_tokens_seen": 14296115,
+      "step": 685
+    },
+    {
+      "epoch": 0.0824866229784164,
+      "grad_norm": 2.027756637615474,
+      "learning_rate": 3.971231820695417e-06,
+      "loss": 0.6818,
+      "num_input_tokens_seen": 14319450,
+      "step": 686
+    },
+    {
+      "epoch": 0.0826068658690555,
+      "grad_norm": 2.4887209020018317,
+      "learning_rate": 3.971100024334193e-06,
+      "loss": 0.8098,
+      "num_input_tokens_seen": 14336690,
+      "step": 687
+    },
+    {
+      "epoch": 0.08272710875969458,
+      "grad_norm": 2.0144004566929765,
+      "learning_rate": 3.970967928958663e-06,
+      "loss": 0.8564,
+      "num_input_tokens_seen": 14353525,
+      "step": 688
+    },
+    {
+      "epoch": 0.08284735165033368,
+      "grad_norm": 1.7497338639848277,
+      "learning_rate": 3.970835534588865e-06,
+      "loss": 0.8309,
+      "num_input_tokens_seen": 14370740,
+      "step": 689
+    },
+    {
+      "epoch": 0.08296759454097276,
+      "grad_norm": 1.7206593880512195,
+      "learning_rate": 3.970702841244883e-06,
+      "loss": 0.8536,
+      "num_input_tokens_seen": 14388780,
+      "step": 690
+    },
+    {
+      "epoch": 0.08308783743161186,
+      "grad_norm": 2.031411152548322,
+      "learning_rate": 3.970569848946847e-06,
+      "loss": 0.8273,
+      "num_input_tokens_seen": 14408315,
+      "step": 691
+    },
+    {
+      "epoch": 0.08320808032225095,
+      "grad_norm": 2.6741784887884057,
+      "learning_rate": 3.970436557714932e-06,
+      "loss": 0.825,
+      "num_input_tokens_seen": 14424555,
+      "step": 692
+    },
+    {
+      "epoch": 0.08332832321289003,
+      "grad_norm": 2.0452193555434426,
+      "learning_rate": 3.970302967569358e-06,
+      "loss": 0.8547,
+      "num_input_tokens_seen": 14442865,
+      "step": 693
+    },
+    {
+      "epoch": 0.08344856610352913,
+      "grad_norm": 1.963202225097715,
+      "learning_rate": 3.9701690785303896e-06,
+      "loss": 0.687,
+      "num_input_tokens_seen": 14461780,
+      "step": 694
+    },
+    {
+      "epoch": 0.08356880899416821,
+      "grad_norm": 2.6151209058241403,
+      "learning_rate": 3.970034890618339e-06,
+      "loss": 0.8796,
+      "num_input_tokens_seen": 14481190,
+      "step": 695
+    },
+    {
+      "epoch": 0.08368905188480731,
+      "grad_norm": 2.020238884687549,
+      "learning_rate": 3.969900403853562e-06,
+      "loss": 0.8734,
+      "num_input_tokens_seen": 14499950,
+      "step": 696
+    },
+    {
+      "epoch": 0.08380929477544641,
+      "grad_norm": 2.798472618251497,
+      "learning_rate": 3.96976561825646e-06,
+      "loss": 0.7787,
+      "num_input_tokens_seen": 14516760,
+      "step": 697
+    },
+    {
+      "epoch": 0.08392953766608549,
+      "grad_norm": 2.0945113681639334,
+      "learning_rate": 3.969630533847479e-06,
+      "loss": 0.8741,
+      "num_input_tokens_seen": 14535440,
+      "step": 698
+    },
+    {
+      "epoch": 0.08404978055672459,
+      "grad_norm": 1.9581714131200367,
+      "learning_rate": 3.969495150647113e-06,
+      "loss": 0.8508,
+      "num_input_tokens_seen": 14553330,
+      "step": 699
+    },
+    {
+      "epoch": 0.08417002344736367,
+      "grad_norm": 2.5493550795950877,
+      "learning_rate": 3.969359468675899e-06,
+      "loss": 0.7649,
+      "num_input_tokens_seen": 14573180,
+      "step": 700
+    },
+    {
+      "epoch": 0.08429026633800277,
+      "grad_norm": 1.9733296071230897,
+      "learning_rate": 3.969223487954418e-06,
+      "loss": 0.8948,
+      "num_input_tokens_seen": 14590360,
+      "step": 701
+    },
+    {
+      "epoch": 0.08441050922864185,
+      "grad_norm": 2.1040598215163953,
+      "learning_rate": 3.969087208503301e-06,
+      "loss": 0.8243,
+      "num_input_tokens_seen": 14610160,
+      "step": 702
+    },
+    {
+      "epoch": 0.08453075211928095,
+      "grad_norm": 3.247026909668937,
+      "learning_rate": 3.968950630343219e-06,
+      "loss": 0.8413,
+      "num_input_tokens_seen": 14626865,
+      "step": 703
+    },
+    {
+      "epoch": 0.08465099500992004,
+      "grad_norm": 3.107297568930794,
+      "learning_rate": 3.968813753494892e-06,
+      "loss": 0.9291,
+      "num_input_tokens_seen": 14644745,
+      "step": 704
+    },
+    {
+      "epoch": 0.08477123790055913,
+      "grad_norm": 14.507405002489918,
+      "learning_rate": 3.968676577979084e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 14664015,
+      "step": 705
+    },
+    {
+      "epoch": 0.08489148079119822,
+      "grad_norm": 3.2302177816524655,
+      "learning_rate": 3.968539103816605e-06,
+      "loss": 0.7812,
+      "num_input_tokens_seen": 14681535,
+      "step": 706
+    },
+    {
+      "epoch": 0.0850117236818373,
+      "grad_norm": 2.161604433236569,
+      "learning_rate": 3.9684013310283085e-06,
+      "loss": 0.8893,
+      "num_input_tokens_seen": 14699940,
+      "step": 707
+    },
+    {
+      "epoch": 0.0851319665724764,
+      "grad_norm": 3.902670916707156,
+      "learning_rate": 3.9682632596350956e-06,
+      "loss": 0.6461,
+      "num_input_tokens_seen": 14720825,
+      "step": 708
+    },
+    {
+      "epoch": 0.0852522094631155,
+      "grad_norm": 2.375580443107337,
+      "learning_rate": 3.968124889657911e-06,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 14735645,
+      "step": 709
+    },
+    {
+      "epoch": 0.08537245235375458,
+      "grad_norm": 4.3180750225749795,
+      "learning_rate": 3.967986221117746e-06,
+      "loss": 0.9005,
+      "num_input_tokens_seen": 14751305,
+      "step": 710
+    },
+    {
+      "epoch": 0.08549269524439368,
+      "grad_norm": 3.2832451594733776,
+      "learning_rate": 3.967847254035635e-06,
+      "loss": 0.8662,
+      "num_input_tokens_seen": 14770410,
+      "step": 711
+    },
+    {
+      "epoch": 0.08561293813503276,
+      "grad_norm": 10.396563553214172,
+      "learning_rate": 3.967707988432661e-06,
+      "loss": 0.87,
+      "num_input_tokens_seen": 14787835,
+      "step": 712
+    },
+    {
+      "epoch": 0.08573318102567186,
+      "grad_norm": 2.7680758992832253,
+      "learning_rate": 3.967568424329949e-06,
+      "loss": 0.8809,
+      "num_input_tokens_seen": 14807980,
+      "step": 713
+    },
+    {
+      "epoch": 0.08585342391631094,
+      "grad_norm": 0.8830994370187513,
+      "learning_rate": 3.967428561748671e-06,
+      "loss": 0.5977,
+      "num_input_tokens_seen": 14875670,
+      "step": 714
+    },
+    {
+      "epoch": 0.08597366680695004,
+      "grad_norm": 3.127153628250816,
+      "learning_rate": 3.967288400710045e-06,
+      "loss": 0.8724,
+      "num_input_tokens_seen": 14894855,
+      "step": 715
+    },
+    {
+      "epoch": 0.08609390969758914,
+      "grad_norm": 2.7594761089139164,
+      "learning_rate": 3.9671479412353335e-06,
+      "loss": 0.8783,
+      "num_input_tokens_seen": 14913040,
+      "step": 716
+    },
+    {
+      "epoch": 0.08621415258822822,
+      "grad_norm": 2.4554005012340547,
+      "learning_rate": 3.967007183345843e-06,
+      "loss": 0.7454,
+      "num_input_tokens_seen": 14932615,
+      "step": 717
+    },
+    {
+      "epoch": 0.08633439547886732,
+      "grad_norm": 2.362434679121161,
+      "learning_rate": 3.966866127062927e-06,
+      "loss": 0.8987,
+      "num_input_tokens_seen": 14949460,
+      "step": 718
+    },
+    {
+      "epoch": 0.0864546383695064,
+      "grad_norm": 0.9745794720445365,
+      "learning_rate": 3.966724772407982e-06,
+      "loss": 0.6818,
+      "num_input_tokens_seen": 15006695,
+      "step": 719
+    },
+    {
+      "epoch": 0.0865748812601455,
+      "grad_norm": 23.01756729782883,
+      "learning_rate": 3.966583119402454e-06,
+      "loss": 0.8794,
+      "num_input_tokens_seen": 15023180,
+      "step": 720
+    },
+    {
+      "epoch": 0.08669512415078459,
+      "grad_norm": 1.6756533688020472,
+      "learning_rate": 3.9664411680678305e-06,
+      "loss": 0.8196,
+      "num_input_tokens_seen": 15044655,
+      "step": 721
+    },
+    {
+      "epoch": 0.08681536704142367,
+      "grad_norm": 0.9522742056414344,
+      "learning_rate": 3.966298918425644e-06,
+      "loss": 0.6426,
+      "num_input_tokens_seen": 15101865,
+      "step": 722
+    },
+    {
+      "epoch": 0.08693560993206277,
+      "grad_norm": 2.4526747326126492,
+      "learning_rate": 3.966156370497476e-06,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 15125195,
+      "step": 723
+    },
+    {
+      "epoch": 0.08705585282270185,
+      "grad_norm": 1.9182045782137724,
+      "learning_rate": 3.96601352430495e-06,
+      "loss": 0.8852,
+      "num_input_tokens_seen": 15144685,
+      "step": 724
+    },
+    {
+      "epoch": 0.08717609571334095,
+      "grad_norm": 2.2229430420804173,
+      "learning_rate": 3.965870379869735e-06,
+      "loss": 0.8289,
+      "num_input_tokens_seen": 15166450,
+      "step": 725
+    },
+    {
+      "epoch": 0.08729633860398003,
+      "grad_norm": 6.086371157458483,
+      "learning_rate": 3.965726937213547e-06,
+      "loss": 0.8609,
+      "num_input_tokens_seen": 15184805,
+      "step": 726
+    },
+    {
+      "epoch": 0.08741658149461913,
+      "grad_norm": 2.522919711400508,
+      "learning_rate": 3.965583196358144e-06,
+      "loss": 0.798,
+      "num_input_tokens_seen": 15203560,
+      "step": 727
+    },
+    {
+      "epoch": 0.08753682438525823,
+      "grad_norm": 2.2648971305235137,
+      "learning_rate": 3.965439157325335e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 15220645,
+      "step": 728
+    },
+    {
+      "epoch": 0.08765706727589731,
+      "grad_norm": 2.5509349094510654,
+      "learning_rate": 3.965294820136968e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 15242165,
+      "step": 729
+    },
+    {
+      "epoch": 0.08777731016653641,
+      "grad_norm": 2.6537668532063443,
+      "learning_rate": 3.965150184814938e-06,
+      "loss": 0.8634,
+      "num_input_tokens_seen": 15261370,
+      "step": 730
+    },
+    {
+      "epoch": 0.08789755305717549,
+      "grad_norm": 3.0779170657643014,
+      "learning_rate": 3.965005251381189e-06,
+      "loss": 0.752,
+      "num_input_tokens_seen": 15279025,
+      "step": 731
+    },
+    {
+      "epoch": 0.08801779594781459,
+      "grad_norm": 0.9427581676011325,
+      "learning_rate": 3.964860019857705e-06,
+      "loss": 0.6676,
+      "num_input_tokens_seen": 15343660,
+      "step": 732
+    },
+    {
+      "epoch": 0.08813803883845367,
+      "grad_norm": 3.0093943758101096,
+      "learning_rate": 3.964714490266518e-06,
+      "loss": 0.8339,
+      "num_input_tokens_seen": 15364025,
+      "step": 733
+    },
+    {
+      "epoch": 0.08825828172909277,
+      "grad_norm": 0.9335780088441266,
+      "learning_rate": 3.964568662629706e-06,
+      "loss": 0.6608,
+      "num_input_tokens_seen": 15425050,
+      "step": 734
+    },
+    {
+      "epoch": 0.08837852461973186,
+      "grad_norm": 2.791396337546597,
+      "learning_rate": 3.9644225369693895e-06,
+      "loss": 0.8389,
+      "num_input_tokens_seen": 15445070,
+      "step": 735
+    },
+    {
+      "epoch": 0.08849876751037095,
+      "grad_norm": 2.35220968070547,
+      "learning_rate": 3.964276113307735e-06,
+      "loss": 0.8763,
+      "num_input_tokens_seen": 15464755,
+      "step": 736
+    },
+    {
+      "epoch": 0.08861901040101004,
+      "grad_norm": 2.2433476881964216,
+      "learning_rate": 3.9641293916669574e-06,
+      "loss": 0.8068,
+      "num_input_tokens_seen": 15483435,
+      "step": 737
+    },
+    {
+      "epoch": 0.08873925329164913,
+      "grad_norm": 1.8917270344456076,
+      "learning_rate": 3.9639823720693115e-06,
+      "loss": 0.8288,
+      "num_input_tokens_seen": 15505010,
+      "step": 738
+    },
+    {
+      "epoch": 0.08885949618228822,
+      "grad_norm": 0.9172565789983289,
+      "learning_rate": 3.963835054537102e-06,
+      "loss": 0.6523,
+      "num_input_tokens_seen": 15573695,
+      "step": 739
+    },
+    {
+      "epoch": 0.08897973907292732,
+      "grad_norm": 3.213698472725518,
+      "learning_rate": 3.963687439092676e-06,
+      "loss": 0.6137,
+      "num_input_tokens_seen": 15594100,
+      "step": 740
+    },
+    {
+      "epoch": 0.0890999819635664,
+      "grad_norm": 3.2958593131421465,
+      "learning_rate": 3.963539525758427e-06,
+      "loss": 0.8,
+      "num_input_tokens_seen": 15613380,
+      "step": 741
+    },
+    {
+      "epoch": 0.0892202248542055,
+      "grad_norm": 2.008049575068697,
+      "learning_rate": 3.9633913145567925e-06,
+      "loss": 0.6789,
+      "num_input_tokens_seen": 15633590,
+      "step": 742
+    },
+    {
+      "epoch": 0.08934046774484458,
+      "grad_norm": 2.429284120730918,
+      "learning_rate": 3.9632428055102575e-06,
+      "loss": 0.8113,
+      "num_input_tokens_seen": 15653320,
+      "step": 743
+    },
+    {
+      "epoch": 0.08946071063548368,
+      "grad_norm": 2.0947660088439384,
+      "learning_rate": 3.9630939986413495e-06,
+      "loss": 0.6739,
+      "num_input_tokens_seen": 15674840,
+      "step": 744
+    },
+    {
+      "epoch": 0.08958095352612276,
+      "grad_norm": 1.7908076624810088,
+      "learning_rate": 3.962944893972643e-06,
+      "loss": 0.7824,
+      "num_input_tokens_seen": 15693010,
+      "step": 745
+    },
+    {
+      "epoch": 0.08970119641676186,
+      "grad_norm": 3.521169557438839,
+      "learning_rate": 3.962795491526756e-06,
+      "loss": 0.9163,
+      "num_input_tokens_seen": 15709890,
+      "step": 746
+    },
+    {
+      "epoch": 0.08982143930740095,
+      "grad_norm": 3.377591863232413,
+      "learning_rate": 3.962645791326354e-06,
+      "loss": 0.891,
+      "num_input_tokens_seen": 15728865,
+      "step": 747
+    },
+    {
+      "epoch": 0.08994168219804004,
+      "grad_norm": 2.127211643559933,
+      "learning_rate": 3.962495793394146e-06,
+      "loss": 0.8281,
+      "num_input_tokens_seen": 15747775,
+      "step": 748
+    },
+    {
+      "epoch": 0.09006192508867913,
+      "grad_norm": 0.8535635028227297,
+      "learning_rate": 3.9623454977528864e-06,
+      "loss": 0.614,
+      "num_input_tokens_seen": 15806150,
+      "step": 749
+    },
+    {
+      "epoch": 0.09018216797931822,
+      "grad_norm": 2.1415739231061433,
+      "learning_rate": 3.962194904425375e-06,
+      "loss": 0.8434,
+      "num_input_tokens_seen": 15826500,
+      "step": 750
+    },
+    {
+      "epoch": 0.09030241086995731,
+      "grad_norm": 2.131746767776979,
+      "learning_rate": 3.9620440134344566e-06,
+      "loss": 0.6823,
+      "num_input_tokens_seen": 15844375,
+      "step": 751
+    },
+    {
+      "epoch": 0.09042265376059641,
+      "grad_norm": 2.1946167727437853,
+      "learning_rate": 3.9618928248030215e-06,
+      "loss": 0.8209,
+      "num_input_tokens_seen": 15863605,
+      "step": 752
+    },
+    {
+      "epoch": 0.0905428966512355,
+      "grad_norm": 3.333589034009174,
+      "learning_rate": 3.961741338554005e-06,
+      "loss": 0.8355,
+      "num_input_tokens_seen": 15881665,
+      "step": 753
+    },
+    {
+      "epoch": 0.09066313954187459,
+      "grad_norm": 2.691967012825218,
+      "learning_rate": 3.9615895547103865e-06,
+      "loss": 0.7532,
+      "num_input_tokens_seen": 15905030,
+      "step": 754
+    },
+    {
+      "epoch": 0.09078338243251367,
+      "grad_norm": 2.1934455670244337,
+      "learning_rate": 3.961437473295193e-06,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 15924895,
+      "step": 755
+    },
+    {
+      "epoch": 0.09090362532315277,
+      "grad_norm": 2.6805744877034945,
+      "learning_rate": 3.961285094331495e-06,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 15942530,
+      "step": 756
+    },
+    {
+      "epoch": 0.09102386821379185,
+      "grad_norm": 2.083728331683761,
+      "learning_rate": 3.961132417842406e-06,
+      "loss": 0.8555,
+      "num_input_tokens_seen": 15962035,
+      "step": 757
+    },
+    {
+      "epoch": 0.09114411110443095,
+      "grad_norm": 4.301934916923718,
+      "learning_rate": 3.960979443851089e-06,
+      "loss": 0.7509,
+      "num_input_tokens_seen": 15978780,
+      "step": 758
+    },
+    {
+      "epoch": 0.09126435399507005,
+      "grad_norm": 1.8769589630604666,
+      "learning_rate": 3.96082617238075e-06,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 16001125,
+      "step": 759
+    },
+    {
+      "epoch": 0.09138459688570913,
+      "grad_norm": 3.416100826591661,
+      "learning_rate": 3.960672603454639e-06,
+      "loss": 0.7973,
+      "num_input_tokens_seen": 16020825,
+      "step": 760
+    },
+    {
+      "epoch": 0.09150483977634823,
+      "grad_norm": 3.035661154129103,
+      "learning_rate": 3.960518737096054e-06,
+      "loss": 0.7705,
+      "num_input_tokens_seen": 16040175,
+      "step": 761
+    },
+    {
+      "epoch": 0.09162508266698731,
+      "grad_norm": 2.568595606205626,
+      "learning_rate": 3.960364573328334e-06,
+      "loss": 0.7339,
+      "num_input_tokens_seen": 16059220,
+      "step": 762
+    },
+    {
+      "epoch": 0.0917453255576264,
+      "grad_norm": 3.238048025015355,
+      "learning_rate": 3.9602101121748675e-06,
+      "loss": 0.8778,
+      "num_input_tokens_seen": 16079435,
+      "step": 763
+    },
+    {
+      "epoch": 0.0918655684482655,
+      "grad_norm": 2.807019953961182,
+      "learning_rate": 3.960055353659085e-06,
+      "loss": 0.7216,
+      "num_input_tokens_seen": 16096265,
+      "step": 764
+    },
+    {
+      "epoch": 0.09198581133890459,
+      "grad_norm": 2.526834797827245,
+      "learning_rate": 3.959900297804465e-06,
+      "loss": 0.8372,
+      "num_input_tokens_seen": 16116155,
+      "step": 765
+    },
+    {
+      "epoch": 0.09210605422954368,
+      "grad_norm": 2.090176035795695,
+      "learning_rate": 3.9597449446345276e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 16133120,
+      "step": 766
+    },
+    {
+      "epoch": 0.09222629712018277,
+      "grad_norm": 2.4614232962855316,
+      "learning_rate": 3.95958929417284e-06,
+      "loss": 0.8304,
+      "num_input_tokens_seen": 16150995,
+      "step": 767
+    },
+    {
+      "epoch": 0.09234654001082186,
+      "grad_norm": 0.8098031810162347,
+      "learning_rate": 3.9594333464430145e-06,
+      "loss": 0.6114,
+      "num_input_tokens_seen": 16220205,
+      "step": 768
+    },
+    {
+      "epoch": 0.09246678290146094,
+      "grad_norm": 2.3417263053559667,
+      "learning_rate": 3.959277101468709e-06,
+      "loss": 0.8784,
+      "num_input_tokens_seen": 16239475,
+      "step": 769
+    },
+    {
+      "epoch": 0.09258702579210004,
+      "grad_norm": 2.8853235259600254,
+      "learning_rate": 3.959120559273624e-06,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 16256980,
+      "step": 770
+    },
+    {
+      "epoch": 0.09270726868273914,
+      "grad_norm": 3.0027926659166178,
+      "learning_rate": 3.958963719881509e-06,
+      "loss": 0.8404,
+      "num_input_tokens_seen": 16274790,
+      "step": 771
+    },
+    {
+      "epoch": 0.09282751157337822,
+      "grad_norm": 2.2175153618136956,
+      "learning_rate": 3.958806583316154e-06,
+      "loss": 0.9344,
+      "num_input_tokens_seen": 16292480,
+      "step": 772
+    },
+    {
+      "epoch": 0.09294775446401732,
+      "grad_norm": 15.453202655575843,
+      "learning_rate": 3.9586491496013985e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 16314595,
+      "step": 773
+    },
+    {
+      "epoch": 0.0930679973546564,
+      "grad_norm": 2.940489711871276,
+      "learning_rate": 3.958491418761124e-06,
+      "loss": 0.8164,
+      "num_input_tokens_seen": 16331885,
+      "step": 774
+    },
+    {
+      "epoch": 0.0931882402452955,
+      "grad_norm": 3.8848968379515907,
+      "learning_rate": 3.958333390819258e-06,
+      "loss": 0.7305,
+      "num_input_tokens_seen": 16348535,
+      "step": 775
+    },
+    {
+      "epoch": 0.0933084831359346,
+      "grad_norm": 2.3641481097103534,
+      "learning_rate": 3.9581750657997754e-06,
+      "loss": 0.8015,
+      "num_input_tokens_seen": 16367620,
+      "step": 776
+    },
+    {
+      "epoch": 0.09342872602657368,
+      "grad_norm": 1.845269461294319,
+      "learning_rate": 3.95801644372669e-06,
+      "loss": 0.8904,
+      "num_input_tokens_seen": 16387245,
+      "step": 777
+    },
+    {
+      "epoch": 0.09354896891721277,
+      "grad_norm": 2.3842978254680824,
+      "learning_rate": 3.957857524624068e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 16405845,
+      "step": 778
+    },
+    {
+      "epoch": 0.09366921180785186,
+      "grad_norm": 1.8456349204348133,
+      "learning_rate": 3.957698308516016e-06,
+      "loss": 0.8984,
+      "num_input_tokens_seen": 16426865,
+      "step": 779
+    },
+    {
+      "epoch": 0.09378945469849095,
+      "grad_norm": 2.1804303757190144,
+      "learning_rate": 3.957538795426688e-06,
+      "loss": 0.8203,
+      "num_input_tokens_seen": 16444010,
+      "step": 780
+    },
+    {
+      "epoch": 0.09390969758913004,
+      "grad_norm": 2.5682612779600635,
+      "learning_rate": 3.9573789853802804e-06,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 16462205,
+      "step": 781
+    },
+    {
+      "epoch": 0.09402994047976913,
+      "grad_norm": 3.000839422113299,
+      "learning_rate": 3.957218878401037e-06,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 16480415,
+      "step": 782
+    },
+    {
+      "epoch": 0.09415018337040823,
+      "grad_norm": 3.5719921084625597,
+      "learning_rate": 3.957058474513246e-06,
+      "loss": 0.8985,
+      "num_input_tokens_seen": 16499990,
+      "step": 783
+    },
+    {
+      "epoch": 0.09427042626104731,
+      "grad_norm": 1.9410620635859839,
+      "learning_rate": 3.956897773741241e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 16518700,
+      "step": 784
+    },
+    {
+      "epoch": 0.09439066915168641,
+      "grad_norm": 1.8008475285382617,
+      "learning_rate": 3.956736776109398e-06,
+      "loss": 0.717,
+      "num_input_tokens_seen": 16539595,
+      "step": 785
+    },
+    {
+      "epoch": 0.09451091204232549,
+      "grad_norm": 2.308861285673119,
+      "learning_rate": 3.956575481642143e-06,
+      "loss": 0.8331,
+      "num_input_tokens_seen": 16558205,
+      "step": 786
+    },
+    {
+      "epoch": 0.09463115493296459,
+      "grad_norm": 4.338081045427215,
+      "learning_rate": 3.956413890363943e-06,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 16574905,
+      "step": 787
+    },
+    {
+      "epoch": 0.09475139782360369,
+      "grad_norm": 2.424918293313003,
+      "learning_rate": 3.956252002299312e-06,
+      "loss": 0.8215,
+      "num_input_tokens_seen": 16590525,
+      "step": 788
+    },
+    {
+      "epoch": 0.09487164071424277,
+      "grad_norm": 2.3242193035105796,
+      "learning_rate": 3.956089817472807e-06,
+      "loss": 0.904,
+      "num_input_tokens_seen": 16607550,
+      "step": 789
+    },
+    {
+      "epoch": 0.09499188360488187,
+      "grad_norm": 2.6751752028985742,
+      "learning_rate": 3.955927335909032e-06,
+      "loss": 0.8571,
+      "num_input_tokens_seen": 16630480,
+      "step": 790
+    },
+    {
+      "epoch": 0.09511212649552095,
+      "grad_norm": 3.5030845084728113,
+      "learning_rate": 3.955764557632634e-06,
+      "loss": 0.76,
+      "num_input_tokens_seen": 16650010,
+      "step": 791
+    },
+    {
+      "epoch": 0.09523236938616005,
+      "grad_norm": 2.6085712476066827,
+      "learning_rate": 3.955601482668309e-06,
+      "loss": 0.9435,
+      "num_input_tokens_seen": 16667590,
+      "step": 792
+    },
+    {
+      "epoch": 0.09535261227679913,
+      "grad_norm": 2.196506061146064,
+      "learning_rate": 3.955438111040794e-06,
+      "loss": 0.8821,
+      "num_input_tokens_seen": 16685585,
+      "step": 793
+    },
+    {
+      "epoch": 0.09547285516743823,
+      "grad_norm": 2.1789414118550616,
+      "learning_rate": 3.955274442774873e-06,
+      "loss": 0.8053,
+      "num_input_tokens_seen": 16703885,
+      "step": 794
+    },
+    {
+      "epoch": 0.09559309805807732,
+      "grad_norm": 2.9322161737975017,
+      "learning_rate": 3.9551104778953725e-06,
+      "loss": 0.7192,
+      "num_input_tokens_seen": 16723900,
+      "step": 795
+    },
+    {
+      "epoch": 0.0957133409487164,
+      "grad_norm": 2.3579054752705932,
+      "learning_rate": 3.954946216427167e-06,
+      "loss": 0.8552,
+      "num_input_tokens_seen": 16744080,
+      "step": 796
+    },
+    {
+      "epoch": 0.0958335838393555,
+      "grad_norm": 0.848150156353711,
+      "learning_rate": 3.954781658395176e-06,
+      "loss": 0.6461,
+      "num_input_tokens_seen": 16800055,
+      "step": 797
+    },
+    {
+      "epoch": 0.09595382672999458,
+      "grad_norm": 2.065885571928616,
+      "learning_rate": 3.95461680382436e-06,
+      "loss": 0.9196,
+      "num_input_tokens_seen": 16818700,
+      "step": 798
+    },
+    {
+      "epoch": 0.09607406962063368,
+      "grad_norm": 2.774878814838025,
+      "learning_rate": 3.9544516527397295e-06,
+      "loss": 0.8571,
+      "num_input_tokens_seen": 16834770,
+      "step": 799
+    },
+    {
+      "epoch": 0.09619431251127276,
+      "grad_norm": 3.5550330091887403,
+      "learning_rate": 3.954286205166338e-06,
+      "loss": 0.8034,
+      "num_input_tokens_seen": 16855655,
+      "step": 800
+    },
+    {
+      "epoch": 0.09631455540191186,
+      "grad_norm": 2.7915733861373297,
+      "learning_rate": 3.954120461129282e-06,
+      "loss": 0.8364,
+      "num_input_tokens_seen": 16872785,
+      "step": 801
+    },
+    {
+      "epoch": 0.09643479829255096,
+      "grad_norm": 2.0158892148812684,
+      "learning_rate": 3.953954420653706e-06,
+      "loss": 0.8388,
+      "num_input_tokens_seen": 16889530,
+      "step": 802
+    },
+    {
+      "epoch": 0.09655504118319004,
+      "grad_norm": 1.8801797237659632,
+      "learning_rate": 3.953788083764798e-06,
+      "loss": 0.8761,
+      "num_input_tokens_seen": 16908485,
+      "step": 803
+    },
+    {
+      "epoch": 0.09667528407382914,
+      "grad_norm": 2.2157862217841924,
+      "learning_rate": 3.953621450487792e-06,
+      "loss": 0.9219,
+      "num_input_tokens_seen": 16926825,
+      "step": 804
+    },
+    {
+      "epoch": 0.09679552696446822,
+      "grad_norm": 0.8805269388389226,
+      "learning_rate": 3.953454520847964e-06,
+      "loss": 0.6507,
+      "num_input_tokens_seen": 16991390,
+      "step": 805
+    },
+    {
+      "epoch": 0.09691576985510732,
+      "grad_norm": 2.1653822360162933,
+      "learning_rate": 3.9532872948706395e-06,
+      "loss": 0.7341,
+      "num_input_tokens_seen": 17010605,
+      "step": 806
+    },
+    {
+      "epoch": 0.09703601274574641,
+      "grad_norm": 3.455326125989557,
+      "learning_rate": 3.9531197725811845e-06,
+      "loss": 0.8278,
+      "num_input_tokens_seen": 17025710,
+      "step": 807
+    },
+    {
+      "epoch": 0.0971562556363855,
+      "grad_norm": 1.995702753388604,
+      "learning_rate": 3.952951954005013e-06,
+      "loss": 0.8692,
+      "num_input_tokens_seen": 17045115,
+      "step": 808
+    },
+    {
+      "epoch": 0.0972764985270246,
+      "grad_norm": 1.79485665333028,
+      "learning_rate": 3.952783839167584e-06,
+      "loss": 0.8502,
+      "num_input_tokens_seen": 17064880,
+      "step": 809
+    },
+    {
+      "epoch": 0.09739674141766368,
+      "grad_norm": 4.291058457118844,
+      "learning_rate": 3.952615428094398e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 17084120,
+      "step": 810
+    },
+    {
+      "epoch": 0.09751698430830277,
+      "grad_norm": 1.9438052169624325,
+      "learning_rate": 3.952446720811004e-06,
+      "loss": 0.7379,
+      "num_input_tokens_seen": 17102165,
+      "step": 811
+    },
+    {
+      "epoch": 0.09763722719894186,
+      "grad_norm": 0.8610778902561823,
+      "learning_rate": 3.952277717342995e-06,
+      "loss": 0.6713,
+      "num_input_tokens_seen": 17168320,
+      "step": 812
+    },
+    {
+      "epoch": 0.09775747008958095,
+      "grad_norm": 6.33912804714025,
+      "learning_rate": 3.952108417716009e-06,
+      "loss": 0.8563,
+      "num_input_tokens_seen": 17187495,
+      "step": 813
+    },
+    {
+      "epoch": 0.09787771298022005,
+      "grad_norm": 2.1429295127776364,
+      "learning_rate": 3.951938821955727e-06,
+      "loss": 0.8494,
+      "num_input_tokens_seen": 17206615,
+      "step": 814
+    },
+    {
+      "epoch": 0.09799795587085913,
+      "grad_norm": 1.8863557679900163,
+      "learning_rate": 3.9517689300878786e-06,
+      "loss": 0.7577,
+      "num_input_tokens_seen": 17226070,
+      "step": 815
+    },
+    {
+      "epoch": 0.09811819876149823,
+      "grad_norm": 2.4820682398711233,
+      "learning_rate": 3.951598742138236e-06,
+      "loss": 0.7807,
+      "num_input_tokens_seen": 17244515,
+      "step": 816
+    },
+    {
+      "epoch": 0.09823844165213731,
+      "grad_norm": 3.2675054957132144,
+      "learning_rate": 3.951428258132615e-06,
+      "loss": 0.7863,
+      "num_input_tokens_seen": 17262355,
+      "step": 817
+    },
+    {
+      "epoch": 0.09835868454277641,
+      "grad_norm": 2.1469827370775434,
+      "learning_rate": 3.951257478096879e-06,
+      "loss": 0.8413,
+      "num_input_tokens_seen": 17280440,
+      "step": 818
+    },
+    {
+      "epoch": 0.0984789274334155,
+      "grad_norm": 2.9853358604500015,
+      "learning_rate": 3.951086402056936e-06,
+      "loss": 0.6787,
+      "num_input_tokens_seen": 17294760,
+      "step": 819
+    },
+    {
+      "epoch": 0.09859917032405459,
+      "grad_norm": 1.8856098284383407,
+      "learning_rate": 3.950915030038735e-06,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 17314275,
+      "step": 820
+    },
+    {
+      "epoch": 0.09871941321469369,
+      "grad_norm": 2.958071816735845,
+      "learning_rate": 3.9507433620682765e-06,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 17330930,
+      "step": 821
+    },
+    {
+      "epoch": 0.09883965610533277,
+      "grad_norm": 2.2258653599664338,
+      "learning_rate": 3.9505713981716e-06,
+      "loss": 0.8725,
+      "num_input_tokens_seen": 17353480,
+      "step": 822
+    },
+    {
+      "epoch": 0.09895989899597187,
+      "grad_norm": 1.8557577210763645,
+      "learning_rate": 3.950399138374795e-06,
+      "loss": 0.8091,
+      "num_input_tokens_seen": 17372280,
+      "step": 823
+    },
+    {
+      "epoch": 0.09908014188661095,
+      "grad_norm": 2.1965356814223607,
+      "learning_rate": 3.95022658270399e-06,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 17392365,
+      "step": 824
+    },
+    {
+      "epoch": 0.09920038477725004,
+      "grad_norm": 2.6909562335043167,
+      "learning_rate": 3.9500537311853635e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 17410040,
+      "step": 825
+    },
+    {
+      "epoch": 0.09932062766788914,
+      "grad_norm": 2.850219724681345,
+      "learning_rate": 3.949880583845136e-06,
+      "loss": 0.8289,
+      "num_input_tokens_seen": 17427835,
+      "step": 826
+    },
+    {
+      "epoch": 0.09944087055852822,
+      "grad_norm": 2.4005268900928414,
+      "learning_rate": 3.949707140709575e-06,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 17447285,
+      "step": 827
+    },
+    {
+      "epoch": 0.09956111344916732,
+      "grad_norm": 2.3337804237662905,
+      "learning_rate": 3.949533401804991e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 17463910,
+      "step": 828
+    },
+    {
+      "epoch": 0.0996813563398064,
+      "grad_norm": 2.3990471959789197,
+      "learning_rate": 3.949359367157739e-06,
+      "loss": 0.9032,
+      "num_input_tokens_seen": 17482325,
+      "step": 829
+    },
+    {
+      "epoch": 0.0998015992304455,
+      "grad_norm": 2.1356542482128553,
+      "learning_rate": 3.949185036794222e-06,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 17500055,
+      "step": 830
+    },
+    {
+      "epoch": 0.0999218421210846,
+      "grad_norm": 1.6977171190632112,
+      "learning_rate": 3.949010410740884e-06,
+      "loss": 0.7766,
+      "num_input_tokens_seen": 17522600,
+      "step": 831
+    },
+    {
+      "epoch": 0.10004208501172368,
+      "grad_norm": 1.8981337506416434,
+      "learning_rate": 3.948835489024216e-06,
+      "loss": 0.8663,
+      "num_input_tokens_seen": 17542055,
+      "step": 832
+    },
+    {
+      "epoch": 0.10016232790236278,
+      "grad_norm": 1.9123374217461606,
+      "learning_rate": 3.948660271670755e-06,
+      "loss": 0.8918,
+      "num_input_tokens_seen": 17558925,
+      "step": 833
+    },
+    {
+      "epoch": 0.10028257079300186,
+      "grad_norm": 2.2080318984779277,
+      "learning_rate": 3.948484758707079e-06,
+      "loss": 0.8442,
+      "num_input_tokens_seen": 17578245,
+      "step": 834
+    },
+    {
+      "epoch": 0.10040281368364096,
+      "grad_norm": 2.594852234622079,
+      "learning_rate": 3.948308950159815e-06,
+      "loss": 0.8347,
+      "num_input_tokens_seen": 17596645,
+      "step": 835
+    },
+    {
+      "epoch": 0.10052305657428004,
+      "grad_norm": 2.430344276047469,
+      "learning_rate": 3.9481328460556326e-06,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 17613585,
+      "step": 836
+    },
+    {
+      "epoch": 0.10064329946491914,
+      "grad_norm": 2.4197432529470997,
+      "learning_rate": 3.9479564464212455e-06,
+      "loss": 0.8894,
+      "num_input_tokens_seen": 17632465,
+      "step": 837
+    },
+    {
+      "epoch": 0.10076354235555823,
+      "grad_norm": 2.619392329816884,
+      "learning_rate": 3.947779751283414e-06,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 17649355,
+      "step": 838
+    },
+    {
+      "epoch": 0.10088378524619732,
+      "grad_norm": 1.9800604616141009,
+      "learning_rate": 3.947602760668944e-06,
+      "loss": 0.7534,
+      "num_input_tokens_seen": 17668865,
+      "step": 839
+    },
+    {
+      "epoch": 0.10100402813683641,
+      "grad_norm": 1.9335559704439675,
+      "learning_rate": 3.947425474604684e-06,
+      "loss": 0.7134,
+      "num_input_tokens_seen": 17692520,
+      "step": 840
+    },
+    {
+      "epoch": 0.1011242710274755,
+      "grad_norm": 2.1978515842196957,
+      "learning_rate": 3.947247893117528e-06,
+      "loss": 0.9242,
+      "num_input_tokens_seen": 17710745,
+      "step": 841
+    },
+    {
+      "epoch": 0.10124451391811459,
+      "grad_norm": 7.388582666770699,
+      "learning_rate": 3.947070016234413e-06,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 17726255,
+      "step": 842
+    },
+    {
+      "epoch": 0.10136475680875369,
+      "grad_norm": 2.6642207060516934,
+      "learning_rate": 3.946891843982326e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 17743640,
+      "step": 843
+    },
+    {
+      "epoch": 0.10148499969939277,
+      "grad_norm": 15.751139988427198,
+      "learning_rate": 3.9467133763882935e-06,
+      "loss": 0.742,
+      "num_input_tokens_seen": 17761825,
+      "step": 844
+    },
+    {
+      "epoch": 0.10160524259003187,
+      "grad_norm": 1.954345176234008,
+      "learning_rate": 3.9465346134793905e-06,
+      "loss": 0.8621,
+      "num_input_tokens_seen": 17781355,
+      "step": 845
+    },
+    {
+      "epoch": 0.10172548548067095,
+      "grad_norm": 2.024418013849752,
+      "learning_rate": 3.9463555552827335e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 17798245,
+      "step": 846
+    },
+    {
+      "epoch": 0.10184572837131005,
+      "grad_norm": 2.6641568638193243,
+      "learning_rate": 3.946176201825487e-06,
+      "loss": 0.8667,
+      "num_input_tokens_seen": 17816000,
+      "step": 847
+    },
+    {
+      "epoch": 0.10196597126194913,
+      "grad_norm": 2.868935123915064,
+      "learning_rate": 3.9459965531348575e-06,
+      "loss": 0.838,
+      "num_input_tokens_seen": 17835375,
+      "step": 848
+    },
+    {
+      "epoch": 0.10208621415258823,
+      "grad_norm": 2.2581213699127827,
+      "learning_rate": 3.945816609238098e-06,
+      "loss": 0.8584,
+      "num_input_tokens_seen": 17854505,
+      "step": 849
+    },
+    {
+      "epoch": 0.10220645704322733,
+      "grad_norm": 1.8912363106877021,
+      "learning_rate": 3.945636370162507e-06,
+      "loss": 0.8493,
+      "num_input_tokens_seen": 17874335,
+      "step": 850
+    },
+    {
+      "epoch": 0.10232669993386641,
+      "grad_norm": 1.7434850982417882,
+      "learning_rate": 3.945455835935425e-06,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 17893240,
+      "step": 851
+    },
+    {
+      "epoch": 0.1024469428245055,
+      "grad_norm": 2.830988142985482,
+      "learning_rate": 3.94527500658424e-06,
+      "loss": 0.7414,
+      "num_input_tokens_seen": 17910625,
+      "step": 852
+    },
+    {
+      "epoch": 0.10256718571514459,
+      "grad_norm": 1.974619927085637,
+      "learning_rate": 3.945093882136382e-06,
+      "loss": 0.8126,
+      "num_input_tokens_seen": 17934120,
+      "step": 853
+    },
+    {
+      "epoch": 0.10268742860578368,
+      "grad_norm": 1.9756341736380256,
+      "learning_rate": 3.944912462619329e-06,
+      "loss": 0.8376,
+      "num_input_tokens_seen": 17952805,
+      "step": 854
+    },
+    {
+      "epoch": 0.10280767149642277,
+      "grad_norm": 2.1421202171865175,
+      "learning_rate": 3.9447307480606025e-06,
+      "loss": 0.809,
+      "num_input_tokens_seen": 17972610,
+      "step": 855
+    },
+    {
+      "epoch": 0.10292791438706186,
+      "grad_norm": 2.349347297268663,
+      "learning_rate": 3.944548738487767e-06,
+      "loss": 0.8997,
+      "num_input_tokens_seen": 17989845,
+      "step": 856
+    },
+    {
+      "epoch": 0.10304815727770096,
+      "grad_norm": 2.669528222583635,
+      "learning_rate": 3.944366433928434e-06,
+      "loss": 0.8976,
+      "num_input_tokens_seen": 18009545,
+      "step": 857
+    },
+    {
+      "epoch": 0.10316840016834004,
+      "grad_norm": 1.7060248167467518,
+      "learning_rate": 3.9441838344102594e-06,
+      "loss": 0.8274,
+      "num_input_tokens_seen": 18028990,
+      "step": 858
+    },
+    {
+      "epoch": 0.10328864305897914,
+      "grad_norm": 2.404724311856886,
+      "learning_rate": 3.944000939960943e-06,
+      "loss": 0.6744,
+      "num_input_tokens_seen": 18047435,
+      "step": 859
+    },
+    {
+      "epoch": 0.10340888594961822,
+      "grad_norm": 1.6697933656545882,
+      "learning_rate": 3.943817750608229e-06,
+      "loss": 0.8018,
+      "num_input_tokens_seen": 18069705,
+      "step": 860
+    },
+    {
+      "epoch": 0.10352912884025732,
+      "grad_norm": 2.7203473805398,
+      "learning_rate": 3.943634266379908e-06,
+      "loss": 0.822,
+      "num_input_tokens_seen": 18086320,
+      "step": 861
+    },
+    {
+      "epoch": 0.10364937173089642,
+      "grad_norm": 3.164258252910918,
+      "learning_rate": 3.943450487303815e-06,
+      "loss": 0.8487,
+      "num_input_tokens_seen": 18106535,
+      "step": 862
+    },
+    {
+      "epoch": 0.1037696146215355,
+      "grad_norm": 2.4741336976674484,
+      "learning_rate": 3.943266413407827e-06,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 18125530,
+      "step": 863
+    },
+    {
+      "epoch": 0.1038898575121746,
+      "grad_norm": 1.752134948496544,
+      "learning_rate": 3.94308204471987e-06,
+      "loss": 0.8451,
+      "num_input_tokens_seen": 18144265,
+      "step": 864
+    },
+    {
+      "epoch": 0.10401010040281368,
+      "grad_norm": 10.398919262719307,
+      "learning_rate": 3.942897381267912e-06,
+      "loss": 0.7419,
+      "num_input_tokens_seen": 18160350,
+      "step": 865
+    },
+    {
+      "epoch": 0.10413034329345278,
+      "grad_norm": 2.7296546889668636,
+      "learning_rate": 3.942712423079965e-06,
+      "loss": 0.6748,
+      "num_input_tokens_seen": 18176460,
+      "step": 866
+    },
+    {
+      "epoch": 0.10425058618409186,
+      "grad_norm": 2.2475341257305628,
+      "learning_rate": 3.942527170184088e-06,
+      "loss": 0.8979,
+      "num_input_tokens_seen": 18192800,
+      "step": 867
+    },
+    {
+      "epoch": 0.10437082907473096,
+      "grad_norm": 5.726105634050731,
+      "learning_rate": 3.942341622608385e-06,
+      "loss": 0.7819,
+      "num_input_tokens_seen": 18209550,
+      "step": 868
+    },
+    {
+      "epoch": 0.10449107196537005,
+      "grad_norm": 1.5963926021940098,
+      "learning_rate": 3.942155780381001e-06,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 18233005,
+      "step": 869
+    },
+    {
+      "epoch": 0.10461131485600914,
+      "grad_norm": 1.914487223740198,
+      "learning_rate": 3.94196964353013e-06,
+      "loss": 0.7613,
+      "num_input_tokens_seen": 18252175,
+      "step": 870
+    },
+    {
+      "epoch": 0.10473155774664823,
+      "grad_norm": 2.1994763345851345,
+      "learning_rate": 3.941783212084008e-06,
+      "loss": 0.8013,
+      "num_input_tokens_seen": 18269650,
+      "step": 871
+    },
+    {
+      "epoch": 0.10485180063728732,
+      "grad_norm": 4.017514211352629,
+      "learning_rate": 3.941596486070916e-06,
+      "loss": 0.775,
+      "num_input_tokens_seen": 18287415,
+      "step": 872
+    },
+    {
+      "epoch": 0.10497204352792641,
+      "grad_norm": 3.2320445141072556,
+      "learning_rate": 3.941409465519182e-06,
+      "loss": 0.5768,
+      "num_input_tokens_seen": 18307660,
+      "step": 873
+    },
+    {
+      "epoch": 0.10509228641856551,
+      "grad_norm": 1.7936104972660718,
+      "learning_rate": 3.941222150457176e-06,
+      "loss": 0.846,
+      "num_input_tokens_seen": 18330635,
+      "step": 874
+    },
+    {
+      "epoch": 0.10521252930920459,
+      "grad_norm": 3.2910709671512812,
+      "learning_rate": 3.941034540913311e-06,
+      "loss": 0.7261,
+      "num_input_tokens_seen": 18347885,
+      "step": 875
+    },
+    {
+      "epoch": 0.10533277219984369,
+      "grad_norm": 1.9848623920406179,
+      "learning_rate": 3.940846636916051e-06,
+      "loss": 0.8232,
+      "num_input_tokens_seen": 18367640,
+      "step": 876
+    },
+    {
+      "epoch": 0.10545301509048277,
+      "grad_norm": 2.407521735295685,
+      "learning_rate": 3.940658438493899e-06,
+      "loss": 0.8626,
+      "num_input_tokens_seen": 18385205,
+      "step": 877
+    },
+    {
+      "epoch": 0.10557325798112187,
+      "grad_norm": 4.9600778152505205,
+      "learning_rate": 3.940469945675405e-06,
+      "loss": 0.7583,
+      "num_input_tokens_seen": 18403310,
+      "step": 878
+    },
+    {
+      "epoch": 0.10569350087176095,
+      "grad_norm": 1.910106903459763,
+      "learning_rate": 3.940281158489163e-06,
+      "loss": 0.9092,
+      "num_input_tokens_seen": 18422260,
+      "step": 879
+    },
+    {
+      "epoch": 0.10581374376240005,
+      "grad_norm": 1.7108246591914542,
+      "learning_rate": 3.940092076963812e-06,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 18439475,
+      "step": 880
+    },
+    {
+      "epoch": 0.10593398665303914,
+      "grad_norm": 5.191199562171951,
+      "learning_rate": 3.9399027011280355e-06,
+      "loss": 0.7905,
+      "num_input_tokens_seen": 18461290,
+      "step": 881
+    },
+    {
+      "epoch": 0.10605422954367823,
+      "grad_norm": 2.4991051124957524,
+      "learning_rate": 3.939713031010561e-06,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 18479375,
+      "step": 882
+    },
+    {
+      "epoch": 0.10617447243431732,
+      "grad_norm": 2.8233917929338457,
+      "learning_rate": 3.939523066640163e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 18497990,
+      "step": 883
+    },
+    {
+      "epoch": 0.10629471532495641,
+      "grad_norm": 2.2332503412616016,
+      "learning_rate": 3.939332808045657e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 18517360,
+      "step": 884
+    },
+    {
+      "epoch": 0.1064149582155955,
+      "grad_norm": 2.545677704667773,
+      "learning_rate": 3.939142255255906e-06,
+      "loss": 0.8394,
+      "num_input_tokens_seen": 18537965,
+      "step": 885
+    },
+    {
+      "epoch": 0.1065352011062346,
+      "grad_norm": 5.0264005574938055,
+      "learning_rate": 3.938951408299817e-06,
+      "loss": 0.874,
+      "num_input_tokens_seen": 18556525,
+      "step": 886
+    },
+    {
+      "epoch": 0.10665544399687368,
+      "grad_norm": 0.853644560270433,
+      "learning_rate": 3.938760267206342e-06,
+      "loss": 0.577,
+      "num_input_tokens_seen": 18618065,
+      "step": 887
+    },
+    {
+      "epoch": 0.10677568688751278,
+      "grad_norm": 2.5750565493754607,
+      "learning_rate": 3.938568832004475e-06,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 18636490,
+      "step": 888
+    },
+    {
+      "epoch": 0.10689592977815186,
+      "grad_norm": 2.2761256889057613,
+      "learning_rate": 3.938377102723257e-06,
+      "loss": 0.7513,
+      "num_input_tokens_seen": 18653345,
+      "step": 889
+    },
+    {
+      "epoch": 0.10701617266879096,
+      "grad_norm": 4.025603558156583,
+      "learning_rate": 3.938185079391774e-06,
+      "loss": 0.8382,
+      "num_input_tokens_seen": 18670110,
+      "step": 890
+    },
+    {
+      "epoch": 0.10713641555943004,
+      "grad_norm": 3.6265857661300474,
+      "learning_rate": 3.937992762039157e-06,
+      "loss": 1.0576,
+      "num_input_tokens_seen": 18683155,
+      "step": 891
+    },
+    {
+      "epoch": 0.10725665845006914,
+      "grad_norm": 1.8686906155825185,
+      "learning_rate": 3.937800150694577e-06,
+      "loss": 0.7968,
+      "num_input_tokens_seen": 18704050,
+      "step": 892
+    },
+    {
+      "epoch": 0.10737690134070824,
+      "grad_norm": 2.2303912319943313,
+      "learning_rate": 3.937607245387255e-06,
+      "loss": 0.7513,
+      "num_input_tokens_seen": 18723135,
+      "step": 893
+    },
+    {
+      "epoch": 0.10749714423134732,
+      "grad_norm": 2.220753653754816,
+      "learning_rate": 3.937414046146455e-06,
+      "loss": 0.7252,
+      "num_input_tokens_seen": 18740810,
+      "step": 894
+    },
+    {
+      "epoch": 0.10761738712198642,
+      "grad_norm": 1.9937587790455347,
+      "learning_rate": 3.9372205530014845e-06,
+      "loss": 0.7551,
+      "num_input_tokens_seen": 18759010,
+      "step": 895
+    },
+    {
+      "epoch": 0.1077376300126255,
+      "grad_norm": 2.1924064618632255,
+      "learning_rate": 3.937026765981696e-06,
+      "loss": 0.7207,
+      "num_input_tokens_seen": 18778800,
+      "step": 896
+    },
+    {
+      "epoch": 0.1078578729032646,
+      "grad_norm": 5.057603383128183,
+      "learning_rate": 3.936832685116488e-06,
+      "loss": 0.7882,
+      "num_input_tokens_seen": 18796615,
+      "step": 897
+    },
+    {
+      "epoch": 0.10797811579390369,
+      "grad_norm": 2.267029215323489,
+      "learning_rate": 3.936638310435301e-06,
+      "loss": 0.8897,
+      "num_input_tokens_seen": 18814200,
+      "step": 898
+    },
+    {
+      "epoch": 0.10809835868454278,
+      "grad_norm": 2.525104824510146,
+      "learning_rate": 3.936443641967623e-06,
+      "loss": 0.8212,
+      "num_input_tokens_seen": 18832750,
+      "step": 899
+    },
+    {
+      "epoch": 0.10821860157518187,
+      "grad_norm": 2.827751131493035,
+      "learning_rate": 3.936248679742983e-06,
+      "loss": 0.8215,
+      "num_input_tokens_seen": 18850965,
+      "step": 900
+    },
+    {
+      "epoch": 0.10833884446582095,
+      "grad_norm": 1.0736845773898236,
+      "learning_rate": 3.936053423790958e-06,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 18899005,
+      "step": 901
+    },
+    {
+      "epoch": 0.10845908735646005,
+      "grad_norm": 3.4983442194194265,
+      "learning_rate": 3.935857874141168e-06,
+      "loss": 0.769,
+      "num_input_tokens_seen": 18917560,
+      "step": 902
+    },
+    {
+      "epoch": 0.10857933024709913,
+      "grad_norm": 2.4660467962538304,
+      "learning_rate": 3.935662030823279e-06,
+      "loss": 0.8318,
+      "num_input_tokens_seen": 18933465,
+      "step": 903
+    },
+    {
+      "epoch": 0.10869957313773823,
+      "grad_norm": 3.905884307399311,
+      "learning_rate": 3.935465893866998e-06,
+      "loss": 0.7177,
+      "num_input_tokens_seen": 18951410,
+      "step": 904
+    },
+    {
+      "epoch": 0.10881981602837733,
+      "grad_norm": 4.987969152361887,
+      "learning_rate": 3.935269463302079e-06,
+      "loss": 0.8016,
+      "num_input_tokens_seen": 18969335,
+      "step": 905
+    },
+    {
+      "epoch": 0.10894005891901641,
+      "grad_norm": 2.9447559752636643,
+      "learning_rate": 3.935072739158322e-06,
+      "loss": 0.7721,
+      "num_input_tokens_seen": 18988765,
+      "step": 906
+    },
+    {
+      "epoch": 0.10906030180965551,
+      "grad_norm": 2.216283978262971,
+      "learning_rate": 3.934875721465569e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 19008905,
+      "step": 907
+    },
+    {
+      "epoch": 0.10918054470029459,
+      "grad_norm": 2.8980127532769284,
+      "learning_rate": 3.9346784102537076e-06,
+      "loss": 0.7044,
+      "num_input_tokens_seen": 19030760,
+      "step": 908
+    },
+    {
+      "epoch": 0.10930078759093369,
+      "grad_norm": 2.0406317290886573,
+      "learning_rate": 3.934480805552669e-06,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 19051490,
+      "step": 909
+    },
+    {
+      "epoch": 0.10942103048157277,
+      "grad_norm": 2.9997279204440845,
+      "learning_rate": 3.93428290739243e-06,
+      "loss": 0.8799,
+      "num_input_tokens_seen": 19070580,
+      "step": 910
+    },
+    {
+      "epoch": 0.10954127337221187,
+      "grad_norm": 3.1081643562199774,
+      "learning_rate": 3.9340847158030125e-06,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 19083880,
+      "step": 911
+    },
+    {
+      "epoch": 0.10966151626285096,
+      "grad_norm": 2.1473590426929823,
+      "learning_rate": 3.9338862308144814e-06,
+      "loss": 0.7521,
+      "num_input_tokens_seen": 19102420,
+      "step": 912
+    },
+    {
+      "epoch": 0.10978175915349005,
+      "grad_norm": 2.0226160343646176,
+      "learning_rate": 3.933687452456946e-06,
+      "loss": 0.8446,
+      "num_input_tokens_seen": 19122040,
+      "step": 913
+    },
+    {
+      "epoch": 0.10990200204412914,
+      "grad_norm": 3.1411197451802195,
+      "learning_rate": 3.933488380760563e-06,
+      "loss": 0.8606,
+      "num_input_tokens_seen": 19141120,
+      "step": 914
+    },
+    {
+      "epoch": 0.11002224493476823,
+      "grad_norm": 3.052820266998868,
+      "learning_rate": 3.9332890157555286e-06,
+      "loss": 0.873,
+      "num_input_tokens_seen": 19157775,
+      "step": 915
+    },
+    {
+      "epoch": 0.11014248782540732,
+      "grad_norm": 2.3795171308120757,
+      "learning_rate": 3.933089357472088e-06,
+      "loss": 0.768,
+      "num_input_tokens_seen": 19175525,
+      "step": 916
+    },
+    {
+      "epoch": 0.11026273071604642,
+      "grad_norm": 2.0161641117628393,
+      "learning_rate": 3.932889405940529e-06,
+      "loss": 0.8564,
+      "num_input_tokens_seen": 19193340,
+      "step": 917
+    },
+    {
+      "epoch": 0.1103829736066855,
+      "grad_norm": 2.794965972704058,
+      "learning_rate": 3.932689161191184e-06,
+      "loss": 0.8012,
+      "num_input_tokens_seen": 19210765,
+      "step": 918
+    },
+    {
+      "epoch": 0.1105032164973246,
+      "grad_norm": 6.322072739647685,
+      "learning_rate": 3.93248862325443e-06,
+      "loss": 0.8685,
+      "num_input_tokens_seen": 19229390,
+      "step": 919
+    },
+    {
+      "epoch": 0.11062345938796368,
+      "grad_norm": 1.0306669843172556,
+      "learning_rate": 3.932287792160688e-06,
+      "loss": 0.6544,
+      "num_input_tokens_seen": 19287570,
+      "step": 920
+    },
+    {
+      "epoch": 0.11074370227860278,
+      "grad_norm": 7.896774559115421,
+      "learning_rate": 3.932086667940424e-06,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 19303995,
+      "step": 921
+    },
+    {
+      "epoch": 0.11086394516924186,
+      "grad_norm": 2.0862080976663204,
+      "learning_rate": 3.93188525062415e-06,
+      "loss": 0.8101,
+      "num_input_tokens_seen": 19324180,
+      "step": 922
+    },
+    {
+      "epoch": 0.11098418805988096,
+      "grad_norm": 2.89765987559848,
+      "learning_rate": 3.931683540242418e-06,
+      "loss": 0.8573,
+      "num_input_tokens_seen": 19344965,
+      "step": 923
+    },
+    {
+      "epoch": 0.11110443095052006,
+      "grad_norm": 5.435721533856892,
+      "learning_rate": 3.9314815368258295e-06,
+      "loss": 0.9106,
+      "num_input_tokens_seen": 19361165,
+      "step": 924
+    },
+    {
+      "epoch": 0.11122467384115914,
+      "grad_norm": 1.7220035487791467,
+      "learning_rate": 3.9312792404050275e-06,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 19378940,
+      "step": 925
+    },
+    {
+      "epoch": 0.11134491673179824,
+      "grad_norm": 1.8868908922111025,
+      "learning_rate": 3.9310766510107e-06,
+      "loss": 0.7763,
+      "num_input_tokens_seen": 19397835,
+      "step": 926
+    },
+    {
+      "epoch": 0.11146515962243732,
+      "grad_norm": 2.324027861265247,
+      "learning_rate": 3.9308737686735806e-06,
+      "loss": 0.9125,
+      "num_input_tokens_seen": 19417515,
+      "step": 927
+    },
+    {
+      "epoch": 0.11158540251307641,
+      "grad_norm": 2.2059056940865562,
+      "learning_rate": 3.9306705934244455e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 19437315,
+      "step": 928
+    },
+    {
+      "epoch": 0.11170564540371551,
+      "grad_norm": 1.9078024371301123,
+      "learning_rate": 3.930467125294116e-06,
+      "loss": 0.8771,
+      "num_input_tokens_seen": 19456585,
+      "step": 929
+    },
+    {
+      "epoch": 0.1118258882943546,
+      "grad_norm": 1.0558164853706429,
+      "learning_rate": 3.930263364313458e-06,
+      "loss": 0.6383,
+      "num_input_tokens_seen": 19506875,
+      "step": 930
+    },
+    {
+      "epoch": 0.11194613118499369,
+      "grad_norm": 2.621749334394303,
+      "learning_rate": 3.930059310513384e-06,
+      "loss": 0.8334,
+      "num_input_tokens_seen": 19525635,
+      "step": 931
+    },
+    {
+      "epoch": 0.11206637407563277,
+      "grad_norm": 1.99741159031062,
+      "learning_rate": 3.929854963924846e-06,
+      "loss": 0.8276,
+      "num_input_tokens_seen": 19545620,
+      "step": 932
+    },
+    {
+      "epoch": 0.11218661696627187,
+      "grad_norm": 1.888590085152523,
+      "learning_rate": 3.929650324578845e-06,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 19564805,
+      "step": 933
+    },
+    {
+      "epoch": 0.11230685985691095,
+      "grad_norm": 3.0609832268736654,
+      "learning_rate": 3.929445392506423e-06,
+      "loss": 0.8238,
+      "num_input_tokens_seen": 19582465,
+      "step": 934
+    },
+    {
+      "epoch": 0.11242710274755005,
+      "grad_norm": 2.151583165283819,
+      "learning_rate": 3.92924016773867e-06,
+      "loss": 0.7609,
+      "num_input_tokens_seen": 19598680,
+      "step": 935
+    },
+    {
+      "epoch": 0.11254734563818915,
+      "grad_norm": 2.765652973327632,
+      "learning_rate": 3.9290346503067175e-06,
+      "loss": 0.7417,
+      "num_input_tokens_seen": 19615065,
+      "step": 936
+    },
+    {
+      "epoch": 0.11266758852882823,
+      "grad_norm": 7.650724153254989,
+      "learning_rate": 3.9288288402417415e-06,
+      "loss": 0.7871,
+      "num_input_tokens_seen": 19641045,
+      "step": 937
+    },
+    {
+      "epoch": 0.11278783141946733,
+      "grad_norm": 3.689071534542488,
+      "learning_rate": 3.928622737574964e-06,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 19656100,
+      "step": 938
+    },
+    {
+      "epoch": 0.11290807431010641,
+      "grad_norm": 2.3187054336088995,
+      "learning_rate": 3.928416342337652e-06,
+      "loss": 0.9029,
+      "num_input_tokens_seen": 19675555,
+      "step": 939
+    },
+    {
+      "epoch": 0.1130283172007455,
+      "grad_norm": 4.7892279579408825,
+      "learning_rate": 3.928209654561113e-06,
+      "loss": 0.828,
+      "num_input_tokens_seen": 19696110,
+      "step": 940
+    },
+    {
+      "epoch": 0.1131485600913846,
+      "grad_norm": 2.093727068657475,
+      "learning_rate": 3.928002674276703e-06,
+      "loss": 0.8032,
+      "num_input_tokens_seen": 19715220,
+      "step": 941
+    },
+    {
+      "epoch": 0.11326880298202369,
+      "grad_norm": 2.7878164036963633,
+      "learning_rate": 3.92779540151582e-06,
+      "loss": 0.7591,
+      "num_input_tokens_seen": 19732025,
+      "step": 942
+    },
+    {
+      "epoch": 0.11338904587266278,
+      "grad_norm": 2.0445347263254963,
+      "learning_rate": 3.927587836309907e-06,
+      "loss": 0.8567,
+      "num_input_tokens_seen": 19749575,
+      "step": 943
+    },
+    {
+      "epoch": 0.11350928876330187,
+      "grad_norm": 2.599662481174319,
+      "learning_rate": 3.927379978690452e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 19768560,
+      "step": 944
+    },
+    {
+      "epoch": 0.11362953165394096,
+      "grad_norm": 2.667051665108858,
+      "learning_rate": 3.927171828688987e-06,
+      "loss": 0.869,
+      "num_input_tokens_seen": 19787805,
+      "step": 945
+    },
+    {
+      "epoch": 0.11374977454458005,
+      "grad_norm": 7.916041774460033,
+      "learning_rate": 3.926963386337088e-06,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 19805755,
+      "step": 946
+    },
+    {
+      "epoch": 0.11387001743521914,
+      "grad_norm": 2.2942675925445126,
+      "learning_rate": 3.926754651666375e-06,
+      "loss": 0.7011,
+      "num_input_tokens_seen": 19826035,
+      "step": 947
+    },
+    {
+      "epoch": 0.11399026032585824,
+      "grad_norm": 5.09330277816299,
+      "learning_rate": 3.926545624708513e-06,
+      "loss": 0.7891,
+      "num_input_tokens_seen": 19844995,
+      "step": 948
+    },
+    {
+      "epoch": 0.11411050321649732,
+      "grad_norm": 2.2372894025517556,
+      "learning_rate": 3.926336305495213e-06,
+      "loss": 0.8604,
+      "num_input_tokens_seen": 19863275,
+      "step": 949
+    },
+    {
+      "epoch": 0.11423074610713642,
+      "grad_norm": 2.2218400320876595,
+      "learning_rate": 3.926126694058226e-06,
+      "loss": 0.8908,
+      "num_input_tokens_seen": 19882145,
+      "step": 950
+    },
+    {
+      "epoch": 0.1143509889977755,
+      "grad_norm": 1.5071397877876678,
+      "learning_rate": 3.92591679042935e-06,
+      "loss": 0.8138,
+      "num_input_tokens_seen": 19901755,
+      "step": 951
+    },
+    {
+      "epoch": 0.1144712318884146,
+      "grad_norm": 1.6842076144983666,
+      "learning_rate": 3.925706594640429e-06,
+      "loss": 0.8203,
+      "num_input_tokens_seen": 19919535,
+      "step": 952
+    },
+    {
+      "epoch": 0.1145914747790537,
+      "grad_norm": 1.8638045607682039,
+      "learning_rate": 3.925496106723349e-06,
+      "loss": 0.7929,
+      "num_input_tokens_seen": 19936695,
+      "step": 953
+    },
+    {
+      "epoch": 0.11471171766969278,
+      "grad_norm": 2.3482208978510597,
+      "learning_rate": 3.9252853267100405e-06,
+      "loss": 0.8341,
+      "num_input_tokens_seen": 19955660,
+      "step": 954
+    },
+    {
+      "epoch": 0.11483196056033187,
+      "grad_norm": 1.8519764025922612,
+      "learning_rate": 3.9250742546324786e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 19975615,
+      "step": 955
+    },
+    {
+      "epoch": 0.11495220345097096,
+      "grad_norm": 2.1497665036345435,
+      "learning_rate": 3.924862890522683e-06,
+      "loss": 0.8615,
+      "num_input_tokens_seen": 19995345,
+      "step": 956
+    },
+    {
+      "epoch": 0.11507244634161005,
+      "grad_norm": 2.336480404613175,
+      "learning_rate": 3.9246512344127174e-06,
+      "loss": 0.8589,
+      "num_input_tokens_seen": 20012725,
+      "step": 957
+    },
+    {
+      "epoch": 0.11519268923224914,
+      "grad_norm": 2.126235469311675,
+      "learning_rate": 3.9244392863346895e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 20031850,
+      "step": 958
+    },
+    {
+      "epoch": 0.11531293212288823,
+      "grad_norm": 3.695723619848682,
+      "learning_rate": 3.9242270463207524e-06,
+      "loss": 0.9144,
+      "num_input_tokens_seen": 20049960,
+      "step": 959
+    },
+    {
+      "epoch": 0.11543317501352733,
+      "grad_norm": 3.419444421071364,
+      "learning_rate": 3.924014514403102e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 20065835,
+      "step": 960
+    },
+    {
+      "epoch": 0.11555341790416641,
+      "grad_norm": 3.286700961716625,
+      "learning_rate": 3.92380169061398e-06,
+      "loss": 0.9194,
+      "num_input_tokens_seen": 20083335,
+      "step": 961
+    },
+    {
+      "epoch": 0.11567366079480551,
+      "grad_norm": 2.292653429085566,
+      "learning_rate": 3.9235885749856705e-06,
+      "loss": 0.8381,
+      "num_input_tokens_seen": 20101735,
+      "step": 962
+    },
+    {
+      "epoch": 0.1157939036854446,
+      "grad_norm": 2.2215081064510747,
+      "learning_rate": 3.9233751675505035e-06,
+      "loss": 0.828,
+      "num_input_tokens_seen": 20120165,
+      "step": 963
+    },
+    {
+      "epoch": 0.11591414657608369,
+      "grad_norm": 3.520903868869938,
+      "learning_rate": 3.923161468340853e-06,
+      "loss": 0.8403,
+      "num_input_tokens_seen": 20139720,
+      "step": 964
+    },
+    {
+      "epoch": 0.11603438946672277,
+      "grad_norm": 2.77945729393548,
+      "learning_rate": 3.9229474773891374e-06,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 20157980,
+      "step": 965
+    },
+    {
+      "epoch": 0.11615463235736187,
+      "grad_norm": 2.1320400900549217,
+      "learning_rate": 3.922733194727818e-06,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 20177495,
+      "step": 966
+    },
+    {
+      "epoch": 0.11627487524800097,
+      "grad_norm": 2.6623691789189525,
+      "learning_rate": 3.922518620389402e-06,
+      "loss": 0.8605,
+      "num_input_tokens_seen": 20194080,
+      "step": 967
+    },
+    {
+      "epoch": 0.11639511813864005,
+      "grad_norm": 1.9312183690154034,
+      "learning_rate": 3.922303754406439e-06,
+      "loss": 0.9012,
+      "num_input_tokens_seen": 20211640,
+      "step": 968
+    },
+    {
+      "epoch": 0.11651536102927915,
+      "grad_norm": 1.9809686321716689,
+      "learning_rate": 3.922088596811526e-06,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 20230490,
+      "step": 969
+    },
+    {
+      "epoch": 0.11663560391991823,
+      "grad_norm": 2.479195251601029,
+      "learning_rate": 3.9218731476373e-06,
+      "loss": 0.8561,
+      "num_input_tokens_seen": 20246395,
+      "step": 970
+    },
+    {
+      "epoch": 0.11675584681055733,
+      "grad_norm": 2.820202020849856,
+      "learning_rate": 3.9216574069164455e-06,
+      "loss": 0.8528,
+      "num_input_tokens_seen": 20265090,
+      "step": 971
+    },
+    {
+      "epoch": 0.11687608970119642,
+      "grad_norm": 5.011204560838788,
+      "learning_rate": 3.921441374681691e-06,
+      "loss": 0.8041,
+      "num_input_tokens_seen": 20284870,
+      "step": 972
+    },
+    {
+      "epoch": 0.1169963325918355,
+      "grad_norm": 1.8417517234955263,
+      "learning_rate": 3.921225050965808e-06,
+      "loss": 0.6458,
+      "num_input_tokens_seen": 20304475,
+      "step": 973
+    },
+    {
+      "epoch": 0.1171165754824746,
+      "grad_norm": 6.733886307733302,
+      "learning_rate": 3.921008435801612e-06,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 20323280,
+      "step": 974
+    },
+    {
+      "epoch": 0.11723681837311369,
+      "grad_norm": 2.9710460195717996,
+      "learning_rate": 3.920791529221963e-06,
+      "loss": 0.7562,
+      "num_input_tokens_seen": 20341675,
+      "step": 975
+    },
+    {
+      "epoch": 0.11735706126375278,
+      "grad_norm": 1.7988373669345004,
+      "learning_rate": 3.920574331259768e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 20362595,
+      "step": 976
+    },
+    {
+      "epoch": 0.11747730415439187,
+      "grad_norm": 2.4923285832027133,
+      "learning_rate": 3.9203568419479716e-06,
+      "loss": 0.7975,
+      "num_input_tokens_seen": 20382870,
+      "step": 977
+    },
+    {
+      "epoch": 0.11759754704503096,
+      "grad_norm": 2.134094773202001,
+      "learning_rate": 3.92013906131957e-06,
+      "loss": 0.7447,
+      "num_input_tokens_seen": 20401520,
+      "step": 978
+    },
+    {
+      "epoch": 0.11771778993567006,
+      "grad_norm": 1.6159386837941472,
+      "learning_rate": 3.9199209894076e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 20421555,
+      "step": 979
+    },
+    {
+      "epoch": 0.11783803282630914,
+      "grad_norm": 2.4867872755139033,
+      "learning_rate": 3.919702626245142e-06,
+      "loss": 0.8962,
+      "num_input_tokens_seen": 20440930,
+      "step": 980
+    },
+    {
+      "epoch": 0.11795827571694824,
+      "grad_norm": 4.779206276194037,
+      "learning_rate": 3.919483971865322e-06,
+      "loss": 0.6499,
+      "num_input_tokens_seen": 20460645,
+      "step": 981
+    },
+    {
+      "epoch": 0.11807851860758732,
+      "grad_norm": 2.323180913127229,
+      "learning_rate": 3.91926502630131e-06,
+      "loss": 0.8693,
+      "num_input_tokens_seen": 20480980,
+      "step": 982
+    },
+    {
+      "epoch": 0.11819876149822642,
+      "grad_norm": 1.9410805261900594,
+      "learning_rate": 3.91904578958632e-06,
+      "loss": 0.7243,
+      "num_input_tokens_seen": 20500115,
+      "step": 983
+    },
+    {
+      "epoch": 0.11831900438886551,
+      "grad_norm": 2.1025062805630483,
+      "learning_rate": 3.918826261753608e-06,
+      "loss": 0.847,
+      "num_input_tokens_seen": 20519415,
+      "step": 984
+    },
+    {
+      "epoch": 0.1184392472795046,
+      "grad_norm": 3.050376686172714,
+      "learning_rate": 3.918606442836478e-06,
+      "loss": 0.7016,
+      "num_input_tokens_seen": 20541355,
+      "step": 985
+    },
+    {
+      "epoch": 0.1185594901701437,
+      "grad_norm": 2.0459997571507005,
+      "learning_rate": 3.918386332868277e-06,
+      "loss": 0.7731,
+      "num_input_tokens_seen": 20559045,
+      "step": 986
+    },
+    {
+      "epoch": 0.11867973306078278,
+      "grad_norm": 3.8614565445852893,
+      "learning_rate": 3.918165931882394e-06,
+      "loss": 0.9354,
+      "num_input_tokens_seen": 20577165,
+      "step": 987
+    },
+    {
+      "epoch": 0.11879997595142187,
+      "grad_norm": 2.324334236100757,
+      "learning_rate": 3.917945239912264e-06,
+      "loss": 0.7612,
+      "num_input_tokens_seen": 20594360,
+      "step": 988
+    },
+    {
+      "epoch": 0.11892021884206096,
+      "grad_norm": 3.0301608156612527,
+      "learning_rate": 3.917724256991367e-06,
+      "loss": 0.762,
+      "num_input_tokens_seen": 20612825,
+      "step": 989
+    },
+    {
+      "epoch": 0.11904046173270005,
+      "grad_norm": 2.2905780921530687,
+      "learning_rate": 3.9175029831532245e-06,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 20632060,
+      "step": 990
+    },
+    {
+      "epoch": 0.11916070462333915,
+      "grad_norm": 2.6714023035010817,
+      "learning_rate": 3.917281418431404e-06,
+      "loss": 0.8768,
+      "num_input_tokens_seen": 20650825,
+      "step": 991
+    },
+    {
+      "epoch": 0.11928094751397823,
+      "grad_norm": 2.6250784998794505,
+      "learning_rate": 3.917059562859516e-06,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 20669870,
+      "step": 992
+    },
+    {
+      "epoch": 0.11940119040461733,
+      "grad_norm": 2.200459161672563,
+      "learning_rate": 3.916837416471218e-06,
+      "loss": 0.8788,
+      "num_input_tokens_seen": 20686210,
+      "step": 993
+    },
+    {
+      "epoch": 0.11952143329525641,
+      "grad_norm": 3.6026629209602627,
+      "learning_rate": 3.916614979300207e-06,
+      "loss": 0.7256,
+      "num_input_tokens_seen": 20700775,
+      "step": 994
+    },
+    {
+      "epoch": 0.11964167618589551,
+      "grad_norm": 1.8927753154406566,
+      "learning_rate": 3.9163922513802274e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 20722830,
+      "step": 995
+    },
+    {
+      "epoch": 0.1197619190765346,
+      "grad_norm": 4.283772290180489,
+      "learning_rate": 3.916169232745067e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 20740225,
+      "step": 996
+    },
+    {
+      "epoch": 0.11988216196717369,
+      "grad_norm": 2.917180180786399,
+      "learning_rate": 3.915945923428559e-06,
+      "loss": 0.915,
+      "num_input_tokens_seen": 20756470,
+      "step": 997
+    },
+    {
+      "epoch": 0.12000240485781279,
+      "grad_norm": 2.5285769235357507,
+      "learning_rate": 3.915722323464577e-06,
+      "loss": 0.8316,
+      "num_input_tokens_seen": 20774795,
+      "step": 998
+    },
+    {
+      "epoch": 0.12012264774845187,
+      "grad_norm": 2.701686022509613,
+      "learning_rate": 3.91549843288704e-06,
+      "loss": 0.7064,
+      "num_input_tokens_seen": 20798195,
+      "step": 999
+    },
+    {
+      "epoch": 0.12024289063909097,
+      "grad_norm": 2.649452779955016,
+      "learning_rate": 3.915274251729916e-06,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 20819205,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12036313352973005,
+      "grad_norm": 1.9241338945506175,
+      "learning_rate": 3.91504978002721e-06,
+      "loss": 0.9043,
+      "num_input_tokens_seen": 20837980,
+      "step": 1001
+    },
+    {
+      "epoch": 0.12048337642036915,
+      "grad_norm": 2.666476453767147,
+      "learning_rate": 3.914825017812974e-06,
+      "loss": 0.7677,
+      "num_input_tokens_seen": 20854350,
+      "step": 1002
+    },
+    {
+      "epoch": 0.12060361931100824,
+      "grad_norm": 7.471144322025861,
+      "learning_rate": 3.9145999651213065e-06,
+      "loss": 0.7273,
+      "num_input_tokens_seen": 20873310,
+      "step": 1003
+    },
+    {
+      "epoch": 0.12072386220164733,
+      "grad_norm": 2.302110803966868,
+      "learning_rate": 3.9143746219863465e-06,
+      "loss": 0.8869,
+      "num_input_tokens_seen": 20890135,
+      "step": 1004
+    },
+    {
+      "epoch": 0.12084410509228642,
+      "grad_norm": 1.0194162197377155,
+      "learning_rate": 3.914148988442278e-06,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 20945645,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1209643479829255,
+      "grad_norm": 2.7086078865004373,
+      "learning_rate": 3.91392306452333e-06,
+      "loss": 0.9599,
+      "num_input_tokens_seen": 20962440,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1210845908735646,
+      "grad_norm": 3.1018479466761737,
+      "learning_rate": 3.913696850263774e-06,
+      "loss": 0.6682,
+      "num_input_tokens_seen": 20976525,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1212048337642037,
+      "grad_norm": 3.538119670553433,
+      "learning_rate": 3.913470345697929e-06,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 20994875,
+      "step": 1008
+    },
+    {
+      "epoch": 0.12132507665484278,
+      "grad_norm": 2.1362114783801416,
+      "learning_rate": 3.913243550860153e-06,
+      "loss": 0.8598,
+      "num_input_tokens_seen": 21012360,
+      "step": 1009
+    },
+    {
+      "epoch": 0.12144531954548188,
+      "grad_norm": 1.9169075846518047,
+      "learning_rate": 3.913016465784852e-06,
+      "loss": 0.7601,
+      "num_input_tokens_seen": 21032755,
+      "step": 1010
+    },
+    {
+      "epoch": 0.12156556243612096,
+      "grad_norm": 3.128298270693645,
+      "learning_rate": 3.912789090506474e-06,
+      "loss": 0.7149,
+      "num_input_tokens_seen": 21051735,
+      "step": 1011
+    },
+    {
+      "epoch": 0.12168580532676006,
+      "grad_norm": 3.752731760917993,
+      "learning_rate": 3.9125614250595114e-06,
+      "loss": 0.7249,
+      "num_input_tokens_seen": 21067665,
+      "step": 1012
+    },
+    {
+      "epoch": 0.12180604821739914,
+      "grad_norm": 6.600958391176295,
+      "learning_rate": 3.912333469478502e-06,
+      "loss": 0.8889,
+      "num_input_tokens_seen": 21085350,
+      "step": 1013
+    },
+    {
+      "epoch": 0.12192629110803824,
+      "grad_norm": 1.967355705997668,
+      "learning_rate": 3.912105223798025e-06,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 21104490,
+      "step": 1014
+    },
+    {
+      "epoch": 0.12204653399867733,
+      "grad_norm": 1.1129023768347945,
+      "learning_rate": 3.9118766880527065e-06,
+      "loss": 0.72,
+      "num_input_tokens_seen": 21158645,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12216677688931642,
+      "grad_norm": 2.140429764695701,
+      "learning_rate": 3.9116478622772145e-06,
+      "loss": 0.7227,
+      "num_input_tokens_seen": 21176940,
+      "step": 1016
+    },
+    {
+      "epoch": 0.12228701977995551,
+      "grad_norm": 1.8088031100412274,
+      "learning_rate": 3.911418746506261e-06,
+      "loss": 0.8741,
+      "num_input_tokens_seen": 21196790,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1224072626705946,
+      "grad_norm": 2.165186597708197,
+      "learning_rate": 3.911189340774604e-06,
+      "loss": 0.7746,
+      "num_input_tokens_seen": 21216640,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1225275055612337,
+      "grad_norm": 1.7605926864044437,
+      "learning_rate": 3.910959645117043e-06,
+      "loss": 0.7985,
+      "num_input_tokens_seen": 21235695,
+      "step": 1019
+    },
+    {
+      "epoch": 0.12264774845187278,
+      "grad_norm": 0.832855925571204,
+      "learning_rate": 3.910729659568423e-06,
+      "loss": 0.5911,
+      "num_input_tokens_seen": 21292600,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12276799134251187,
+      "grad_norm": 4.733778872145136,
+      "learning_rate": 3.9104993841636344e-06,
+      "loss": 0.8259,
+      "num_input_tokens_seen": 21312890,
+      "step": 1021
+    },
+    {
+      "epoch": 0.12288823423315097,
+      "grad_norm": 2.1100773879392847,
+      "learning_rate": 3.910268818937608e-06,
+      "loss": 0.8065,
+      "num_input_tokens_seen": 21330765,
+      "step": 1022
+    },
+    {
+      "epoch": 0.12300847712379005,
+      "grad_norm": 3.069906921858533,
+      "learning_rate": 3.9100379639253196e-06,
+      "loss": 0.8746,
+      "num_input_tokens_seen": 21347205,
+      "step": 1023
+    },
+    {
+      "epoch": 0.12312872001442915,
+      "grad_norm": 3.255031621170376,
+      "learning_rate": 3.909806819161791e-06,
+      "loss": 0.856,
+      "num_input_tokens_seen": 21362400,
+      "step": 1024
+    },
+    {
+      "epoch": 0.12324896290506823,
+      "grad_norm": 1.9260925007559073,
+      "learning_rate": 3.909575384682086e-06,
+      "loss": 0.8602,
+      "num_input_tokens_seen": 21381000,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12336920579570733,
+      "grad_norm": 1.8598593797901357,
+      "learning_rate": 3.9093436605213144e-06,
+      "loss": 0.687,
+      "num_input_tokens_seen": 21401220,
+      "step": 1026
+    },
+    {
+      "epoch": 0.12348944868634643,
+      "grad_norm": 3.4914720893430338,
+      "learning_rate": 3.909111646714627e-06,
+      "loss": 0.787,
+      "num_input_tokens_seen": 21421785,
+      "step": 1027
+    },
+    {
+      "epoch": 0.12360969157698551,
+      "grad_norm": 2.132347825699078,
+      "learning_rate": 3.9088793432972206e-06,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 21440325,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1237299344676246,
+      "grad_norm": 2.1913524532986215,
+      "learning_rate": 3.908646750304336e-06,
+      "loss": 0.8169,
+      "num_input_tokens_seen": 21457730,
+      "step": 1029
+    },
+    {
+      "epoch": 0.12385017735826369,
+      "grad_norm": 1.573529337128982,
+      "learning_rate": 3.908413867771257e-06,
+      "loss": 0.8606,
+      "num_input_tokens_seen": 21476360,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12397042024890279,
+      "grad_norm": 2.4042228735044846,
+      "learning_rate": 3.908180695733311e-06,
+      "loss": 0.8079,
+      "num_input_tokens_seen": 21495570,
+      "step": 1031
+    },
+    {
+      "epoch": 0.12409066313954187,
+      "grad_norm": 3.0539586376361907,
+      "learning_rate": 3.907947234225871e-06,
+      "loss": 0.8273,
+      "num_input_tokens_seen": 21514300,
+      "step": 1032
+    },
+    {
+      "epoch": 0.12421090603018096,
+      "grad_norm": 2.0547908614224784,
+      "learning_rate": 3.907713483284352e-06,
+      "loss": 0.8697,
+      "num_input_tokens_seen": 21533495,
+      "step": 1033
+    },
+    {
+      "epoch": 0.12433114892082006,
+      "grad_norm": 2.850318331985092,
+      "learning_rate": 3.907479442944216e-06,
+      "loss": 0.9769,
+      "num_input_tokens_seen": 21551620,
+      "step": 1034
+    },
+    {
+      "epoch": 0.12445139181145914,
+      "grad_norm": 2.3318014283313295,
+      "learning_rate": 3.907245113240963e-06,
+      "loss": 0.9121,
+      "num_input_tokens_seen": 21569460,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12457163470209824,
+      "grad_norm": 2.020499722546725,
+      "learning_rate": 3.907010494210144e-06,
+      "loss": 0.7464,
+      "num_input_tokens_seen": 21591840,
+      "step": 1036
+    },
+    {
+      "epoch": 0.12469187759273732,
+      "grad_norm": 2.3653044803214476,
+      "learning_rate": 3.9067755858873495e-06,
+      "loss": 0.9167,
+      "num_input_tokens_seen": 21608360,
+      "step": 1037
+    },
+    {
+      "epoch": 0.12481212048337642,
+      "grad_norm": 0.911710098495449,
+      "learning_rate": 3.906540388308214e-06,
+      "loss": 0.6611,
+      "num_input_tokens_seen": 21667665,
+      "step": 1038
+    },
+    {
+      "epoch": 0.12493236337401552,
+      "grad_norm": 2.1144263629763174,
+      "learning_rate": 3.906304901508417e-06,
+      "loss": 0.8162,
+      "num_input_tokens_seen": 21686285,
+      "step": 1039
+    },
+    {
+      "epoch": 0.12505260626465461,
+      "grad_norm": 2.741549915883754,
+      "learning_rate": 3.9060691255236835e-06,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 21706570,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1251728491552937,
+      "grad_norm": 2.553860671191147,
+      "learning_rate": 3.905833060389778e-06,
+      "loss": 0.8162,
+      "num_input_tokens_seen": 21730410,
+      "step": 1041
+    },
+    {
+      "epoch": 0.12529309204593278,
+      "grad_norm": 2.2693661185320333,
+      "learning_rate": 3.905596706142513e-06,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 21751540,
+      "step": 1042
+    },
+    {
+      "epoch": 0.12541333493657186,
+      "grad_norm": 2.1536887635831037,
+      "learning_rate": 3.9053600628177435e-06,
+      "loss": 0.8558,
+      "num_input_tokens_seen": 21770870,
+      "step": 1043
+    },
+    {
+      "epoch": 0.12553357782721097,
+      "grad_norm": 2.289806123084169,
+      "learning_rate": 3.905123130451367e-06,
+      "loss": 0.8469,
+      "num_input_tokens_seen": 21791690,
+      "step": 1044
+    },
+    {
+      "epoch": 0.12565382071785006,
+      "grad_norm": 1.8934132417416718,
+      "learning_rate": 3.904885909079326e-06,
+      "loss": 0.7916,
+      "num_input_tokens_seen": 21810195,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12577406360848914,
+      "grad_norm": 3.4966685366285666,
+      "learning_rate": 3.904648398737607e-06,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 21828480,
+      "step": 1046
+    },
+    {
+      "epoch": 0.12589430649912825,
+      "grad_norm": 2.02358282457232,
+      "learning_rate": 3.9044105994622406e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 21849345,
+      "step": 1047
+    },
+    {
+      "epoch": 0.12601454938976733,
+      "grad_norm": 2.0805301061473367,
+      "learning_rate": 3.9041725112893005e-06,
+      "loss": 0.824,
+      "num_input_tokens_seen": 21870290,
+      "step": 1048
+    },
+    {
+      "epoch": 0.12613479228040642,
+      "grad_norm": 1.8387633852821261,
+      "learning_rate": 3.903934134254904e-06,
+      "loss": 0.751,
+      "num_input_tokens_seen": 21887800,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1262550351710455,
+      "grad_norm": 2.211872450775569,
+      "learning_rate": 3.903695468395213e-06,
+      "loss": 0.8444,
+      "num_input_tokens_seen": 21905390,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1263752780616846,
+      "grad_norm": 3.2615181254555936,
+      "learning_rate": 3.903456513746434e-06,
+      "loss": 0.5644,
+      "num_input_tokens_seen": 21926085,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1264955209523237,
+      "grad_norm": 1.9405068365942182,
+      "learning_rate": 3.903217270344815e-06,
+      "loss": 0.8759,
+      "num_input_tokens_seen": 21946055,
+      "step": 1052
+    },
+    {
+      "epoch": 0.12661576384296278,
+      "grad_norm": 12.250098913373938,
+      "learning_rate": 3.902977738226648e-06,
+      "loss": 0.8227,
+      "num_input_tokens_seen": 21966510,
+      "step": 1053
+    },
+    {
+      "epoch": 0.12673600673360189,
+      "grad_norm": 2.665855025520238,
+      "learning_rate": 3.902737917428273e-06,
+      "loss": 0.9069,
+      "num_input_tokens_seen": 21984395,
+      "step": 1054
+    },
+    {
+      "epoch": 0.12685624962424097,
+      "grad_norm": 3.1758880872398545,
+      "learning_rate": 3.902497807986068e-06,
+      "loss": 0.837,
+      "num_input_tokens_seen": 22004135,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12697649251488005,
+      "grad_norm": 3.7721605135438323,
+      "learning_rate": 3.902257409936458e-06,
+      "loss": 0.829,
+      "num_input_tokens_seen": 22024620,
+      "step": 1056
+    },
+    {
+      "epoch": 0.12709673540551916,
+      "grad_norm": 2.211102985639801,
+      "learning_rate": 3.902016723315912e-06,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 22042280,
+      "step": 1057
+    },
+    {
+      "epoch": 0.12721697829615825,
+      "grad_norm": 6.10875660756262,
+      "learning_rate": 3.901775748160941e-06,
+      "loss": 0.6903,
+      "num_input_tokens_seen": 22061180,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12733722118679733,
+      "grad_norm": 0.852554714532998,
+      "learning_rate": 3.901534484508101e-06,
+      "loss": 0.6373,
+      "num_input_tokens_seen": 22123575,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1274574640774364,
+      "grad_norm": 2.495044616275202,
+      "learning_rate": 3.901292932393991e-06,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 22142175,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12757770696807552,
+      "grad_norm": 3.8646298932163345,
+      "learning_rate": 3.9010510918552555e-06,
+      "loss": 0.8501,
+      "num_input_tokens_seen": 22160970,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1276979498587146,
+      "grad_norm": 6.437951000804665,
+      "learning_rate": 3.900808962928581e-06,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 22178305,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1278181927493537,
+      "grad_norm": 2.1712122903532634,
+      "learning_rate": 3.900566545650698e-06,
+      "loss": 0.885,
+      "num_input_tokens_seen": 22195695,
+      "step": 1063
+    },
+    {
+      "epoch": 0.1279384356399928,
+      "grad_norm": 2.5731375189546593,
+      "learning_rate": 3.900323840058381e-06,
+      "loss": 0.808,
+      "num_input_tokens_seen": 22213125,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12805867853063188,
+      "grad_norm": 1.998690081017005,
+      "learning_rate": 3.900080846188449e-06,
+      "loss": 0.8126,
+      "num_input_tokens_seen": 22231435,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12817892142127096,
+      "grad_norm": 1.87419399396287,
+      "learning_rate": 3.8998375640777625e-06,
+      "loss": 0.8097,
+      "num_input_tokens_seen": 22249025,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12829916431191005,
+      "grad_norm": 0.7547878965101757,
+      "learning_rate": 3.899593993763228e-06,
+      "loss": 0.5505,
+      "num_input_tokens_seen": 22309705,
+      "step": 1067
+    },
+    {
+      "epoch": 0.12841940720254916,
+      "grad_norm": 3.25408413968592,
+      "learning_rate": 3.899350135281796e-06,
+      "loss": 0.8095,
+      "num_input_tokens_seen": 22330425,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12853965009318824,
+      "grad_norm": 2.730455823675955,
+      "learning_rate": 3.899105988670458e-06,
+      "loss": 0.7862,
+      "num_input_tokens_seen": 22349650,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12865989298382732,
+      "grad_norm": 2.9839235693342987,
+      "learning_rate": 3.898861553966252e-06,
+      "loss": 0.8201,
+      "num_input_tokens_seen": 22369020,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12878013587446643,
+      "grad_norm": 1.7265713224793218,
+      "learning_rate": 3.898616831206257e-06,
+      "loss": 0.8716,
+      "num_input_tokens_seen": 22389165,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12890037876510552,
+      "grad_norm": 2.426067571234945,
+      "learning_rate": 3.8983718204276e-06,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 22411105,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1290206216557446,
+      "grad_norm": 2.105795372440824,
+      "learning_rate": 3.898126521667446e-06,
+      "loss": 0.8403,
+      "num_input_tokens_seen": 22430980,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12914086454638368,
+      "grad_norm": 8.424387604368214,
+      "learning_rate": 3.897880934963007e-06,
+      "loss": 0.8327,
+      "num_input_tokens_seen": 22450250,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1292611074370228,
+      "grad_norm": 2.5104778899307876,
+      "learning_rate": 3.89763506035154e-06,
+      "loss": 0.781,
+      "num_input_tokens_seen": 22467820,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12938135032766188,
+      "grad_norm": 1.91814868795964,
+      "learning_rate": 3.897388897870343e-06,
+      "loss": 0.8137,
+      "num_input_tokens_seen": 22488180,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12950159321830096,
+      "grad_norm": 1.8650031206643212,
+      "learning_rate": 3.89714244755676e-06,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 22509260,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12962183610894007,
+      "grad_norm": 3.3126320206993563,
+      "learning_rate": 3.896895709448175e-06,
+      "loss": 0.8583,
+      "num_input_tokens_seen": 22528730,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12974207899957915,
+      "grad_norm": 2.7257410358076313,
+      "learning_rate": 3.896648683582019e-06,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 22543785,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12986232189021824,
+      "grad_norm": 2.309038443497535,
+      "learning_rate": 3.896401369995766e-06,
+      "loss": 0.8048,
+      "num_input_tokens_seen": 22563310,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12998256478085732,
+      "grad_norm": 1.7422256657293502,
+      "learning_rate": 3.896153768726932e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 22583340,
+      "step": 1081
+    },
+    {
+      "epoch": 0.13010280767149643,
+      "grad_norm": 6.013861478407679,
+      "learning_rate": 3.8959058798130806e-06,
+      "loss": 0.8733,
+      "num_input_tokens_seen": 22601035,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1302230505621355,
+      "grad_norm": 2.109035342626544,
+      "learning_rate": 3.895657703291814e-06,
+      "loss": 0.7487,
+      "num_input_tokens_seen": 22620860,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1303432934527746,
+      "grad_norm": 2.74506499055514,
+      "learning_rate": 3.895409239200781e-06,
+      "loss": 0.7993,
+      "num_input_tokens_seen": 22636465,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1304635363434137,
+      "grad_norm": 2.459451829805312,
+      "learning_rate": 3.895160487577673e-06,
+      "loss": 0.9195,
+      "num_input_tokens_seen": 22653755,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1305837792340528,
+      "grad_norm": 0.8034069745463782,
+      "learning_rate": 3.894911448460226e-06,
+      "loss": 0.6314,
+      "num_input_tokens_seen": 22712790,
+      "step": 1086
+    },
+    {
+      "epoch": 0.13070402212469187,
+      "grad_norm": 2.087647873250765,
+      "learning_rate": 3.8946621218862195e-06,
+      "loss": 0.7254,
+      "num_input_tokens_seen": 22733510,
+      "step": 1087
+    },
+    {
+      "epoch": 0.13082426501533098,
+      "grad_norm": 2.3356661577486677,
+      "learning_rate": 3.894412507893475e-06,
+      "loss": 0.8837,
+      "num_input_tokens_seen": 22753510,
+      "step": 1088
+    },
+    {
+      "epoch": 0.13094450790597006,
+      "grad_norm": 4.958812919249834,
+      "learning_rate": 3.894162606519859e-06,
+      "loss": 0.7124,
+      "num_input_tokens_seen": 22772180,
+      "step": 1089
+    },
+    {
+      "epoch": 0.13106475079660915,
+      "grad_norm": 3.8791353870661482,
+      "learning_rate": 3.893912417803282e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 22791615,
+      "step": 1090
+    },
+    {
+      "epoch": 0.13118499368724823,
+      "grad_norm": 2.155677741131767,
+      "learning_rate": 3.8936619417816975e-06,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 22811665,
+      "step": 1091
+    },
+    {
+      "epoch": 0.13130523657788734,
+      "grad_norm": 4.0205827523181386,
+      "learning_rate": 3.8934111784931015e-06,
+      "loss": 0.7188,
+      "num_input_tokens_seen": 22828835,
+      "step": 1092
+    },
+    {
+      "epoch": 0.13142547946852642,
+      "grad_norm": 1.0558013032592701,
+      "learning_rate": 3.893160127975535e-06,
+      "loss": 0.6405,
+      "num_input_tokens_seen": 22889245,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1315457223591655,
+      "grad_norm": 3.3651356696016794,
+      "learning_rate": 3.8929087902670826e-06,
+      "loss": 0.803,
+      "num_input_tokens_seen": 22910595,
+      "step": 1094
+    },
+    {
+      "epoch": 0.13166596524980462,
+      "grad_norm": 0.9509244316896768,
+      "learning_rate": 3.8926571654058715e-06,
+      "loss": 0.6285,
+      "num_input_tokens_seen": 22966960,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1317862081404437,
+      "grad_norm": 3.37905317603145,
+      "learning_rate": 3.892405253430074e-06,
+      "loss": 0.7687,
+      "num_input_tokens_seen": 22984200,
+      "step": 1096
+    },
+    {
+      "epoch": 0.13190645103108278,
+      "grad_norm": 3.0841499998169977,
+      "learning_rate": 3.892153054377904e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 23001325,
+      "step": 1097
+    },
+    {
+      "epoch": 0.13202669392172187,
+      "grad_norm": 1.044069258755872,
+      "learning_rate": 3.891900568287619e-06,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 23053430,
+      "step": 1098
+    },
+    {
+      "epoch": 0.13214693681236098,
+      "grad_norm": 3.69115157410813,
+      "learning_rate": 3.891647795197523e-06,
+      "loss": 0.727,
+      "num_input_tokens_seen": 23069190,
+      "step": 1099
+    },
+    {
+      "epoch": 0.13226717970300006,
+      "grad_norm": 2.7243687240948855,
+      "learning_rate": 3.8913947351459605e-06,
+      "loss": 0.6874,
+      "num_input_tokens_seen": 23086450,
+      "step": 1100
+    },
+    {
+      "epoch": 0.13238742259363914,
+      "grad_norm": 1.8820422997154056,
+      "learning_rate": 3.89114138817132e-06,
+      "loss": 0.6712,
+      "num_input_tokens_seen": 23102835,
+      "step": 1101
+    },
+    {
+      "epoch": 0.13250766548427825,
+      "grad_norm": 1.938249978049348,
+      "learning_rate": 3.890887754312035e-06,
+      "loss": 0.8472,
+      "num_input_tokens_seen": 23120800,
+      "step": 1102
+    },
+    {
+      "epoch": 0.13262790837491734,
+      "grad_norm": 1.7591436046001099,
+      "learning_rate": 3.890633833606581e-06,
+      "loss": 0.8703,
+      "num_input_tokens_seen": 23140210,
+      "step": 1103
+    },
+    {
+      "epoch": 0.13274815126555642,
+      "grad_norm": 2.832583402007141,
+      "learning_rate": 3.890379626093477e-06,
+      "loss": 0.6933,
+      "num_input_tokens_seen": 23159680,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1328683941561955,
+      "grad_norm": 2.7227603818785644,
+      "learning_rate": 3.890125131811287e-06,
+      "loss": 0.914,
+      "num_input_tokens_seen": 23177450,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1329886370468346,
+      "grad_norm": 1.9389941127744519,
+      "learning_rate": 3.889870350798618e-06,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 23194515,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1331088799374737,
+      "grad_norm": 2.737836930131518,
+      "learning_rate": 3.889615283094119e-06,
+      "loss": 0.7856,
+      "num_input_tokens_seen": 23213425,
+      "step": 1107
+    },
+    {
+      "epoch": 0.13322912282811278,
+      "grad_norm": 3.8473974026545963,
+      "learning_rate": 3.889359928736485e-06,
+      "loss": 0.8443,
+      "num_input_tokens_seen": 23231090,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1333493657187519,
+      "grad_norm": 2.3842930921577516,
+      "learning_rate": 3.889104287764451e-06,
+      "loss": 0.9012,
+      "num_input_tokens_seen": 23251185,
+      "step": 1109
+    },
+    {
+      "epoch": 0.13346960860939097,
+      "grad_norm": 3.2756687011181267,
+      "learning_rate": 3.888848360216798e-06,
+      "loss": 0.9028,
+      "num_input_tokens_seen": 23268550,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13358985150003005,
+      "grad_norm": 0.8176872946321924,
+      "learning_rate": 3.888592146132351e-06,
+      "loss": 0.5815,
+      "num_input_tokens_seen": 23329540,
+      "step": 1111
+    },
+    {
+      "epoch": 0.13371009439066917,
+      "grad_norm": 11.073247614933557,
+      "learning_rate": 3.888335645549978e-06,
+      "loss": 0.7844,
+      "num_input_tokens_seen": 23349680,
+      "step": 1112
+    },
+    {
+      "epoch": 0.13383033728130825,
+      "grad_norm": 2.7809865837993284,
+      "learning_rate": 3.888078858508588e-06,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 23369260,
+      "step": 1113
+    },
+    {
+      "epoch": 0.13395058017194733,
+      "grad_norm": 5.213570399769174,
+      "learning_rate": 3.8878217850471365e-06,
+      "loss": 0.84,
+      "num_input_tokens_seen": 23388895,
+      "step": 1114
+    },
+    {
+      "epoch": 0.13407082306258641,
+      "grad_norm": 2.2506166832773884,
+      "learning_rate": 3.887564425204621e-06,
+      "loss": 0.7373,
+      "num_input_tokens_seen": 23410300,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13419106595322552,
+      "grad_norm": 0.8916455472722215,
+      "learning_rate": 3.887306779020083e-06,
+      "loss": 0.5797,
+      "num_input_tokens_seen": 23464675,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1343113088438646,
+      "grad_norm": 2.6633588282881187,
+      "learning_rate": 3.887048846532608e-06,
+      "loss": 0.7234,
+      "num_input_tokens_seen": 23481370,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1344315517345037,
+      "grad_norm": 0.8345272787021525,
+      "learning_rate": 3.8867906277813224e-06,
+      "loss": 0.5951,
+      "num_input_tokens_seen": 23539245,
+      "step": 1118
+    },
+    {
+      "epoch": 0.1345517946251428,
+      "grad_norm": 2.1196222409841594,
+      "learning_rate": 3.886532122805399e-06,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 23561445,
+      "step": 1119
+    },
+    {
+      "epoch": 0.13467203751578188,
+      "grad_norm": 3.642784471392457,
+      "learning_rate": 3.886273331644053e-06,
+      "loss": 0.8865,
+      "num_input_tokens_seen": 23580035,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13479228040642097,
+      "grad_norm": 8.108535940319493,
+      "learning_rate": 3.886014254336542e-06,
+      "loss": 0.82,
+      "num_input_tokens_seen": 23596230,
+      "step": 1121
+    },
+    {
+      "epoch": 0.13491252329706005,
+      "grad_norm": 1.7142577001519765,
+      "learning_rate": 3.885754890922168e-06,
+      "loss": 0.9222,
+      "num_input_tokens_seen": 23616280,
+      "step": 1122
+    },
+    {
+      "epoch": 0.13503276618769916,
+      "grad_norm": 1.8992419122116653,
+      "learning_rate": 3.885495241440277e-06,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 23640095,
+      "step": 1123
+    },
+    {
+      "epoch": 0.13515300907833824,
+      "grad_norm": 2.100107135973512,
+      "learning_rate": 3.885235305930257e-06,
+      "loss": 0.744,
+      "num_input_tokens_seen": 23658015,
+      "step": 1124
+    },
+    {
+      "epoch": 0.13527325196897733,
+      "grad_norm": 1.9459895809476313,
+      "learning_rate": 3.884975084431539e-06,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 23672685,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13539349485961644,
+      "grad_norm": 2.4850537161117248,
+      "learning_rate": 3.8847145769836e-06,
+      "loss": 0.9214,
+      "num_input_tokens_seen": 23688825,
+      "step": 1126
+    },
+    {
+      "epoch": 0.13551373775025552,
+      "grad_norm": 2.455654196382161,
+      "learning_rate": 3.884453783625959e-06,
+      "loss": 0.671,
+      "num_input_tokens_seen": 23706155,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1356339806408946,
+      "grad_norm": 5.7450600171540085,
+      "learning_rate": 3.884192704398176e-06,
+      "loss": 0.8617,
+      "num_input_tokens_seen": 23723075,
+      "step": 1128
+    },
+    {
+      "epoch": 0.13575422353153369,
+      "grad_norm": 1.75469449157673,
+      "learning_rate": 3.883931339339858e-06,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 23747180,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1358744664221728,
+      "grad_norm": 2.3289690118552713,
+      "learning_rate": 3.883669688490654e-06,
+      "loss": 0.79,
+      "num_input_tokens_seen": 23764670,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13599470931281188,
+      "grad_norm": 1.9796934311098506,
+      "learning_rate": 3.883407751890256e-06,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 23782995,
+      "step": 1131
+    },
+    {
+      "epoch": 0.13611495220345096,
+      "grad_norm": 1.7913152849158285,
+      "learning_rate": 3.8831455295783994e-06,
+      "loss": 0.8452,
+      "num_input_tokens_seen": 23801965,
+      "step": 1132
+    },
+    {
+      "epoch": 0.13623519509409007,
+      "grad_norm": 1.761249765237672,
+      "learning_rate": 3.882883021594864e-06,
+      "loss": 0.7337,
+      "num_input_tokens_seen": 23819825,
+      "step": 1133
+    },
+    {
+      "epoch": 0.13635543798472916,
+      "grad_norm": 2.7216820945681186,
+      "learning_rate": 3.8826202279794705e-06,
+      "loss": 0.8683,
+      "num_input_tokens_seen": 23836605,
+      "step": 1134
+    },
+    {
+      "epoch": 0.13647568087536824,
+      "grad_norm": 2.6393650700210767,
+      "learning_rate": 3.882357148772085e-06,
+      "loss": 0.7014,
+      "num_input_tokens_seen": 23853750,
+      "step": 1135
+    },
+    {
+      "epoch": 0.13659592376600732,
+      "grad_norm": 2.6877366932657223,
+      "learning_rate": 3.882093784012617e-06,
+      "loss": 0.8511,
+      "num_input_tokens_seen": 23872110,
+      "step": 1136
+    },
+    {
+      "epoch": 0.13671616665664643,
+      "grad_norm": 2.424823620199298,
+      "learning_rate": 3.881830133741019e-06,
+      "loss": 0.8305,
+      "num_input_tokens_seen": 23890695,
+      "step": 1137
+    },
+    {
+      "epoch": 0.13683640954728551,
+      "grad_norm": 3.212446380043299,
+      "learning_rate": 3.881566197997285e-06,
+      "loss": 0.7615,
+      "num_input_tokens_seen": 23906850,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1369566524379246,
+      "grad_norm": 1.8330378764762854,
+      "learning_rate": 3.881301976821456e-06,
+      "loss": 0.7467,
+      "num_input_tokens_seen": 23926600,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1370768953285637,
+      "grad_norm": 2.2959177787613845,
+      "learning_rate": 3.881037470253612e-06,
+      "loss": 0.9021,
+      "num_input_tokens_seen": 23945835,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1371971382192028,
+      "grad_norm": 2.5923940068779285,
+      "learning_rate": 3.88077267833388e-06,
+      "loss": 0.7963,
+      "num_input_tokens_seen": 23962070,
+      "step": 1141
+    },
+    {
+      "epoch": 0.13731738110984187,
+      "grad_norm": 2.120445260520303,
+      "learning_rate": 3.880507601102427e-06,
+      "loss": 0.8338,
+      "num_input_tokens_seen": 23979725,
+      "step": 1142
+    },
+    {
+      "epoch": 0.13743762400048098,
+      "grad_norm": 1.8588715922031163,
+      "learning_rate": 3.880242238599467e-06,
+      "loss": 0.8158,
+      "num_input_tokens_seen": 23995970,
+      "step": 1143
+    },
+    {
+      "epoch": 0.13755786689112007,
+      "grad_norm": 1.9458720166020516,
+      "learning_rate": 3.879976590865254e-06,
+      "loss": 0.8272,
+      "num_input_tokens_seen": 24015145,
+      "step": 1144
+    },
+    {
+      "epoch": 0.13767810978175915,
+      "grad_norm": 2.4217223962576906,
+      "learning_rate": 3.879710657940087e-06,
+      "loss": 0.8691,
+      "num_input_tokens_seen": 24033815,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13779835267239823,
+      "grad_norm": 2.248425995860865,
+      "learning_rate": 3.879444439864308e-06,
+      "loss": 0.6979,
+      "num_input_tokens_seen": 24053110,
+      "step": 1146
+    },
+    {
+      "epoch": 0.13791859556303734,
+      "grad_norm": 1.680901706438742,
+      "learning_rate": 3.879177936678301e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 24071835,
+      "step": 1147
+    },
+    {
+      "epoch": 0.13803883845367643,
+      "grad_norm": 2.3156656239983002,
+      "learning_rate": 3.878911148422496e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 24093030,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1381590813443155,
+      "grad_norm": 7.553188026840432,
+      "learning_rate": 3.878644075137364e-06,
+      "loss": 0.6975,
+      "num_input_tokens_seen": 24113400,
+      "step": 1149
+    },
+    {
+      "epoch": 0.13827932423495462,
+      "grad_norm": 2.4076782871457167,
+      "learning_rate": 3.878376716863418e-06,
+      "loss": 0.7869,
+      "num_input_tokens_seen": 24129420,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1383995671255937,
+      "grad_norm": 2.8081918755301905,
+      "learning_rate": 3.878109073641218e-06,
+      "loss": 0.7126,
+      "num_input_tokens_seen": 24148170,
+      "step": 1151
+    },
+    {
+      "epoch": 0.13851981001623279,
+      "grad_norm": 2.2826032507747973,
+      "learning_rate": 3.877841145511366e-06,
+      "loss": 0.8057,
+      "num_input_tokens_seen": 24170630,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13864005290687187,
+      "grad_norm": 1.7538648544210549,
+      "learning_rate": 3.8775729325145035e-06,
+      "loss": 0.8215,
+      "num_input_tokens_seen": 24189585,
+      "step": 1153
+    },
+    {
+      "epoch": 0.13876029579751098,
+      "grad_norm": 0.8428788057744127,
+      "learning_rate": 3.877304434691321e-06,
+      "loss": 0.6512,
+      "num_input_tokens_seen": 24256155,
+      "step": 1154
+    },
+    {
+      "epoch": 0.13888053868815006,
+      "grad_norm": 1.8341979651892155,
+      "learning_rate": 3.877035652082548e-06,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 24275320,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13900078157878915,
+      "grad_norm": 2.099726460213305,
+      "learning_rate": 3.87676658472896e-06,
+      "loss": 0.8435,
+      "num_input_tokens_seen": 24293850,
+      "step": 1156
+    },
+    {
+      "epoch": 0.13912102446942826,
+      "grad_norm": 2.593916888642099,
+      "learning_rate": 3.876497232671372e-06,
+      "loss": 0.8548,
+      "num_input_tokens_seen": 24313525,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13924126736006734,
+      "grad_norm": 2.3880264166108436,
+      "learning_rate": 3.876227595950647e-06,
+      "loss": 0.8347,
+      "num_input_tokens_seen": 24332675,
+      "step": 1158
+    },
+    {
+      "epoch": 0.13936151025070642,
+      "grad_norm": 1.579093652675001,
+      "learning_rate": 3.875957674607686e-06,
+      "loss": 0.7818,
+      "num_input_tokens_seen": 24354670,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1394817531413455,
+      "grad_norm": 1.989418947919792,
+      "learning_rate": 3.8756874686834386e-06,
+      "loss": 0.8812,
+      "num_input_tokens_seen": 24372605,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13960199603198462,
+      "grad_norm": 1.795260215261753,
+      "learning_rate": 3.875416978218893e-06,
+      "loss": 0.8001,
+      "num_input_tokens_seen": 24395520,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1397222389226237,
+      "grad_norm": 1.9946191462393907,
+      "learning_rate": 3.8751462032550835e-06,
+      "loss": 0.8245,
+      "num_input_tokens_seen": 24412245,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13984248181326278,
+      "grad_norm": 3.345937205454328,
+      "learning_rate": 3.874875143833085e-06,
+      "loss": 0.8183,
+      "num_input_tokens_seen": 24430205,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1399627247039019,
+      "grad_norm": 1.8249065040436567,
+      "learning_rate": 3.874603799994019e-06,
+      "loss": 0.6892,
+      "num_input_tokens_seen": 24453460,
+      "step": 1164
+    },
+    {
+      "epoch": 0.14008296759454097,
+      "grad_norm": 2.221714377845785,
+      "learning_rate": 3.874332171779046e-06,
+      "loss": 0.8753,
+      "num_input_tokens_seen": 24468060,
+      "step": 1165
+    },
+    {
+      "epoch": 0.14020321048518006,
+      "grad_norm": 2.2640638477222894,
+      "learning_rate": 3.874060259229373e-06,
+      "loss": 0.7578,
+      "num_input_tokens_seen": 24489355,
+      "step": 1166
+    },
+    {
+      "epoch": 0.14032345337581917,
+      "grad_norm": 2.40342453054275,
+      "learning_rate": 3.873788062386249e-06,
+      "loss": 0.9238,
+      "num_input_tokens_seen": 24507335,
+      "step": 1167
+    },
+    {
+      "epoch": 0.14044369626645825,
+      "grad_norm": 1.8149702578368532,
+      "learning_rate": 3.873515581290965e-06,
+      "loss": 0.8199,
+      "num_input_tokens_seen": 24531860,
+      "step": 1168
+    },
+    {
+      "epoch": 0.14056393915709733,
+      "grad_norm": 4.275014581038032,
+      "learning_rate": 3.8732428159848575e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 24550555,
+      "step": 1169
+    },
+    {
+      "epoch": 0.14068418204773642,
+      "grad_norm": 2.3863887431943245,
+      "learning_rate": 3.872969766509304e-06,
+      "loss": 0.7874,
+      "num_input_tokens_seen": 24570830,
+      "step": 1170
+    },
+    {
+      "epoch": 0.14080442493837553,
+      "grad_norm": 0.8493337014314831,
+      "learning_rate": 3.872696432905726e-06,
+      "loss": 0.5836,
+      "num_input_tokens_seen": 24631370,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1409246678290146,
+      "grad_norm": 2.19758273516171,
+      "learning_rate": 3.872422815215589e-06,
+      "loss": 0.7175,
+      "num_input_tokens_seen": 24650170,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1410449107196537,
+      "grad_norm": 2.1380036691566966,
+      "learning_rate": 3.8721489134803994e-06,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 24668680,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1411651536102928,
+      "grad_norm": 2.965164562060746,
+      "learning_rate": 3.871874727741707e-06,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 24685630,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1412853965009319,
+      "grad_norm": 1.8456976649706474,
+      "learning_rate": 3.871600258041108e-06,
+      "loss": 0.9595,
+      "num_input_tokens_seen": 24704875,
+      "step": 1175
+    },
+    {
+      "epoch": 0.14140563939157097,
+      "grad_norm": 2.582217555983381,
+      "learning_rate": 3.8713255044202375e-06,
+      "loss": 0.8531,
+      "num_input_tokens_seen": 24723585,
+      "step": 1176
+    },
+    {
+      "epoch": 0.14152588228221005,
+      "grad_norm": 2.077489827008881,
+      "learning_rate": 3.871050466920776e-06,
+      "loss": 0.811,
+      "num_input_tokens_seen": 24743210,
+      "step": 1177
+    },
+    {
+      "epoch": 0.14164612517284916,
+      "grad_norm": 2.104516187419509,
+      "learning_rate": 3.870775145584447e-06,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 24760710,
+      "step": 1178
+    },
+    {
+      "epoch": 0.14176636806348825,
+      "grad_norm": 3.075178549562538,
+      "learning_rate": 3.8704995404530145e-06,
+      "loss": 0.629,
+      "num_input_tokens_seen": 24776055,
+      "step": 1179
+    },
+    {
+      "epoch": 0.14188661095412733,
+      "grad_norm": 2.334877802904178,
+      "learning_rate": 3.870223651568289e-06,
+      "loss": 0.8465,
+      "num_input_tokens_seen": 24796490,
+      "step": 1180
+    },
+    {
+      "epoch": 0.14200685384476644,
+      "grad_norm": 2.0850337068461675,
+      "learning_rate": 3.869947478972123e-06,
+      "loss": 0.7993,
+      "num_input_tokens_seen": 24817235,
+      "step": 1181
+    },
+    {
+      "epoch": 0.14212709673540552,
+      "grad_norm": 3.1769878158552447,
+      "learning_rate": 3.869671022706412e-06,
+      "loss": 0.8095,
+      "num_input_tokens_seen": 24835685,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1422473396260446,
+      "grad_norm": 2.0978937815177945,
+      "learning_rate": 3.869394282813092e-06,
+      "loss": 0.6501,
+      "num_input_tokens_seen": 24854605,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1423675825166837,
+      "grad_norm": 2.6718273633011638,
+      "learning_rate": 3.869117259334147e-06,
+      "loss": 0.8945,
+      "num_input_tokens_seen": 24872250,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1424878254073228,
+      "grad_norm": 1.9983488153671112,
+      "learning_rate": 3.868839952311599e-06,
+      "loss": 0.822,
+      "num_input_tokens_seen": 24889925,
+      "step": 1185
+    },
+    {
+      "epoch": 0.14260806829796188,
+      "grad_norm": 2.3074406235932354,
+      "learning_rate": 3.868562361787516e-06,
+      "loss": 0.8033,
+      "num_input_tokens_seen": 24908775,
+      "step": 1186
+    },
+    {
+      "epoch": 0.14272831118860096,
+      "grad_norm": 2.238242931165128,
+      "learning_rate": 3.868284487804009e-06,
+      "loss": 0.6936,
+      "num_input_tokens_seen": 24927725,
+      "step": 1187
+    },
+    {
+      "epoch": 0.14284855407924008,
+      "grad_norm": 1.8743448081365774,
+      "learning_rate": 3.86800633040323e-06,
+      "loss": 0.7796,
+      "num_input_tokens_seen": 24948035,
+      "step": 1188
+    },
+    {
+      "epoch": 0.14296879696987916,
+      "grad_norm": 2.3079073454308787,
+      "learning_rate": 3.867727889627376e-06,
+      "loss": 0.7833,
+      "num_input_tokens_seen": 24967370,
+      "step": 1189
+    },
+    {
+      "epoch": 0.14308903986051824,
+      "grad_norm": 3.4205602945925473,
+      "learning_rate": 3.867449165518687e-06,
+      "loss": 0.7817,
+      "num_input_tokens_seen": 24983560,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14320928275115732,
+      "grad_norm": 1.917858774660838,
+      "learning_rate": 3.867170158119443e-06,
+      "loss": 0.7064,
+      "num_input_tokens_seen": 25002280,
+      "step": 1191
+    },
+    {
+      "epoch": 0.14332952564179643,
+      "grad_norm": 3.484199753147754,
+      "learning_rate": 3.866890867471972e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 25020470,
+      "step": 1192
+    },
+    {
+      "epoch": 0.14344976853243552,
+      "grad_norm": 2.5817815420350017,
+      "learning_rate": 3.86661129361864e-06,
+      "loss": 0.8962,
+      "num_input_tokens_seen": 25034680,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1435700114230746,
+      "grad_norm": 4.966148827242535,
+      "learning_rate": 3.866331436601859e-06,
+      "loss": 0.8584,
+      "num_input_tokens_seen": 25052395,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1436902543137137,
+      "grad_norm": 2.5516887621876903,
+      "learning_rate": 3.866051296464083e-06,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 25070950,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1438104972043528,
+      "grad_norm": 9.030350797148046,
+      "learning_rate": 3.86577087324781e-06,
+      "loss": 0.839,
+      "num_input_tokens_seen": 25087160,
+      "step": 1196
+    },
+    {
+      "epoch": 0.14393074009499188,
+      "grad_norm": 4.152997075747749,
+      "learning_rate": 3.865490166995578e-06,
+      "loss": 0.7686,
+      "num_input_tokens_seen": 25105110,
+      "step": 1197
+    },
+    {
+      "epoch": 0.144050982985631,
+      "grad_norm": 3.8598736239759592,
+      "learning_rate": 3.86520917774997e-06,
+      "loss": 0.835,
+      "num_input_tokens_seen": 25124265,
+      "step": 1198
+    },
+    {
+      "epoch": 0.14417122587627007,
+      "grad_norm": 2.280213964280314,
+      "learning_rate": 3.864927905553614e-06,
+      "loss": 0.7521,
+      "num_input_tokens_seen": 25141895,
+      "step": 1199
+    },
+    {
+      "epoch": 0.14429146876690915,
+      "grad_norm": 1.8087576353408834,
+      "learning_rate": 3.8646463504491765e-06,
+      "loss": 0.8816,
+      "num_input_tokens_seen": 25161750,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14441171165754824,
+      "grad_norm": 2.434974778757965,
+      "learning_rate": 3.8643645124793705e-06,
+      "loss": 0.8337,
+      "num_input_tokens_seen": 25180370,
+      "step": 1201
+    },
+    {
+      "epoch": 0.14453195454818735,
+      "grad_norm": 1.9602187323739744,
+      "learning_rate": 3.8640823916869515e-06,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 25204400,
+      "step": 1202
+    },
+    {
+      "epoch": 0.14465219743882643,
+      "grad_norm": 3.1561425020460825,
+      "learning_rate": 3.863799988114714e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 25226150,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1447724403294655,
+      "grad_norm": 3.02076886404522,
+      "learning_rate": 3.863517301805502e-06,
+      "loss": 0.7006,
+      "num_input_tokens_seen": 25244260,
+      "step": 1204
+    },
+    {
+      "epoch": 0.14489268322010462,
+      "grad_norm": 2.5779882368855245,
+      "learning_rate": 3.863234332802196e-06,
+      "loss": 0.9654,
+      "num_input_tokens_seen": 25256185,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1450129261107437,
+      "grad_norm": 3.602838935776098,
+      "learning_rate": 3.862951081147723e-06,
+      "loss": 0.742,
+      "num_input_tokens_seen": 25276070,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1451331690013828,
+      "grad_norm": 2.04430040283684,
+      "learning_rate": 3.862667546885053e-06,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 25294340,
+      "step": 1207
+    },
+    {
+      "epoch": 0.14525341189202187,
+      "grad_norm": 2.28504922120642,
+      "learning_rate": 3.8623837300571965e-06,
+      "loss": 0.728,
+      "num_input_tokens_seen": 25313045,
+      "step": 1208
+    },
+    {
+      "epoch": 0.14537365478266098,
+      "grad_norm": 2.2193805628405454,
+      "learning_rate": 3.8620996307072085e-06,
+      "loss": 0.8338,
+      "num_input_tokens_seen": 25333470,
+      "step": 1209
+    },
+    {
+      "epoch": 0.14549389767330007,
+      "grad_norm": 2.018436636492823,
+      "learning_rate": 3.861815248878188e-06,
+      "loss": 0.6379,
+      "num_input_tokens_seen": 25350675,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14561414056393915,
+      "grad_norm": 2.4206716335751204,
+      "learning_rate": 3.861530584613274e-06,
+      "loss": 0.8005,
+      "num_input_tokens_seen": 25368395,
+      "step": 1211
+    },
+    {
+      "epoch": 0.14573438345457826,
+      "grad_norm": 3.1399369198095357,
+      "learning_rate": 3.86124563795565e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 25386930,
+      "step": 1212
+    },
+    {
+      "epoch": 0.14585462634521734,
+      "grad_norm": 2.0542085685361515,
+      "learning_rate": 3.860960408948543e-06,
+      "loss": 0.7042,
+      "num_input_tokens_seen": 25408400,
+      "step": 1213
+    },
+    {
+      "epoch": 0.14597486923585642,
+      "grad_norm": 2.3518370057811695,
+      "learning_rate": 3.860674897635222e-06,
+      "loss": 0.9048,
+      "num_input_tokens_seen": 25424605,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1460951121264955,
+      "grad_norm": 1.786375568969041,
+      "learning_rate": 3.860389104058998e-06,
+      "loss": 0.8263,
+      "num_input_tokens_seen": 25442555,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14621535501713462,
+      "grad_norm": 2.0782593913866605,
+      "learning_rate": 3.860103028263227e-06,
+      "loss": 0.7223,
+      "num_input_tokens_seen": 25465380,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1463355979077737,
+      "grad_norm": 2.278154609023106,
+      "learning_rate": 3.859816670291304e-06,
+      "loss": 0.6918,
+      "num_input_tokens_seen": 25484195,
+      "step": 1217
+    },
+    {
+      "epoch": 0.14645584079841278,
+      "grad_norm": 2.378237717036832,
+      "learning_rate": 3.859530030186672e-06,
+      "loss": 0.8961,
+      "num_input_tokens_seen": 25500925,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1465760836890519,
+      "grad_norm": 2.344043564337135,
+      "learning_rate": 3.859243107992813e-06,
+      "loss": 0.8223,
+      "num_input_tokens_seen": 25519450,
+      "step": 1219
+    },
+    {
+      "epoch": 0.14669632657969098,
+      "grad_norm": 5.219753942097025,
+      "learning_rate": 3.858955903753252e-06,
+      "loss": 0.784,
+      "num_input_tokens_seen": 25537810,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14681656947033006,
+      "grad_norm": 1.669532118271603,
+      "learning_rate": 3.858668417511559e-06,
+      "loss": 0.8276,
+      "num_input_tokens_seen": 25560280,
+      "step": 1221
+    },
+    {
+      "epoch": 0.14693681236096917,
+      "grad_norm": 2.092838163063582,
+      "learning_rate": 3.8583806493113445e-06,
+      "loss": 0.7609,
+      "num_input_tokens_seen": 25578345,
+      "step": 1222
+    },
+    {
+      "epoch": 0.14705705525160825,
+      "grad_norm": 2.2600462017100504,
+      "learning_rate": 3.858092599196263e-06,
+      "loss": 0.813,
+      "num_input_tokens_seen": 25596020,
+      "step": 1223
+    },
+    {
+      "epoch": 0.14717729814224734,
+      "grad_norm": 2.482597700580532,
+      "learning_rate": 3.857804267210012e-06,
+      "loss": 0.8241,
+      "num_input_tokens_seen": 25615040,
+      "step": 1224
+    },
+    {
+      "epoch": 0.14729754103288642,
+      "grad_norm": 2.093665233275961,
+      "learning_rate": 3.857515653396331e-06,
+      "loss": 0.8799,
+      "num_input_tokens_seen": 25631970,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14741778392352553,
+      "grad_norm": 2.544159244375698,
+      "learning_rate": 3.857226757799001e-06,
+      "loss": 0.863,
+      "num_input_tokens_seen": 25649245,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1475380268141646,
+      "grad_norm": 2.28471473099243,
+      "learning_rate": 3.85693758046185e-06,
+      "loss": 0.7321,
+      "num_input_tokens_seen": 25667255,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1476582697048037,
+      "grad_norm": 4.382976322352578,
+      "learning_rate": 3.8566481214287435e-06,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 25685095,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1477785125954428,
+      "grad_norm": 1.9292208842864993,
+      "learning_rate": 3.8563583807435935e-06,
+      "loss": 0.8977,
+      "num_input_tokens_seen": 25700960,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1478987554860819,
+      "grad_norm": 2.059500723105536,
+      "learning_rate": 3.856068358450353e-06,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 25720630,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14801899837672097,
+      "grad_norm": 3.4310915179565304,
+      "learning_rate": 3.8557780545930186e-06,
+      "loss": 0.8527,
+      "num_input_tokens_seen": 25738765,
+      "step": 1231
+    },
+    {
+      "epoch": 0.14813924126736006,
+      "grad_norm": 2.729718138848321,
+      "learning_rate": 3.855487469215628e-06,
+      "loss": 0.7855,
+      "num_input_tokens_seen": 25757415,
+      "step": 1232
+    },
+    {
+      "epoch": 0.14825948415799917,
+      "grad_norm": 2.6386627463083996,
+      "learning_rate": 3.855196602362264e-06,
+      "loss": 0.7202,
+      "num_input_tokens_seen": 25780055,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14837972704863825,
+      "grad_norm": 2.121744463035372,
+      "learning_rate": 3.854905454077051e-06,
+      "loss": 0.9422,
+      "num_input_tokens_seen": 25797385,
+      "step": 1234
+    },
+    {
+      "epoch": 0.14849996993927733,
+      "grad_norm": 1.9113785942100372,
+      "learning_rate": 3.854614024404155e-06,
+      "loss": 0.8782,
+      "num_input_tokens_seen": 25815415,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14862021282991644,
+      "grad_norm": 1.9757040650042264,
+      "learning_rate": 3.8543223133877865e-06,
+      "loss": 0.8974,
+      "num_input_tokens_seen": 25833730,
+      "step": 1236
+    },
+    {
+      "epoch": 0.14874045572055553,
+      "grad_norm": 2.0258420940263413,
+      "learning_rate": 3.854030321072198e-06,
+      "loss": 0.8721,
+      "num_input_tokens_seen": 25853355,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1488606986111946,
+      "grad_norm": 2.073113965953351,
+      "learning_rate": 3.853738047501682e-06,
+      "loss": 0.7294,
+      "num_input_tokens_seen": 25873635,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1489809415018337,
+      "grad_norm": 2.8219345067994808,
+      "learning_rate": 3.85344549272058e-06,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 25891335,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1491011843924728,
+      "grad_norm": 1.8922606712380388,
+      "learning_rate": 3.853152656773269e-06,
+      "loss": 0.821,
+      "num_input_tokens_seen": 25912490,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14922142728311188,
+      "grad_norm": 1.8515395462787125,
+      "learning_rate": 3.852859539704174e-06,
+      "loss": 0.8409,
+      "num_input_tokens_seen": 25931510,
+      "step": 1241
+    },
+    {
+      "epoch": 0.14934167017375097,
+      "grad_norm": 2.012879530583986,
+      "learning_rate": 3.852566141557759e-06,
+      "loss": 0.7568,
+      "num_input_tokens_seen": 25951360,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14946191306439008,
+      "grad_norm": 2.442529972328362,
+      "learning_rate": 3.852272462378535e-06,
+      "loss": 0.7498,
+      "num_input_tokens_seen": 25968955,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14958215595502916,
+      "grad_norm": 1.9992961144426993,
+      "learning_rate": 3.85197850221105e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 25984975,
+      "step": 1244
+    },
+    {
+      "epoch": 0.14970239884566824,
+      "grad_norm": 1.7058298419999713,
+      "learning_rate": 3.851684261099899e-06,
+      "loss": 0.7608,
+      "num_input_tokens_seen": 26006435,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14982264173630733,
+      "grad_norm": 2.0742106175521626,
+      "learning_rate": 3.851389739089718e-06,
+      "loss": 0.8614,
+      "num_input_tokens_seen": 26022775,
+      "step": 1246
+    },
+    {
+      "epoch": 0.14994288462694644,
+      "grad_norm": 1.9388154032300862,
+      "learning_rate": 3.851094936225186e-06,
+      "loss": 0.7999,
+      "num_input_tokens_seen": 26043380,
+      "step": 1247
+    },
+    {
+      "epoch": 0.15006312751758552,
+      "grad_norm": 1.433384118925013,
+      "learning_rate": 3.850799852551024e-06,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 26065520,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1501833704082246,
+      "grad_norm": 2.5170145102294597,
+      "learning_rate": 3.850504488111995e-06,
+      "loss": 0.8604,
+      "num_input_tokens_seen": 26081915,
+      "step": 1249
+    },
+    {
+      "epoch": 0.15030361329886371,
+      "grad_norm": 1.654428917863714,
+      "learning_rate": 3.850208842952907e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 26100440,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1504238561895028,
+      "grad_norm": 1.7280710308919371,
+      "learning_rate": 3.849912917118608e-06,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 26121200,
+      "step": 1251
+    },
+    {
+      "epoch": 0.15054409908014188,
+      "grad_norm": 1.1027292970909603,
+      "learning_rate": 3.849616710653992e-06,
+      "loss": 0.6267,
+      "num_input_tokens_seen": 26182390,
+      "step": 1252
+    },
+    {
+      "epoch": 0.150664341970781,
+      "grad_norm": 1.7787948775832094,
+      "learning_rate": 3.84932022360399e-06,
+      "loss": 0.7491,
+      "num_input_tokens_seen": 26200775,
+      "step": 1253
+    },
+    {
+      "epoch": 0.15078458486142007,
+      "grad_norm": 2.852026570377626,
+      "learning_rate": 3.849023456013581e-06,
+      "loss": 0.8324,
+      "num_input_tokens_seen": 26218055,
+      "step": 1254
+    },
+    {
+      "epoch": 0.15090482775205916,
+      "grad_norm": 3.0744392624516963,
+      "learning_rate": 3.848726407927784e-06,
+      "loss": 0.6175,
+      "num_input_tokens_seen": 26238160,
+      "step": 1255
+    },
+    {
+      "epoch": 0.15102507064269824,
+      "grad_norm": 2.5917276064979964,
+      "learning_rate": 3.84842907939166e-06,
+      "loss": 0.8701,
+      "num_input_tokens_seen": 26257105,
+      "step": 1256
+    },
+    {
+      "epoch": 0.15114531353333735,
+      "grad_norm": 3.537445758643608,
+      "learning_rate": 3.8481314704503146e-06,
+      "loss": 0.7178,
+      "num_input_tokens_seen": 26276655,
+      "step": 1257
+    },
+    {
+      "epoch": 0.15126555642397643,
+      "grad_norm": 5.814397022407301,
+      "learning_rate": 3.847833581148895e-06,
+      "loss": 0.8826,
+      "num_input_tokens_seen": 26295285,
+      "step": 1258
+    },
+    {
+      "epoch": 0.15138579931461552,
+      "grad_norm": 1.9357352262497578,
+      "learning_rate": 3.84753541153259e-06,
+      "loss": 0.7958,
+      "num_input_tokens_seen": 26314575,
+      "step": 1259
+    },
+    {
+      "epoch": 0.15150604220525463,
+      "grad_norm": 1.659858690748384,
+      "learning_rate": 3.847236961646633e-06,
+      "loss": 0.8292,
+      "num_input_tokens_seen": 26333275,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1516262850958937,
+      "grad_norm": 2.6075052615493886,
+      "learning_rate": 3.846938231536296e-06,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 26348615,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1517465279865328,
+      "grad_norm": 1.9914077629325395,
+      "learning_rate": 3.8466392212468995e-06,
+      "loss": 0.8069,
+      "num_input_tokens_seen": 26368525,
+      "step": 1262
+    },
+    {
+      "epoch": 0.15186677087717187,
+      "grad_norm": 0.8301618911821336,
+      "learning_rate": 3.8463399308238e-06,
+      "loss": 0.645,
+      "num_input_tokens_seen": 26427350,
+      "step": 1263
+    },
+    {
+      "epoch": 0.15198701376781099,
+      "grad_norm": 1.9567545730329718,
+      "learning_rate": 3.846040360312402e-06,
+      "loss": 0.6363,
+      "num_input_tokens_seen": 26450330,
+      "step": 1264
+    },
+    {
+      "epoch": 0.15210725665845007,
+      "grad_norm": 2.7513082621790277,
+      "learning_rate": 3.8457405097581485e-06,
+      "loss": 0.815,
+      "num_input_tokens_seen": 26469040,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15222749954908915,
+      "grad_norm": 2.5777182047653016,
+      "learning_rate": 3.8454403792065275e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 26487580,
+      "step": 1266
+    },
+    {
+      "epoch": 0.15234774243972826,
+      "grad_norm": 2.0513283701240947,
+      "learning_rate": 3.845139968703068e-06,
+      "loss": 0.8525,
+      "num_input_tokens_seen": 26504820,
+      "step": 1267
+    },
+    {
+      "epoch": 0.15246798533036734,
+      "grad_norm": 2.0510147415541278,
+      "learning_rate": 3.844839278293342e-06,
+      "loss": 0.8252,
+      "num_input_tokens_seen": 26525390,
+      "step": 1268
+    },
+    {
+      "epoch": 0.15258822822100643,
+      "grad_norm": 2.5746028380602555,
+      "learning_rate": 3.8445383080229654e-06,
+      "loss": 0.7602,
+      "num_input_tokens_seen": 26541125,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1527084711116455,
+      "grad_norm": 5.025945563031634,
+      "learning_rate": 3.844237057937593e-06,
+      "loss": 0.7308,
+      "num_input_tokens_seen": 26559850,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15282871400228462,
+      "grad_norm": 3.504246336163769,
+      "learning_rate": 3.843935528082926e-06,
+      "loss": 0.7838,
+      "num_input_tokens_seen": 26580595,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1529489568929237,
+      "grad_norm": 2.1537130538107094,
+      "learning_rate": 3.843633718504704e-06,
+      "loss": 0.8484,
+      "num_input_tokens_seen": 26598760,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1530691997835628,
+      "grad_norm": 2.478490561573534,
+      "learning_rate": 3.843331629248715e-06,
+      "loss": 0.8919,
+      "num_input_tokens_seen": 26616080,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1531894426742019,
+      "grad_norm": 2.272641781374053,
+      "learning_rate": 3.843029260360782e-06,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 26634170,
+      "step": 1274
+    },
+    {
+      "epoch": 0.15330968556484098,
+      "grad_norm": 4.182756894710253,
+      "learning_rate": 3.8427266118867755e-06,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 26653640,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15342992845548006,
+      "grad_norm": 2.4697343284779634,
+      "learning_rate": 3.842423683872608e-06,
+      "loss": 0.8208,
+      "num_input_tokens_seen": 26673935,
+      "step": 1276
+    },
+    {
+      "epoch": 0.15355017134611917,
+      "grad_norm": 2.758921912262511,
+      "learning_rate": 3.842120476364232e-06,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 26692105,
+      "step": 1277
+    },
+    {
+      "epoch": 0.15367041423675826,
+      "grad_norm": 2.2516647944219392,
+      "learning_rate": 3.841816989407644e-06,
+      "loss": 0.8296,
+      "num_input_tokens_seen": 26707315,
+      "step": 1278
+    },
+    {
+      "epoch": 0.15379065712739734,
+      "grad_norm": 2.0724141887527585,
+      "learning_rate": 3.841513223048884e-06,
+      "loss": 0.7612,
+      "num_input_tokens_seen": 26727720,
+      "step": 1279
+    },
+    {
+      "epoch": 0.15391090001803642,
+      "grad_norm": 3.016533848443925,
+      "learning_rate": 3.841209177334031e-06,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 26745800,
+      "step": 1280
+    },
+    {
+      "epoch": 0.15403114290867553,
+      "grad_norm": 2.9306784155719496,
+      "learning_rate": 3.84090485230921e-06,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 26763760,
+      "step": 1281
+    },
+    {
+      "epoch": 0.15415138579931462,
+      "grad_norm": 4.179257497991434,
+      "learning_rate": 3.840600248020588e-06,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 26780420,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1542716286899537,
+      "grad_norm": 2.4084822387968345,
+      "learning_rate": 3.840295364514371e-06,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 26797520,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1543918715805928,
+      "grad_norm": 2.437446661749323,
+      "learning_rate": 3.83999020183681e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 26815935,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1545121144712319,
+      "grad_norm": 2.0890401264597958,
+      "learning_rate": 3.839684760034199e-06,
+      "loss": 0.7765,
+      "num_input_tokens_seen": 26833860,
+      "step": 1285
+    },
+    {
+      "epoch": 0.15463235736187098,
+      "grad_norm": 2.5797079363326687,
+      "learning_rate": 3.8393790391528716e-06,
+      "loss": 0.6503,
+      "num_input_tokens_seen": 26854275,
+      "step": 1286
+    },
+    {
+      "epoch": 0.15475260025251006,
+      "grad_norm": 2.1434082151630536,
+      "learning_rate": 3.8390730392392075e-06,
+      "loss": 0.8841,
+      "num_input_tokens_seen": 26873975,
+      "step": 1287
+    },
+    {
+      "epoch": 0.15487284314314917,
+      "grad_norm": 2.3177742584938477,
+      "learning_rate": 3.838766760339626e-06,
+      "loss": 0.794,
+      "num_input_tokens_seen": 26892220,
+      "step": 1288
+    },
+    {
+      "epoch": 0.15499308603378825,
+      "grad_norm": 2.796655175919591,
+      "learning_rate": 3.838460202500587e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 26907730,
+      "step": 1289
+    },
+    {
+      "epoch": 0.15511332892442733,
+      "grad_norm": 2.1343041116359167,
+      "learning_rate": 3.838153365768599e-06,
+      "loss": 0.7384,
+      "num_input_tokens_seen": 26923960,
+      "step": 1290
+    },
+    {
+      "epoch": 0.15523357181506645,
+      "grad_norm": 2.569057553779175,
+      "learning_rate": 3.837846250190206e-06,
+      "loss": 0.7481,
+      "num_input_tokens_seen": 26946545,
+      "step": 1291
+    },
+    {
+      "epoch": 0.15535381470570553,
+      "grad_norm": 2.3866090308302,
+      "learning_rate": 3.837538855811998e-06,
+      "loss": 0.773,
+      "num_input_tokens_seen": 26964440,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1554740575963446,
+      "grad_norm": 2.799496795287829,
+      "learning_rate": 3.837231182680606e-06,
+      "loss": 0.6982,
+      "num_input_tokens_seen": 26982125,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1555943004869837,
+      "grad_norm": 1.8906334803766507,
+      "learning_rate": 3.836923230842706e-06,
+      "loss": 0.7595,
+      "num_input_tokens_seen": 27000960,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1557145433776228,
+      "grad_norm": 2.220523574827949,
+      "learning_rate": 3.836615000345011e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 27018860,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1558347862682619,
+      "grad_norm": 2.207135707949427,
+      "learning_rate": 3.836306491234282e-06,
+      "loss": 0.7722,
+      "num_input_tokens_seen": 27036430,
+      "step": 1296
+    },
+    {
+      "epoch": 0.15595502915890097,
+      "grad_norm": 2.493064488433665,
+      "learning_rate": 3.835997703557317e-06,
+      "loss": 0.7408,
+      "num_input_tokens_seen": 27052890,
+      "step": 1297
+    },
+    {
+      "epoch": 0.15607527204954008,
+      "grad_norm": 2.552935961981001,
+      "learning_rate": 3.83568863736096e-06,
+      "loss": 0.8015,
+      "num_input_tokens_seen": 27071480,
+      "step": 1298
+    },
+    {
+      "epoch": 0.15619551494017916,
+      "grad_norm": 2.439422893072988,
+      "learning_rate": 3.8353792926920975e-06,
+      "loss": 0.8841,
+      "num_input_tokens_seen": 27089850,
+      "step": 1299
+    },
+    {
+      "epoch": 0.15631575783081825,
+      "grad_norm": 2.521620051677975,
+      "learning_rate": 3.835069669597655e-06,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 27107960,
+      "step": 1300
+    },
+    {
+      "epoch": 0.15643600072145733,
+      "grad_norm": 2.1091572306019626,
+      "learning_rate": 3.834759768124603e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 27126555,
+      "step": 1301
+    },
+    {
+      "epoch": 0.15655624361209644,
+      "grad_norm": 76.8875280061683,
+      "learning_rate": 3.834449588319953e-06,
+      "loss": 0.7572,
+      "num_input_tokens_seen": 27144310,
+      "step": 1302
+    },
+    {
+      "epoch": 0.15667648650273552,
+      "grad_norm": 3.3633950303514397,
+      "learning_rate": 3.834139130230758e-06,
+      "loss": 0.8376,
+      "num_input_tokens_seen": 27163335,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1567967293933746,
+      "grad_norm": 1.916679119412341,
+      "learning_rate": 3.833828393904117e-06,
+      "loss": 0.811,
+      "num_input_tokens_seen": 27183335,
+      "step": 1304
+    },
+    {
+      "epoch": 0.15691697228401372,
+      "grad_norm": 2.322982845974901,
+      "learning_rate": 3.833517379387165e-06,
+      "loss": 0.7683,
+      "num_input_tokens_seen": 27199510,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1570372151746528,
+      "grad_norm": 1.8592348824121185,
+      "learning_rate": 3.833206086727085e-06,
+      "loss": 0.89,
+      "num_input_tokens_seen": 27218580,
+      "step": 1306
+    },
+    {
+      "epoch": 0.15715745806529188,
+      "grad_norm": 2.1543387627220256,
+      "learning_rate": 3.8328945159710994e-06,
+      "loss": 0.709,
+      "num_input_tokens_seen": 27238480,
+      "step": 1307
+    },
+    {
+      "epoch": 0.157277700955931,
+      "grad_norm": 2.0037505641022597,
+      "learning_rate": 3.832582667166473e-06,
+      "loss": 0.8788,
+      "num_input_tokens_seen": 27258010,
+      "step": 1308
+    },
+    {
+      "epoch": 0.15739794384657008,
+      "grad_norm": 2.0021781237442995,
+      "learning_rate": 3.8322705403605125e-06,
+      "loss": 0.8171,
+      "num_input_tokens_seen": 27278075,
+      "step": 1309
+    },
+    {
+      "epoch": 0.15751818673720916,
+      "grad_norm": 2.2334019858596665,
+      "learning_rate": 3.831958135600568e-06,
+      "loss": 0.8163,
+      "num_input_tokens_seen": 27295345,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15763842962784824,
+      "grad_norm": 3.429400100664831,
+      "learning_rate": 3.831645452934032e-06,
+      "loss": 0.8001,
+      "num_input_tokens_seen": 27313495,
+      "step": 1311
+    },
+    {
+      "epoch": 0.15775867251848735,
+      "grad_norm": 2.5480806705361196,
+      "learning_rate": 3.831332492408336e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 27334625,
+      "step": 1312
+    },
+    {
+      "epoch": 0.15787891540912644,
+      "grad_norm": 1.9905738779330642,
+      "learning_rate": 3.831019254070957e-06,
+      "loss": 0.6825,
+      "num_input_tokens_seen": 27352130,
+      "step": 1313
+    },
+    {
+      "epoch": 0.15799915829976552,
+      "grad_norm": 2.5776251697101444,
+      "learning_rate": 3.8307057379694135e-06,
+      "loss": 0.9569,
+      "num_input_tokens_seen": 27371185,
+      "step": 1314
+    },
+    {
+      "epoch": 0.15811940119040463,
+      "grad_norm": 2.370712191880963,
+      "learning_rate": 3.830391944151264e-06,
+      "loss": 0.816,
+      "num_input_tokens_seen": 27386785,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1582396440810437,
+      "grad_norm": 1.8733668286459888,
+      "learning_rate": 3.830077872664114e-06,
+      "loss": 0.6696,
+      "num_input_tokens_seen": 27407630,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1583598869716828,
+      "grad_norm": 1.8680214783940308,
+      "learning_rate": 3.829763523555604e-06,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 27427750,
+      "step": 1317
+    },
+    {
+      "epoch": 0.15848012986232188,
+      "grad_norm": 2.3373558070949296,
+      "learning_rate": 3.829448896873423e-06,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 27446570,
+      "step": 1318
+    },
+    {
+      "epoch": 0.158600372752961,
+      "grad_norm": 2.201530039165448,
+      "learning_rate": 3.829133992665299e-06,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 27465415,
+      "step": 1319
+    },
+    {
+      "epoch": 0.15872061564360007,
+      "grad_norm": 2.4817220948294927,
+      "learning_rate": 3.828818810979002e-06,
+      "loss": 0.8923,
+      "num_input_tokens_seen": 27483465,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15884085853423915,
+      "grad_norm": 4.934492941257204,
+      "learning_rate": 3.8285033518623454e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 27503435,
+      "step": 1321
+    },
+    {
+      "epoch": 0.15896110142487826,
+      "grad_norm": 2.664115441836404,
+      "learning_rate": 3.8281876153631845e-06,
+      "loss": 0.8148,
+      "num_input_tokens_seen": 27519910,
+      "step": 1322
+    },
+    {
+      "epoch": 0.15908134431551735,
+      "grad_norm": 2.921151980929316,
+      "learning_rate": 3.827871601529416e-06,
+      "loss": 0.6482,
+      "num_input_tokens_seen": 27538150,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15920158720615643,
+      "grad_norm": 8.772096420329104,
+      "learning_rate": 3.827555310408979e-06,
+      "loss": 0.8006,
+      "num_input_tokens_seen": 27557265,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1593218300967955,
+      "grad_norm": 2.0999620719837644,
+      "learning_rate": 3.827238742049854e-06,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 27577280,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15944207298743462,
+      "grad_norm": 2.057588485196255,
+      "learning_rate": 3.826921896500066e-06,
+      "loss": 0.531,
+      "num_input_tokens_seen": 27598285,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1595623158780737,
+      "grad_norm": 2.317607682973034,
+      "learning_rate": 3.826604773807678e-06,
+      "loss": 0.787,
+      "num_input_tokens_seen": 27615980,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1596825587687128,
+      "grad_norm": 3.2685684682426785,
+      "learning_rate": 3.826287374020798e-06,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 27630505,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1598028016593519,
+      "grad_norm": 3.194547399945437,
+      "learning_rate": 3.825969697187575e-06,
+      "loss": 0.8198,
+      "num_input_tokens_seen": 27649555,
+      "step": 1329
+    },
+    {
+      "epoch": 0.15992304454999098,
+      "grad_norm": 1.9559845177944728,
+      "learning_rate": 3.8256517433562015e-06,
+      "loss": 0.699,
+      "num_input_tokens_seen": 27667215,
+      "step": 1330
+    },
+    {
+      "epoch": 0.16004328744063007,
+      "grad_norm": 3.3455049748582626,
+      "learning_rate": 3.82533351257491e-06,
+      "loss": 0.9236,
+      "num_input_tokens_seen": 27684885,
+      "step": 1331
+    },
+    {
+      "epoch": 0.16016353033126918,
+      "grad_norm": 2.1910572080448065,
+      "learning_rate": 3.825015004891975e-06,
+      "loss": 0.8847,
+      "num_input_tokens_seen": 27703345,
+      "step": 1332
+    },
+    {
+      "epoch": 0.16028377322190826,
+      "grad_norm": 2.730165213193173,
+      "learning_rate": 3.824696220355716e-06,
+      "loss": 0.7603,
+      "num_input_tokens_seen": 27724655,
+      "step": 1333
+    },
+    {
+      "epoch": 0.16040401611254734,
+      "grad_norm": 5.972470411195008,
+      "learning_rate": 3.824377159014491e-06,
+      "loss": 0.7926,
+      "num_input_tokens_seen": 27745270,
+      "step": 1334
+    },
+    {
+      "epoch": 0.16052425900318643,
+      "grad_norm": 5.431198173286477,
+      "learning_rate": 3.824057820916702e-06,
+      "loss": 0.8493,
+      "num_input_tokens_seen": 27762195,
+      "step": 1335
+    },
+    {
+      "epoch": 0.16064450189382554,
+      "grad_norm": 2.8398681747811123,
+      "learning_rate": 3.8237382061107904e-06,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 27778635,
+      "step": 1336
+    },
+    {
+      "epoch": 0.16076474478446462,
+      "grad_norm": 3.616908841007189,
+      "learning_rate": 3.823418314645243e-06,
+      "loss": 0.7888,
+      "num_input_tokens_seen": 27797230,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1608849876751037,
+      "grad_norm": 2.807757964799803,
+      "learning_rate": 3.823098146568588e-06,
+      "loss": 0.7547,
+      "num_input_tokens_seen": 27816655,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1610052305657428,
+      "grad_norm": 3.513618632551631,
+      "learning_rate": 3.822777701929394e-06,
+      "loss": 0.717,
+      "num_input_tokens_seen": 27838200,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1611254734563819,
+      "grad_norm": 3.1066323013685686,
+      "learning_rate": 3.8224569807762714e-06,
+      "loss": 0.7388,
+      "num_input_tokens_seen": 27857240,
+      "step": 1340
+    },
+    {
+      "epoch": 0.16124571634702098,
+      "grad_norm": 2.926871461601744,
+      "learning_rate": 3.822135983157873e-06,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 27876235,
+      "step": 1341
+    },
+    {
+      "epoch": 0.16136595923766006,
+      "grad_norm": 2.7044087534467054,
+      "learning_rate": 3.821814709122896e-06,
+      "loss": 0.8404,
+      "num_input_tokens_seen": 27894005,
+      "step": 1342
+    },
+    {
+      "epoch": 0.16148620212829917,
+      "grad_norm": 6.947638090296441,
+      "learning_rate": 3.821493158720076e-06,
+      "loss": 0.8544,
+      "num_input_tokens_seen": 27912830,
+      "step": 1343
+    },
+    {
+      "epoch": 0.16160644501893826,
+      "grad_norm": 4.806349228038643,
+      "learning_rate": 3.821171331998191e-06,
+      "loss": 0.7316,
+      "num_input_tokens_seen": 27929080,
+      "step": 1344
+    },
+    {
+      "epoch": 0.16172668790957734,
+      "grad_norm": 0.8439097444460514,
+      "learning_rate": 3.820849229006064e-06,
+      "loss": 0.5889,
+      "num_input_tokens_seen": 27996550,
+      "step": 1345
+    },
+    {
+      "epoch": 0.16184693080021645,
+      "grad_norm": 2.631377025327218,
+      "learning_rate": 3.8205268497925564e-06,
+      "loss": 0.7126,
+      "num_input_tokens_seen": 28016740,
+      "step": 1346
+    },
+    {
+      "epoch": 0.16196717369085553,
+      "grad_norm": 3.31699266738122,
+      "learning_rate": 3.8202041944065725e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 28032280,
+      "step": 1347
+    },
+    {
+      "epoch": 0.16208741658149461,
+      "grad_norm": 2.2779831743242087,
+      "learning_rate": 3.819881262897061e-06,
+      "loss": 0.7373,
+      "num_input_tokens_seen": 28050135,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1622076594721337,
+      "grad_norm": 5.853857014923446,
+      "learning_rate": 3.819558055313008e-06,
+      "loss": 0.7329,
+      "num_input_tokens_seen": 28070540,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1623279023627728,
+      "grad_norm": 2.063878564910883,
+      "learning_rate": 3.819234571703444e-06,
+      "loss": 0.7672,
+      "num_input_tokens_seen": 28089085,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1624481452534119,
+      "grad_norm": 4.151467106043026,
+      "learning_rate": 3.8189108121174435e-06,
+      "loss": 0.8506,
+      "num_input_tokens_seen": 28108570,
+      "step": 1351
+    },
+    {
+      "epoch": 0.16256838814405097,
+      "grad_norm": 1.8173507918921392,
+      "learning_rate": 3.818586776604118e-06,
+      "loss": 0.8305,
+      "num_input_tokens_seen": 28128930,
+      "step": 1352
+    },
+    {
+      "epoch": 0.16268863103469008,
+      "grad_norm": 6.114712393327527,
+      "learning_rate": 3.818262465212625e-06,
+      "loss": 0.6209,
+      "num_input_tokens_seen": 28148775,
+      "step": 1353
+    },
+    {
+      "epoch": 0.16280887392532917,
+      "grad_norm": 3.124103804029256,
+      "learning_rate": 3.817937877992161e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 28165790,
+      "step": 1354
+    },
+    {
+      "epoch": 0.16292911681596825,
+      "grad_norm": 2.807021774612654,
+      "learning_rate": 3.817613014991967e-06,
+      "loss": 0.8469,
+      "num_input_tokens_seen": 28181650,
+      "step": 1355
+    },
+    {
+      "epoch": 0.16304935970660733,
+      "grad_norm": 3.3745525051579333,
+      "learning_rate": 3.817287876261323e-06,
+      "loss": 0.756,
+      "num_input_tokens_seen": 28201705,
+      "step": 1356
+    },
+    {
+      "epoch": 0.16316960259724644,
+      "grad_norm": 2.3411486132923782,
+      "learning_rate": 3.816962461849553e-06,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 28223295,
+      "step": 1357
+    },
+    {
+      "epoch": 0.16328984548788553,
+      "grad_norm": 4.4311518968104275,
+      "learning_rate": 3.8166367718060235e-06,
+      "loss": 0.8403,
+      "num_input_tokens_seen": 28242905,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1634100883785246,
+      "grad_norm": 3.4719845149613344,
+      "learning_rate": 3.816310806180139e-06,
+      "loss": 0.7534,
+      "num_input_tokens_seen": 28261035,
+      "step": 1359
+    },
+    {
+      "epoch": 0.16353033126916372,
+      "grad_norm": 2.0423168503709865,
+      "learning_rate": 3.81598456502135e-06,
+      "loss": 0.799,
+      "num_input_tokens_seen": 28280775,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1636505741598028,
+      "grad_norm": 5.1445463270225025,
+      "learning_rate": 3.8156580483791455e-06,
+      "loss": 0.8652,
+      "num_input_tokens_seen": 28295685,
+      "step": 1361
+    },
+    {
+      "epoch": 0.16377081705044189,
+      "grad_norm": 3.16540128174647,
+      "learning_rate": 3.815331256303059e-06,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 28315435,
+      "step": 1362
+    },
+    {
+      "epoch": 0.163891059941081,
+      "grad_norm": 2.9248364321935614,
+      "learning_rate": 3.815004188842665e-06,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 28333195,
+      "step": 1363
+    },
+    {
+      "epoch": 0.16401130283172008,
+      "grad_norm": 1.9862311408795017,
+      "learning_rate": 3.814676846047578e-06,
+      "loss": 0.7913,
+      "num_input_tokens_seen": 28353790,
+      "step": 1364
+    },
+    {
+      "epoch": 0.16413154572235916,
+      "grad_norm": 1.8935930062928639,
+      "learning_rate": 3.8143492279674565e-06,
+      "loss": 0.6979,
+      "num_input_tokens_seen": 28376205,
+      "step": 1365
+    },
+    {
+      "epoch": 0.16425178861299825,
+      "grad_norm": 0.9834768612405342,
+      "learning_rate": 3.8140213346519993e-06,
+      "loss": 0.6581,
+      "num_input_tokens_seen": 28426520,
+      "step": 1366
+    },
+    {
+      "epoch": 0.16437203150363736,
+      "grad_norm": 1.7672564952023309,
+      "learning_rate": 3.813693166150948e-06,
+      "loss": 0.7667,
+      "num_input_tokens_seen": 28446450,
+      "step": 1367
+    },
+    {
+      "epoch": 0.16449227439427644,
+      "grad_norm": 2.883140828353441,
+      "learning_rate": 3.813364722514086e-06,
+      "loss": 0.8555,
+      "num_input_tokens_seen": 28464505,
+      "step": 1368
+    },
+    {
+      "epoch": 0.16461251728491552,
+      "grad_norm": 2.2682538442601166,
+      "learning_rate": 3.8130360037912368e-06,
+      "loss": 0.8092,
+      "num_input_tokens_seen": 28480670,
+      "step": 1369
+    },
+    {
+      "epoch": 0.16473276017555463,
+      "grad_norm": 2.263892416476242,
+      "learning_rate": 3.812707010032268e-06,
+      "loss": 0.813,
+      "num_input_tokens_seen": 28499445,
+      "step": 1370
+    },
+    {
+      "epoch": 0.16485300306619372,
+      "grad_norm": 2.7725838228584827,
+      "learning_rate": 3.8123777412870863e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 28518665,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1649732459568328,
+      "grad_norm": 2.317519558579017,
+      "learning_rate": 3.812048197605643e-06,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 28537280,
+      "step": 1372
+    },
+    {
+      "epoch": 0.16509348884747188,
+      "grad_norm": 2.0705585709478598,
+      "learning_rate": 3.8117183790379277e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 28555450,
+      "step": 1373
+    },
+    {
+      "epoch": 0.165213731738111,
+      "grad_norm": 3.5330293605607874,
+      "learning_rate": 3.811388285633976e-06,
+      "loss": 0.938,
+      "num_input_tokens_seen": 28571155,
+      "step": 1374
+    },
+    {
+      "epoch": 0.16533397462875007,
+      "grad_norm": 2.2344139474291125,
+      "learning_rate": 3.811057917443861e-06,
+      "loss": 0.6152,
+      "num_input_tokens_seen": 28590140,
+      "step": 1375
+    },
+    {
+      "epoch": 0.16545421751938916,
+      "grad_norm": 0.8818225596152021,
+      "learning_rate": 3.8107272745177e-06,
+      "loss": 0.7024,
+      "num_input_tokens_seen": 28662190,
+      "step": 1376
+    },
+    {
+      "epoch": 0.16557446041002827,
+      "grad_norm": 2.0117484218355006,
+      "learning_rate": 3.8103963569056513e-06,
+      "loss": 0.7924,
+      "num_input_tokens_seen": 28681045,
+      "step": 1377
+    },
+    {
+      "epoch": 0.16569470330066735,
+      "grad_norm": 1.611483272360582,
+      "learning_rate": 3.8100651646579146e-06,
+      "loss": 0.8777,
+      "num_input_tokens_seen": 28699975,
+      "step": 1378
+    },
+    {
+      "epoch": 0.16581494619130643,
+      "grad_norm": 2.218300737493285,
+      "learning_rate": 3.8097336978247317e-06,
+      "loss": 0.9224,
+      "num_input_tokens_seen": 28716400,
+      "step": 1379
+    },
+    {
+      "epoch": 0.16593518908194552,
+      "grad_norm": 2.5779844843578634,
+      "learning_rate": 3.8094019564563854e-06,
+      "loss": 0.8914,
+      "num_input_tokens_seen": 28733050,
+      "step": 1380
+    },
+    {
+      "epoch": 0.16605543197258463,
+      "grad_norm": 3.9080921328117517,
+      "learning_rate": 3.809069940603201e-06,
+      "loss": 0.7602,
+      "num_input_tokens_seen": 28750725,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1661756748632237,
+      "grad_norm": 2.594512634178134,
+      "learning_rate": 3.8087376503155452e-06,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 28767930,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1662959177538628,
+      "grad_norm": 1.00309695066994,
+      "learning_rate": 3.808405085643826e-06,
+      "loss": 0.6039,
+      "num_input_tokens_seen": 28832530,
+      "step": 1383
+    },
+    {
+      "epoch": 0.1664161606445019,
+      "grad_norm": 2.059007914928114,
+      "learning_rate": 3.8080722466384925e-06,
+      "loss": 0.8947,
+      "num_input_tokens_seen": 28850100,
+      "step": 1384
+    },
+    {
+      "epoch": 0.166536403535141,
+      "grad_norm": 2.1208931077949362,
+      "learning_rate": 3.8077391333500376e-06,
+      "loss": 0.6986,
+      "num_input_tokens_seen": 28868960,
+      "step": 1385
+    },
+    {
+      "epoch": 0.16665664642578007,
+      "grad_norm": 1.9186469174437826,
+      "learning_rate": 3.8074057458289934e-06,
+      "loss": 0.765,
+      "num_input_tokens_seen": 28889370,
+      "step": 1386
+    },
+    {
+      "epoch": 0.16677688931641918,
+      "grad_norm": 2.286303493467645,
+      "learning_rate": 3.807072084125934e-06,
+      "loss": 0.8121,
+      "num_input_tokens_seen": 28910940,
+      "step": 1387
+    },
+    {
+      "epoch": 0.16689713220705826,
+      "grad_norm": 2.7432692759711776,
+      "learning_rate": 3.806738148291477e-06,
+      "loss": 0.8034,
+      "num_input_tokens_seen": 28927485,
+      "step": 1388
+    },
+    {
+      "epoch": 0.16701737509769735,
+      "grad_norm": 2.0296425136256553,
+      "learning_rate": 3.806403938376279e-06,
+      "loss": 0.7118,
+      "num_input_tokens_seen": 28949570,
+      "step": 1389
+    },
+    {
+      "epoch": 0.16713761798833643,
+      "grad_norm": 2.254303355473648,
+      "learning_rate": 3.8060694544310396e-06,
+      "loss": 0.7697,
+      "num_input_tokens_seen": 28967800,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16725786087897554,
+      "grad_norm": 2.0867246596688096,
+      "learning_rate": 3.8057346965065006e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 28988750,
+      "step": 1391
+    },
+    {
+      "epoch": 0.16737810376961462,
+      "grad_norm": 1.6522322160282428,
+      "learning_rate": 3.805399664653443e-06,
+      "loss": 0.8461,
+      "num_input_tokens_seen": 29010610,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1674983466602537,
+      "grad_norm": 2.61312289615138,
+      "learning_rate": 3.805064358922692e-06,
+      "loss": 0.7552,
+      "num_input_tokens_seen": 29028620,
+      "step": 1393
+    },
+    {
+      "epoch": 0.16761858955089282,
+      "grad_norm": 2.24430929747971,
+      "learning_rate": 3.8047287793651136e-06,
+      "loss": 0.8024,
+      "num_input_tokens_seen": 29049785,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1677388324415319,
+      "grad_norm": 2.4872347511290394,
+      "learning_rate": 3.8043929260316133e-06,
+      "loss": 0.8857,
+      "num_input_tokens_seen": 29067660,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16785907533217098,
+      "grad_norm": 1.974567189010661,
+      "learning_rate": 3.8040567989731417e-06,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 29085325,
+      "step": 1396
+    },
+    {
+      "epoch": 0.16797931822281006,
+      "grad_norm": 4.516959561007454,
+      "learning_rate": 3.8037203982406876e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 29103210,
+      "step": 1397
+    },
+    {
+      "epoch": 0.16809956111344918,
+      "grad_norm": 2.321208463719716,
+      "learning_rate": 3.8033837238852835e-06,
+      "loss": 0.7282,
+      "num_input_tokens_seen": 29119630,
+      "step": 1398
+    },
+    {
+      "epoch": 0.16821980400408826,
+      "grad_norm": 1.9948011370396532,
+      "learning_rate": 3.8030467759580017e-06,
+      "loss": 0.6946,
+      "num_input_tokens_seen": 29140270,
+      "step": 1399
+    },
+    {
+      "epoch": 0.16834004689472734,
+      "grad_norm": 2.55117562694287,
+      "learning_rate": 3.802709554509958e-06,
+      "loss": 0.8709,
+      "num_input_tokens_seen": 29157790,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16846028978536645,
+      "grad_norm": 1.9504766751299423,
+      "learning_rate": 3.8023720595923083e-06,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 29176765,
+      "step": 1401
+    },
+    {
+      "epoch": 0.16858053267600553,
+      "grad_norm": 2.4182892059139087,
+      "learning_rate": 3.80203429125625e-06,
+      "loss": 0.8736,
+      "num_input_tokens_seen": 29194660,
+      "step": 1402
+    },
+    {
+      "epoch": 0.16870077556664462,
+      "grad_norm": 1.9877660607277075,
+      "learning_rate": 3.8016962495530225e-06,
+      "loss": 0.7031,
+      "num_input_tokens_seen": 29213570,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1688210184572837,
+      "grad_norm": 3.720472325750388,
+      "learning_rate": 3.8013579345339063e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 29228155,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1689412613479228,
+      "grad_norm": 1.9936002083985014,
+      "learning_rate": 3.801019346250224e-06,
+      "loss": 0.6983,
+      "num_input_tokens_seen": 29248020,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1690615042385619,
+      "grad_norm": 2.6521733438559085,
+      "learning_rate": 3.8006804847533395e-06,
+      "loss": 0.8325,
+      "num_input_tokens_seen": 29267255,
+      "step": 1406
+    },
+    {
+      "epoch": 0.16918174712920098,
+      "grad_norm": 2.111681775362625,
+      "learning_rate": 3.8003413500946556e-06,
+      "loss": 0.8454,
+      "num_input_tokens_seen": 29287085,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1693019900198401,
+      "grad_norm": 3.3929131761768496,
+      "learning_rate": 3.8000019423256216e-06,
+      "loss": 0.8302,
+      "num_input_tokens_seen": 29304570,
+      "step": 1408
+    },
+    {
+      "epoch": 0.16942223291047917,
+      "grad_norm": 2.173272743604148,
+      "learning_rate": 3.7996622614977234e-06,
+      "loss": 0.877,
+      "num_input_tokens_seen": 29325480,
+      "step": 1409
+    },
+    {
+      "epoch": 0.16954247580111825,
+      "grad_norm": 2.1476726294793718,
+      "learning_rate": 3.799322307662492e-06,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 29343020,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16966271869175734,
+      "grad_norm": 2.754964428139544,
+      "learning_rate": 3.798982080871496e-06,
+      "loss": 0.8396,
+      "num_input_tokens_seen": 29357880,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16978296158239645,
+      "grad_norm": 2.0320749070398714,
+      "learning_rate": 3.798641581176349e-06,
+      "loss": 0.6807,
+      "num_input_tokens_seen": 29379880,
+      "step": 1412
+    },
+    {
+      "epoch": 0.16990320447303553,
+      "grad_norm": 2.0666430109772405,
+      "learning_rate": 3.7983008086287044e-06,
+      "loss": 0.7441,
+      "num_input_tokens_seen": 29400920,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1700234473636746,
+      "grad_norm": 2.4848960376377316,
+      "learning_rate": 3.797959763280257e-06,
+      "loss": 0.8016,
+      "num_input_tokens_seen": 29419325,
+      "step": 1414
+    },
+    {
+      "epoch": 0.17014369025431372,
+      "grad_norm": 2.3103150987471985,
+      "learning_rate": 3.797618445182743e-06,
+      "loss": 0.7874,
+      "num_input_tokens_seen": 29440440,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1702639331449528,
+      "grad_norm": 3.26892625245738,
+      "learning_rate": 3.79727685438794e-06,
+      "loss": 0.8457,
+      "num_input_tokens_seen": 29454350,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1703841760355919,
+      "grad_norm": 0.895684263423864,
+      "learning_rate": 3.796934990947667e-06,
+      "loss": 0.6347,
+      "num_input_tokens_seen": 29515755,
+      "step": 1417
+    },
+    {
+      "epoch": 0.170504418926231,
+      "grad_norm": 0.9889567298394493,
+      "learning_rate": 3.7965928549137854e-06,
+      "loss": 0.6544,
+      "num_input_tokens_seen": 29572290,
+      "step": 1418
+    },
+    {
+      "epoch": 0.17062466181687008,
+      "grad_norm": 2.1182777479787775,
+      "learning_rate": 3.7962504463381953e-06,
+      "loss": 0.769,
+      "num_input_tokens_seen": 29593500,
+      "step": 1419
+    },
+    {
+      "epoch": 0.17074490470750917,
+      "grad_norm": 1.9667559456931794,
+      "learning_rate": 3.7959077652728412e-06,
+      "loss": 0.7832,
+      "num_input_tokens_seen": 29611675,
+      "step": 1420
+    },
+    {
+      "epoch": 0.17086514759814825,
+      "grad_norm": 2.0891537355877974,
+      "learning_rate": 3.795564811769707e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 29629750,
+      "step": 1421
+    },
+    {
+      "epoch": 0.17098539048878736,
+      "grad_norm": 2.1678915640683454,
+      "learning_rate": 3.795221585880818e-06,
+      "loss": 0.7817,
+      "num_input_tokens_seen": 29650150,
+      "step": 1422
+    },
+    {
+      "epoch": 0.17110563337942644,
+      "grad_norm": 1.76719599723931,
+      "learning_rate": 3.794878087658242e-06,
+      "loss": 0.912,
+      "num_input_tokens_seen": 29667640,
+      "step": 1423
+    },
+    {
+      "epoch": 0.17122587627006552,
+      "grad_norm": 3.1538638824249245,
+      "learning_rate": 3.7945343171540873e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 29688235,
+      "step": 1424
+    },
+    {
+      "epoch": 0.17134611916070464,
+      "grad_norm": 2.234589477674962,
+      "learning_rate": 3.7941902744205033e-06,
+      "loss": 0.7888,
+      "num_input_tokens_seen": 29708990,
+      "step": 1425
+    },
+    {
+      "epoch": 0.17146636205134372,
+      "grad_norm": 2.213027658667225,
+      "learning_rate": 3.7938459595096817e-06,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 29727255,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1715866049419828,
+      "grad_norm": 4.949509936531085,
+      "learning_rate": 3.7935013724738545e-06,
+      "loss": 0.8586,
+      "num_input_tokens_seen": 29747475,
+      "step": 1427
+    },
+    {
+      "epoch": 0.17170684783262188,
+      "grad_norm": 1.9819250305976224,
+      "learning_rate": 3.7931565133652945e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 29767270,
+      "step": 1428
+    },
+    {
+      "epoch": 0.171827090723261,
+      "grad_norm": 3.10670289614236,
+      "learning_rate": 3.792811382236317e-06,
+      "loss": 0.6751,
+      "num_input_tokens_seen": 29785500,
+      "step": 1429
+    },
+    {
+      "epoch": 0.17194733361390008,
+      "grad_norm": 4.828839008089451,
+      "learning_rate": 3.792465979139279e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 29807825,
+      "step": 1430
+    },
+    {
+      "epoch": 0.17206757650453916,
+      "grad_norm": 1.0843571136356336,
+      "learning_rate": 3.792120304126576e-06,
+      "loss": 0.733,
+      "num_input_tokens_seen": 29870920,
+      "step": 1431
+    },
+    {
+      "epoch": 0.17218781939517827,
+      "grad_norm": 1.892758435013736,
+      "learning_rate": 3.791774357250649e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 29889470,
+      "step": 1432
+    },
+    {
+      "epoch": 0.17230806228581735,
+      "grad_norm": 2.384290095308587,
+      "learning_rate": 3.7914281385639757e-06,
+      "loss": 0.791,
+      "num_input_tokens_seen": 29907065,
+      "step": 1433
+    },
+    {
+      "epoch": 0.17242830517645644,
+      "grad_norm": 1.8831846089619875,
+      "learning_rate": 3.7910816481190784e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 29926600,
+      "step": 1434
+    },
+    {
+      "epoch": 0.17254854806709552,
+      "grad_norm": 1.8458086399803142,
+      "learning_rate": 3.7907348859685193e-06,
+      "loss": 0.7479,
+      "num_input_tokens_seen": 29948025,
+      "step": 1435
+    },
+    {
+      "epoch": 0.17266879095773463,
+      "grad_norm": 2.6620837148082814,
+      "learning_rate": 3.790387852164902e-06,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 29968475,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1727890338483737,
+      "grad_norm": 2.242297848365326,
+      "learning_rate": 3.7900405467608707e-06,
+      "loss": 0.7739,
+      "num_input_tokens_seen": 29987740,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1729092767390128,
+      "grad_norm": 3.2953358127133536,
+      "learning_rate": 3.7896929698091114e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 30000275,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1730295196296519,
+      "grad_norm": 5.49889199371149,
+      "learning_rate": 3.7893451213623518e-06,
+      "loss": 0.6952,
+      "num_input_tokens_seen": 30017225,
+      "step": 1439
+    },
+    {
+      "epoch": 0.173149762520291,
+      "grad_norm": 2.2572037389059427,
+      "learning_rate": 3.7889970014733606e-06,
+      "loss": 0.8153,
+      "num_input_tokens_seen": 30036050,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17327000541093007,
+      "grad_norm": 1.7027231033888086,
+      "learning_rate": 3.7886486101949463e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 30056950,
+      "step": 1441
+    },
+    {
+      "epoch": 0.17339024830156918,
+      "grad_norm": 2.36995084415686,
+      "learning_rate": 3.7882999475799594e-06,
+      "loss": 0.8704,
+      "num_input_tokens_seen": 30074705,
+      "step": 1442
+    },
+    {
+      "epoch": 0.17351049119220827,
+      "grad_norm": 2.0609998726922147,
+      "learning_rate": 3.787951013681293e-06,
+      "loss": 0.8061,
+      "num_input_tokens_seen": 30092470,
+      "step": 1443
+    },
+    {
+      "epoch": 0.17363073408284735,
+      "grad_norm": 1.9826865241432492,
+      "learning_rate": 3.787601808551879e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 30112005,
+      "step": 1444
+    },
+    {
+      "epoch": 0.17375097697348643,
+      "grad_norm": 2.5125451356639577,
+      "learning_rate": 3.7872523322446926e-06,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 30130610,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17387121986412554,
+      "grad_norm": 2.5177695768721593,
+      "learning_rate": 3.7869025848127478e-06,
+      "loss": 0.6008,
+      "num_input_tokens_seen": 30154525,
+      "step": 1446
+    },
+    {
+      "epoch": 0.17399146275476463,
+      "grad_norm": 2.5390598202422927,
+      "learning_rate": 3.7865525663091018e-06,
+      "loss": 0.8061,
+      "num_input_tokens_seen": 30172455,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1741117056454037,
+      "grad_norm": 2.3620137624981137,
+      "learning_rate": 3.7862022767868517e-06,
+      "loss": 0.865,
+      "num_input_tokens_seen": 30189765,
+      "step": 1448
+    },
+    {
+      "epoch": 0.17423194853604282,
+      "grad_norm": 1.914694058538217,
+      "learning_rate": 3.7858517162991367e-06,
+      "loss": 0.8463,
+      "num_input_tokens_seen": 30209560,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1743521914266819,
+      "grad_norm": 3.4358997147705153,
+      "learning_rate": 3.7855008848991363e-06,
+      "loss": 0.6043,
+      "num_input_tokens_seen": 30227485,
+      "step": 1450
+    },
+    {
+      "epoch": 0.17447243431732098,
+      "grad_norm": 2.0538998885345703,
+      "learning_rate": 3.7851497826400714e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 30247345,
+      "step": 1451
+    },
+    {
+      "epoch": 0.17459267720796007,
+      "grad_norm": 2.200192153615084,
+      "learning_rate": 3.7847984095752034e-06,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 30270520,
+      "step": 1452
+    },
+    {
+      "epoch": 0.17471292009859918,
+      "grad_norm": 2.0587147703885846,
+      "learning_rate": 3.784446765757836e-06,
+      "loss": 0.8058,
+      "num_input_tokens_seen": 30288885,
+      "step": 1453
+    },
+    {
+      "epoch": 0.17483316298923826,
+      "grad_norm": 2.1658740572742463,
+      "learning_rate": 3.7840948512413133e-06,
+      "loss": 0.7723,
+      "num_input_tokens_seen": 30306190,
+      "step": 1454
+    },
+    {
+      "epoch": 0.17495340587987734,
+      "grad_norm": 2.2002313052141362,
+      "learning_rate": 3.7837426660790196e-06,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 30327325,
+      "step": 1455
+    },
+    {
+      "epoch": 0.17507364877051645,
+      "grad_norm": 2.4941086292106673,
+      "learning_rate": 3.783390210324382e-06,
+      "loss": 0.8161,
+      "num_input_tokens_seen": 30346770,
+      "step": 1456
+    },
+    {
+      "epoch": 0.17519389166115554,
+      "grad_norm": 2.5452965406995958,
+      "learning_rate": 3.7830374840308676e-06,
+      "loss": 0.7281,
+      "num_input_tokens_seen": 30366645,
+      "step": 1457
+    },
+    {
+      "epoch": 0.17531413455179462,
+      "grad_norm": 2.560689283840395,
+      "learning_rate": 3.7826844872519842e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 30384220,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1754343774424337,
+      "grad_norm": 2.074247891424234,
+      "learning_rate": 3.782331220041282e-06,
+      "loss": 0.7278,
+      "num_input_tokens_seen": 30404005,
+      "step": 1459
+    },
+    {
+      "epoch": 0.17555462033307281,
+      "grad_norm": 3.1996771160907143,
+      "learning_rate": 3.7819776824523504e-06,
+      "loss": 0.82,
+      "num_input_tokens_seen": 30421590,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1756748632237119,
+      "grad_norm": 3.0689989716256143,
+      "learning_rate": 3.7816238745388213e-06,
+      "loss": 0.8375,
+      "num_input_tokens_seen": 30440855,
+      "step": 1461
+    },
+    {
+      "epoch": 0.17579510611435098,
+      "grad_norm": 2.0200199989520735,
+      "learning_rate": 3.781269796354367e-06,
+      "loss": 0.8758,
+      "num_input_tokens_seen": 30460195,
+      "step": 1462
+    },
+    {
+      "epoch": 0.1759153490049901,
+      "grad_norm": 2.435885321014921,
+      "learning_rate": 3.7809154479527006e-06,
+      "loss": 0.8689,
+      "num_input_tokens_seen": 30479120,
+      "step": 1463
+    },
+    {
+      "epoch": 0.17603559189562917,
+      "grad_norm": 2.6333516128504497,
+      "learning_rate": 3.780560829387577e-06,
+      "loss": 0.83,
+      "num_input_tokens_seen": 30497340,
+      "step": 1464
+    },
+    {
+      "epoch": 0.17615583478626826,
+      "grad_norm": 0.9006979444395685,
+      "learning_rate": 3.7802059407127915e-06,
+      "loss": 0.6157,
+      "num_input_tokens_seen": 30555610,
+      "step": 1465
+    },
+    {
+      "epoch": 0.17627607767690734,
+      "grad_norm": 2.443150298852987,
+      "learning_rate": 3.7798507819821797e-06,
+      "loss": 0.8597,
+      "num_input_tokens_seen": 30572455,
+      "step": 1466
+    },
+    {
+      "epoch": 0.17639632056754645,
+      "grad_norm": 2.4955983915757836,
+      "learning_rate": 3.7794953532496197e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 30588080,
+      "step": 1467
+    },
+    {
+      "epoch": 0.17651656345818553,
+      "grad_norm": 2.3345481875803054,
+      "learning_rate": 3.7791396545690295e-06,
+      "loss": 0.6404,
+      "num_input_tokens_seen": 30649035,
+      "step": 1468
+    },
+    {
+      "epoch": 0.17663680634882462,
+      "grad_norm": 2.0690896275679527,
+      "learning_rate": 3.7787836859943685e-06,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 30667480,
+      "step": 1469
+    },
+    {
+      "epoch": 0.17675704923946373,
+      "grad_norm": 2.49706896908787,
+      "learning_rate": 3.7784274475796363e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 30685830,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1768772921301028,
+      "grad_norm": 2.427799774441815,
+      "learning_rate": 3.7780709393788745e-06,
+      "loss": 0.7697,
+      "num_input_tokens_seen": 30706025,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1769975350207419,
+      "grad_norm": 2.2401360563273083,
+      "learning_rate": 3.7777141614461647e-06,
+      "loss": 0.7484,
+      "num_input_tokens_seen": 30725450,
+      "step": 1472
+    },
+    {
+      "epoch": 0.177117777911381,
+      "grad_norm": 2.5582807268762053,
+      "learning_rate": 3.7773571138356304e-06,
+      "loss": 0.6875,
+      "num_input_tokens_seen": 30745340,
+      "step": 1473
+    },
+    {
+      "epoch": 0.17723802080202009,
+      "grad_norm": 2.3214489684297677,
+      "learning_rate": 3.776999796601435e-06,
+      "loss": 0.8938,
+      "num_input_tokens_seen": 30763820,
+      "step": 1474
+    },
+    {
+      "epoch": 0.17735826369265917,
+      "grad_norm": 2.0717794265190355,
+      "learning_rate": 3.776642209797783e-06,
+      "loss": 0.7166,
+      "num_input_tokens_seen": 30785370,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17747850658329825,
+      "grad_norm": 3.816360025997515,
+      "learning_rate": 3.7762843534789205e-06,
+      "loss": 0.7768,
+      "num_input_tokens_seen": 30803840,
+      "step": 1476
+    },
+    {
+      "epoch": 0.17759874947393736,
+      "grad_norm": 2.5972658467495053,
+      "learning_rate": 3.7759262276991343e-06,
+      "loss": 0.8728,
+      "num_input_tokens_seen": 30821170,
+      "step": 1477
+    },
+    {
+      "epoch": 0.17771899236457644,
+      "grad_norm": 2.421399770959549,
+      "learning_rate": 3.7755678325127506e-06,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 30838570,
+      "step": 1478
+    },
+    {
+      "epoch": 0.17783923525521553,
+      "grad_norm": 1.856629464823198,
+      "learning_rate": 3.7752091679741393e-06,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 30856080,
+      "step": 1479
+    },
+    {
+      "epoch": 0.17795947814585464,
+      "grad_norm": 3.0543108103926246,
+      "learning_rate": 3.774850234137708e-06,
+      "loss": 0.7765,
+      "num_input_tokens_seen": 30873095,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17807972103649372,
+      "grad_norm": 2.96939523560821,
+      "learning_rate": 3.7744910310579076e-06,
+      "loss": 0.8255,
+      "num_input_tokens_seen": 30891740,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1781999639271328,
+      "grad_norm": 2.8133808228960855,
+      "learning_rate": 3.774131558789229e-06,
+      "loss": 0.848,
+      "num_input_tokens_seen": 30910790,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1783202068177719,
+      "grad_norm": 3.394090806278892,
+      "learning_rate": 3.773771817386203e-06,
+      "loss": 0.6914,
+      "num_input_tokens_seen": 30927840,
+      "step": 1483
+    },
+    {
+      "epoch": 0.178440449708411,
+      "grad_norm": 3.9206975838561537,
+      "learning_rate": 3.773411806903403e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 30946640,
+      "step": 1484
+    },
+    {
+      "epoch": 0.17856069259905008,
+      "grad_norm": 2.1575725985502188,
+      "learning_rate": 3.7730515273954415e-06,
+      "loss": 0.9333,
+      "num_input_tokens_seen": 30964970,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17868093548968916,
+      "grad_norm": 3.038165178394511,
+      "learning_rate": 3.772690978916973e-06,
+      "loss": 0.8353,
+      "num_input_tokens_seen": 30984445,
+      "step": 1486
+    },
+    {
+      "epoch": 0.17880117838032827,
+      "grad_norm": 3.5700927006787975,
+      "learning_rate": 3.772330161522693e-06,
+      "loss": 0.8741,
+      "num_input_tokens_seen": 31002075,
+      "step": 1487
+    },
+    {
+      "epoch": 0.17892142127096736,
+      "grad_norm": 2.47945334178659,
+      "learning_rate": 3.7719690752673365e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 31022590,
+      "step": 1488
+    },
+    {
+      "epoch": 0.17904166416160644,
+      "grad_norm": 2.7410631767727565,
+      "learning_rate": 3.7716077202056796e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 31040785,
+      "step": 1489
+    },
+    {
+      "epoch": 0.17916190705224552,
+      "grad_norm": 4.627595195793667,
+      "learning_rate": 3.7712460963925404e-06,
+      "loss": 0.9243,
+      "num_input_tokens_seen": 31056445,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17928214994288463,
+      "grad_norm": 2.0240293846166386,
+      "learning_rate": 3.7708842038827775e-06,
+      "loss": 0.7474,
+      "num_input_tokens_seen": 31075125,
+      "step": 1491
+    },
+    {
+      "epoch": 0.17940239283352372,
+      "grad_norm": 1.8725844728970533,
+      "learning_rate": 3.770522042731288e-06,
+      "loss": 0.8479,
+      "num_input_tokens_seen": 31096740,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1795226357241628,
+      "grad_norm": 2.4820110686434935,
+      "learning_rate": 3.7701596129930122e-06,
+      "loss": 0.8761,
+      "num_input_tokens_seen": 31115185,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1796428786148019,
+      "grad_norm": 2.3434476543630147,
+      "learning_rate": 3.7697969147229315e-06,
+      "loss": 0.7254,
+      "num_input_tokens_seen": 31133065,
+      "step": 1494
+    },
+    {
+      "epoch": 0.179763121505441,
+      "grad_norm": 2.231594758620423,
+      "learning_rate": 3.7694339479760647e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 31151815,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17988336439608008,
+      "grad_norm": 0.817100011807163,
+      "learning_rate": 3.769070712807476e-06,
+      "loss": 0.6057,
+      "num_input_tokens_seen": 31213565,
+      "step": 1496
+    },
+    {
+      "epoch": 0.18000360728671919,
+      "grad_norm": 1.8113649945854873,
+      "learning_rate": 3.768707209272266e-06,
+      "loss": 0.7844,
+      "num_input_tokens_seen": 31233415,
+      "step": 1497
+    },
+    {
+      "epoch": 0.18012385017735827,
+      "grad_norm": 4.42906657152381,
+      "learning_rate": 3.768343437425579e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 31251705,
+      "step": 1498
+    },
+    {
+      "epoch": 0.18024409306799735,
+      "grad_norm": 2.3559640083407154,
+      "learning_rate": 3.7679793973225987e-06,
+      "loss": 0.8584,
+      "num_input_tokens_seen": 31267235,
+      "step": 1499
+    },
+    {
+      "epoch": 0.18036433595863643,
+      "grad_norm": 0.9283757103183825,
+      "learning_rate": 3.767615089018549e-06,
+      "loss": 0.6409,
+      "num_input_tokens_seen": 31329300,
+      "step": 1500
+    },
+    {
+      "epoch": 0.18048457884927555,
+      "grad_norm": 1.983195447459231,
+      "learning_rate": 3.7672505125686966e-06,
+      "loss": 0.85,
+      "num_input_tokens_seen": 31345385,
+      "step": 1501
+    },
+    {
+      "epoch": 0.18060482173991463,
+      "grad_norm": 2.9594970320730645,
+      "learning_rate": 3.7668856680283455e-06,
+      "loss": 0.863,
+      "num_input_tokens_seen": 31362130,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1807250646305537,
+      "grad_norm": 2.2318074820686102,
+      "learning_rate": 3.7665205554528437e-06,
+      "loss": 0.8214,
+      "num_input_tokens_seen": 31381205,
+      "step": 1503
+    },
+    {
+      "epoch": 0.18084530752119282,
+      "grad_norm": 2.361084715862096,
+      "learning_rate": 3.7661551748975782e-06,
+      "loss": 0.7509,
+      "num_input_tokens_seen": 31399100,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1809655504118319,
+      "grad_norm": 0.8497449757372831,
+      "learning_rate": 3.7657895264179772e-06,
+      "loss": 0.6147,
+      "num_input_tokens_seen": 31454795,
+      "step": 1505
+    },
+    {
+      "epoch": 0.181085793302471,
+      "grad_norm": 2.657925669348378,
+      "learning_rate": 3.765423610069509e-06,
+      "loss": 0.7475,
+      "num_input_tokens_seen": 31479905,
+      "step": 1506
+    },
+    {
+      "epoch": 0.18120603619311007,
+      "grad_norm": 2.7126668062439423,
+      "learning_rate": 3.765057425907683e-06,
+      "loss": 0.7256,
+      "num_input_tokens_seen": 31501085,
+      "step": 1507
+    },
+    {
+      "epoch": 0.18132627908374918,
+      "grad_norm": 2.211402235578531,
+      "learning_rate": 3.764690973988048e-06,
+      "loss": 0.78,
+      "num_input_tokens_seen": 31521145,
+      "step": 1508
+    },
+    {
+      "epoch": 0.18144652197438826,
+      "grad_norm": 3.0580549500636818,
+      "learning_rate": 3.7643242543661963e-06,
+      "loss": 0.7427,
+      "num_input_tokens_seen": 31543525,
+      "step": 1509
+    },
+    {
+      "epoch": 0.18156676486502735,
+      "grad_norm": 0.8741238408541412,
+      "learning_rate": 3.7639572670977573e-06,
+      "loss": 0.6402,
+      "num_input_tokens_seen": 31598740,
+      "step": 1510
+    },
+    {
+      "epoch": 0.18168700775566646,
+      "grad_norm": 1.6119891590307953,
+      "learning_rate": 3.7635900122384042e-06,
+      "loss": 0.7635,
+      "num_input_tokens_seen": 31621455,
+      "step": 1511
+    },
+    {
+      "epoch": 0.18180725064630554,
+      "grad_norm": 2.562713905486325,
+      "learning_rate": 3.7632224898438477e-06,
+      "loss": 0.865,
+      "num_input_tokens_seen": 31637650,
+      "step": 1512
+    },
+    {
+      "epoch": 0.18192749353694462,
+      "grad_norm": 1.900212068018805,
+      "learning_rate": 3.762854699969842e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 31657880,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1820477364275837,
+      "grad_norm": 2.0324910335456874,
+      "learning_rate": 3.762486642672179e-06,
+      "loss": 0.7216,
+      "num_input_tokens_seen": 31674540,
+      "step": 1514
+    },
+    {
+      "epoch": 0.18216797931822282,
+      "grad_norm": 2.8958883369781443,
+      "learning_rate": 3.7621183180066946e-06,
+      "loss": 0.8647,
+      "num_input_tokens_seen": 31692220,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1822882222088619,
+      "grad_norm": 1.6972364002689149,
+      "learning_rate": 3.7617497260292625e-06,
+      "loss": 0.7358,
+      "num_input_tokens_seen": 31713995,
+      "step": 1516
+    },
+    {
+      "epoch": 0.18240846509950098,
+      "grad_norm": 3.5413388547440685,
+      "learning_rate": 3.7613808667957967e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 31726405,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1825287079901401,
+      "grad_norm": 2.4730921681730065,
+      "learning_rate": 3.7610117403622547e-06,
+      "loss": 0.9085,
+      "num_input_tokens_seen": 31742685,
+      "step": 1518
+    },
+    {
+      "epoch": 0.18264895088077918,
+      "grad_norm": 1.9939354510108345,
+      "learning_rate": 3.7606423467846313e-06,
+      "loss": 0.897,
+      "num_input_tokens_seen": 31762010,
+      "step": 1519
+    },
+    {
+      "epoch": 0.18276919377141826,
+      "grad_norm": 1.541853824738568,
+      "learning_rate": 3.760272686118964e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 31779950,
+      "step": 1520
+    },
+    {
+      "epoch": 0.18288943666205737,
+      "grad_norm": 1.9722612599950973,
+      "learning_rate": 3.7599027584213297e-06,
+      "loss": 0.9158,
+      "num_input_tokens_seen": 31798550,
+      "step": 1521
+    },
+    {
+      "epoch": 0.18300967955269645,
+      "grad_norm": 2.3414316911970934,
+      "learning_rate": 3.7595325637478465e-06,
+      "loss": 0.774,
+      "num_input_tokens_seen": 31816295,
+      "step": 1522
+    },
+    {
+      "epoch": 0.18312992244333554,
+      "grad_norm": 1.8754532727653939,
+      "learning_rate": 3.7591621021546723e-06,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 31838010,
+      "step": 1523
+    },
+    {
+      "epoch": 0.18325016533397462,
+      "grad_norm": 2.3308493616537382,
+      "learning_rate": 3.7587913736980062e-06,
+      "loss": 0.8117,
+      "num_input_tokens_seen": 31857370,
+      "step": 1524
+    },
+    {
+      "epoch": 0.18337040822461373,
+      "grad_norm": 2.0447281442595626,
+      "learning_rate": 3.7584203784340865e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 31876260,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1834906511152528,
+      "grad_norm": 2.9268539782783476,
+      "learning_rate": 3.7580491164191938e-06,
+      "loss": 0.8485,
+      "num_input_tokens_seen": 31894290,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1836108940058919,
+      "grad_norm": 0.7961140701586151,
+      "learning_rate": 3.757677587709648e-06,
+      "loss": 0.6337,
+      "num_input_tokens_seen": 31957275,
+      "step": 1527
+    },
+    {
+      "epoch": 0.183731136896531,
+      "grad_norm": 2.538589584797089,
+      "learning_rate": 3.7573057923618095e-06,
+      "loss": 0.7563,
+      "num_input_tokens_seen": 31977090,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1838513797871701,
+      "grad_norm": 2.486400045525193,
+      "learning_rate": 3.7569337304320793e-06,
+      "loss": 0.7415,
+      "num_input_tokens_seen": 31996395,
+      "step": 1529
+    },
+    {
+      "epoch": 0.18397162267780917,
+      "grad_norm": 0.8670439149400267,
+      "learning_rate": 3.756561401976899e-06,
+      "loss": 0.6564,
+      "num_input_tokens_seen": 32055820,
+      "step": 1530
+    },
+    {
+      "epoch": 0.18409186556844825,
+      "grad_norm": 2.2571183394544994,
+      "learning_rate": 3.7561888070527514e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 32077580,
+      "step": 1531
+    },
+    {
+      "epoch": 0.18421210845908736,
+      "grad_norm": 2.324072075766497,
+      "learning_rate": 3.7558159457161577e-06,
+      "loss": 0.7989,
+      "num_input_tokens_seen": 32095265,
+      "step": 1532
+    },
+    {
+      "epoch": 0.18433235134972645,
+      "grad_norm": 2.5874483375876918,
+      "learning_rate": 3.755442818023681e-06,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 32114610,
+      "step": 1533
+    },
+    {
+      "epoch": 0.18445259424036553,
+      "grad_norm": 2.1318883676994735,
+      "learning_rate": 3.7550694240319246e-06,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 32132205,
+      "step": 1534
+    },
+    {
+      "epoch": 0.18457283713100464,
+      "grad_norm": 2.292800877515025,
+      "learning_rate": 3.7546957637975326e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 32149335,
+      "step": 1535
+    },
+    {
+      "epoch": 0.18469308002164372,
+      "grad_norm": 1.676362903220582,
+      "learning_rate": 3.7543218373771873e-06,
+      "loss": 0.7427,
+      "num_input_tokens_seen": 32168380,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1848133229122828,
+      "grad_norm": 1.3722554624399752,
+      "learning_rate": 3.7539476448276145e-06,
+      "loss": 0.779,
+      "num_input_tokens_seen": 32191560,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1849335658029219,
+      "grad_norm": 0.8277266087378189,
+      "learning_rate": 3.753573186205579e-06,
+      "loss": 0.5996,
+      "num_input_tokens_seen": 32259400,
+      "step": 1538
+    },
+    {
+      "epoch": 0.185053808693561,
+      "grad_norm": 2.5593223668236327,
+      "learning_rate": 3.753198461567885e-06,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 32276365,
+      "step": 1539
+    },
+    {
+      "epoch": 0.18517405158420008,
+      "grad_norm": 1.8610055825487877,
+      "learning_rate": 3.7528234709713783e-06,
+      "loss": 0.9165,
+      "num_input_tokens_seen": 32298830,
+      "step": 1540
+    },
+    {
+      "epoch": 0.18529429447483917,
+      "grad_norm": 2.1749754840799818,
+      "learning_rate": 3.7524482144729447e-06,
+      "loss": 0.8385,
+      "num_input_tokens_seen": 32318005,
+      "step": 1541
+    },
+    {
+      "epoch": 0.18541453736547828,
+      "grad_norm": 2.2872127328970913,
+      "learning_rate": 3.7520726921295106e-06,
+      "loss": 0.8395,
+      "num_input_tokens_seen": 32334445,
+      "step": 1542
+    },
+    {
+      "epoch": 0.18553478025611736,
+      "grad_norm": 1.981363897263376,
+      "learning_rate": 3.751696903998042e-06,
+      "loss": 0.7258,
+      "num_input_tokens_seen": 32352800,
+      "step": 1543
+    },
+    {
+      "epoch": 0.18565502314675644,
+      "grad_norm": 1.699634041629413,
+      "learning_rate": 3.7513208501355456e-06,
+      "loss": 0.6974,
+      "num_input_tokens_seen": 32373625,
+      "step": 1544
+    },
+    {
+      "epoch": 0.18577526603739553,
+      "grad_norm": 1.969548087294321,
+      "learning_rate": 3.750944530599069e-06,
+      "loss": 0.8388,
+      "num_input_tokens_seen": 32392915,
+      "step": 1545
+    },
+    {
+      "epoch": 0.18589550892803464,
+      "grad_norm": 2.5525588910784704,
+      "learning_rate": 3.7505679454456992e-06,
+      "loss": 0.8125,
+      "num_input_tokens_seen": 32409245,
+      "step": 1546
+    },
+    {
+      "epoch": 0.18601575181867372,
+      "grad_norm": 2.39695210265176,
+      "learning_rate": 3.750191094732564e-06,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 32429830,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1861359947093128,
+      "grad_norm": 2.601566560511127,
+      "learning_rate": 3.749813978516831e-06,
+      "loss": 0.7416,
+      "num_input_tokens_seen": 32450155,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1862562375999519,
+      "grad_norm": 1.8289719031512919,
+      "learning_rate": 3.749436596855709e-06,
+      "loss": 0.7796,
+      "num_input_tokens_seen": 32469175,
+      "step": 1549
+    },
+    {
+      "epoch": 0.186376480490591,
+      "grad_norm": 1.8658897214878047,
+      "learning_rate": 3.749058949806446e-06,
+      "loss": 0.9114,
+      "num_input_tokens_seen": 32485620,
+      "step": 1550
+    },
+    {
+      "epoch": 0.18649672338123008,
+      "grad_norm": 1.9598513917463747,
+      "learning_rate": 3.748681037426331e-06,
+      "loss": 0.833,
+      "num_input_tokens_seen": 32504550,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1866169662718692,
+      "grad_norm": 2.3075000493749407,
+      "learning_rate": 3.748302859772693e-06,
+      "loss": 0.913,
+      "num_input_tokens_seen": 32521040,
+      "step": 1552
+    },
+    {
+      "epoch": 0.18673720916250827,
+      "grad_norm": 2.673151895887735,
+      "learning_rate": 3.7479244169029017e-06,
+      "loss": 0.619,
+      "num_input_tokens_seen": 32540550,
+      "step": 1553
+    },
+    {
+      "epoch": 0.18685745205314735,
+      "grad_norm": 3.721390958786071,
+      "learning_rate": 3.7475457088743658e-06,
+      "loss": 0.7414,
+      "num_input_tokens_seen": 32557520,
+      "step": 1554
+    },
+    {
+      "epoch": 0.18697769494378644,
+      "grad_norm": 3.6636415545223953,
+      "learning_rate": 3.7471667357445348e-06,
+      "loss": 0.7419,
+      "num_input_tokens_seen": 32577070,
+      "step": 1555
+    },
+    {
+      "epoch": 0.18709793783442555,
+      "grad_norm": 2.137534855952005,
+      "learning_rate": 3.7467874975709003e-06,
+      "loss": 0.7225,
+      "num_input_tokens_seen": 32597595,
+      "step": 1556
+    },
+    {
+      "epoch": 0.18721818072506463,
+      "grad_norm": 2.8209618866363457,
+      "learning_rate": 3.7464079944109904e-06,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 32619175,
+      "step": 1557
+    },
+    {
+      "epoch": 0.18733842361570371,
+      "grad_norm": 2.42780704004583,
+      "learning_rate": 3.746028226322376e-06,
+      "loss": 0.7748,
+      "num_input_tokens_seen": 32634775,
+      "step": 1558
+    },
+    {
+      "epoch": 0.18745866650634282,
+      "grad_norm": 2.1734248595822696,
+      "learning_rate": 3.7456481933626686e-06,
+      "loss": 0.7576,
+      "num_input_tokens_seen": 32653850,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1875789093969819,
+      "grad_norm": 3.9478239557000108,
+      "learning_rate": 3.745267895589518e-06,
+      "loss": 0.7294,
+      "num_input_tokens_seen": 32672110,
+      "step": 1560
+    },
+    {
+      "epoch": 0.187699152287621,
+      "grad_norm": 1.9869960281018462,
+      "learning_rate": 3.7448873330606154e-06,
+      "loss": 0.8117,
+      "num_input_tokens_seen": 32689600,
+      "step": 1561
+    },
+    {
+      "epoch": 0.18781939517826007,
+      "grad_norm": 2.5533182962354406,
+      "learning_rate": 3.7445065058336914e-06,
+      "loss": 0.8758,
+      "num_input_tokens_seen": 32708190,
+      "step": 1562
+    },
+    {
+      "epoch": 0.18793963806889918,
+      "grad_norm": 2.0350589320206147,
+      "learning_rate": 3.7441254139665176e-06,
+      "loss": 0.8614,
+      "num_input_tokens_seen": 32724095,
+      "step": 1563
+    },
+    {
+      "epoch": 0.18805988095953827,
+      "grad_norm": 2.07771725044608,
+      "learning_rate": 3.743744057516905e-06,
+      "loss": 0.8237,
+      "num_input_tokens_seen": 32741875,
+      "step": 1564
+    },
+    {
+      "epoch": 0.18818012385017735,
+      "grad_norm": 3.5048968201694155,
+      "learning_rate": 3.743362436542706e-06,
+      "loss": 0.8811,
+      "num_input_tokens_seen": 32756285,
+      "step": 1565
+    },
+    {
+      "epoch": 0.18830036674081646,
+      "grad_norm": 2.218619394655088,
+      "learning_rate": 3.7429805511018115e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 32777665,
+      "step": 1566
+    },
+    {
+      "epoch": 0.18842060963145554,
+      "grad_norm": 1.8716675557089433,
+      "learning_rate": 3.742598401252153e-06,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 32797585,
+      "step": 1567
+    },
+    {
+      "epoch": 0.18854085252209463,
+      "grad_norm": 0.7568653270311523,
+      "learning_rate": 3.7422159870517025e-06,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 32862560,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1886610954127337,
+      "grad_norm": 1.6156463420438107,
+      "learning_rate": 3.7418333085584717e-06,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 32883465,
+      "step": 1569
+    },
+    {
+      "epoch": 0.18878133830337282,
+      "grad_norm": 2.020256887929179,
+      "learning_rate": 3.7414503658305128e-06,
+      "loss": 0.914,
+      "num_input_tokens_seen": 32900420,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1889015811940119,
+      "grad_norm": 6.8800146059166,
+      "learning_rate": 3.7410671589259185e-06,
+      "loss": 0.773,
+      "num_input_tokens_seen": 32918740,
+      "step": 1571
+    },
+    {
+      "epoch": 0.18902182408465099,
+      "grad_norm": 1.8617271403653202,
+      "learning_rate": 3.7406836879028205e-06,
+      "loss": 0.7906,
+      "num_input_tokens_seen": 32938685,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1891420669752901,
+      "grad_norm": 2.1303595084970004,
+      "learning_rate": 3.7402999528193907e-06,
+      "loss": 0.7732,
+      "num_input_tokens_seen": 32957905,
+      "step": 1573
+    },
+    {
+      "epoch": 0.18926230986592918,
+      "grad_norm": 4.408069293320368,
+      "learning_rate": 3.739915953733842e-06,
+      "loss": 0.8498,
+      "num_input_tokens_seen": 32975670,
+      "step": 1574
+    },
+    {
+      "epoch": 0.18938255275656826,
+      "grad_norm": 3.2798886078506246,
+      "learning_rate": 3.7395316907044264e-06,
+      "loss": 0.8096,
+      "num_input_tokens_seen": 32996175,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18950279564720737,
+      "grad_norm": 1.8304855598548924,
+      "learning_rate": 3.7391471637894364e-06,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 33018160,
+      "step": 1576
+    },
+    {
+      "epoch": 0.18962303853784646,
+      "grad_norm": 5.080045571426592,
+      "learning_rate": 3.7387623730472046e-06,
+      "loss": 0.8523,
+      "num_input_tokens_seen": 33037800,
+      "step": 1577
+    },
+    {
+      "epoch": 0.18974328142848554,
+      "grad_norm": 1.657702952235094,
+      "learning_rate": 3.738377318536103e-06,
+      "loss": 0.8264,
+      "num_input_tokens_seen": 33057405,
+      "step": 1578
+    },
+    {
+      "epoch": 0.18986352431912462,
+      "grad_norm": 2.556705142139931,
+      "learning_rate": 3.7379920003145447e-06,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 33071400,
+      "step": 1579
+    },
+    {
+      "epoch": 0.18998376720976373,
+      "grad_norm": 3.5880439538801547,
+      "learning_rate": 3.7376064184409817e-06,
+      "loss": 0.8379,
+      "num_input_tokens_seen": 33090700,
+      "step": 1580
+    },
+    {
+      "epoch": 0.19010401010040281,
+      "grad_norm": 1.747660119237241,
+      "learning_rate": 3.7372205729739063e-06,
+      "loss": 0.8632,
+      "num_input_tokens_seen": 33112235,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1902242529910419,
+      "grad_norm": 2.2904560904065843,
+      "learning_rate": 3.7368344639718514e-06,
+      "loss": 0.7161,
+      "num_input_tokens_seen": 33129890,
+      "step": 1582
+    },
+    {
+      "epoch": 0.190344495881681,
+      "grad_norm": 1.6127607413479657,
+      "learning_rate": 3.7364480914933895e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 33149850,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1904647387723201,
+      "grad_norm": 2.594618153342351,
+      "learning_rate": 3.7360614555971325e-06,
+      "loss": 0.8129,
+      "num_input_tokens_seen": 33169225,
+      "step": 1584
+    },
+    {
+      "epoch": 0.19058498166295917,
+      "grad_norm": 2.4926917021728574,
+      "learning_rate": 3.735674556341733e-06,
+      "loss": 0.8523,
+      "num_input_tokens_seen": 33188560,
+      "step": 1585
+    },
+    {
+      "epoch": 0.19070522455359826,
+      "grad_norm": 2.3769816338372762,
+      "learning_rate": 3.7352873937858835e-06,
+      "loss": 0.8316,
+      "num_input_tokens_seen": 33209815,
+      "step": 1586
+    },
+    {
+      "epoch": 0.19082546744423737,
+      "grad_norm": 2.150547181117884,
+      "learning_rate": 3.7348999679883155e-06,
+      "loss": 0.717,
+      "num_input_tokens_seen": 33227715,
+      "step": 1587
+    },
+    {
+      "epoch": 0.19094571033487645,
+      "grad_norm": 1.9412753327084251,
+      "learning_rate": 3.7345122790078026e-06,
+      "loss": 0.8399,
+      "num_input_tokens_seen": 33245000,
+      "step": 1588
+    },
+    {
+      "epoch": 0.19106595322551553,
+      "grad_norm": 2.31855575847196,
+      "learning_rate": 3.7341243269031556e-06,
+      "loss": 0.9352,
+      "num_input_tokens_seen": 33263710,
+      "step": 1589
+    },
+    {
+      "epoch": 0.19118619611615464,
+      "grad_norm": 1.9440815032483427,
+      "learning_rate": 3.7337361117332275e-06,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 33285170,
+      "step": 1590
+    },
+    {
+      "epoch": 0.19130643900679373,
+      "grad_norm": 2.0384982790293864,
+      "learning_rate": 3.7333476335569087e-06,
+      "loss": 0.7621,
+      "num_input_tokens_seen": 33302890,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1914266818974328,
+      "grad_norm": 2.499944301921766,
+      "learning_rate": 3.7329588924331325e-06,
+      "loss": 0.6631,
+      "num_input_tokens_seen": 33323815,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1915469247880719,
+      "grad_norm": 1.8056664987383,
+      "learning_rate": 3.732569888420871e-06,
+      "loss": 0.8186,
+      "num_input_tokens_seen": 33343070,
+      "step": 1593
+    },
+    {
+      "epoch": 0.191667167678711,
+      "grad_norm": 2.727366731428961,
+      "learning_rate": 3.732180621579134e-06,
+      "loss": 0.8272,
+      "num_input_tokens_seen": 33362005,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1917874105693501,
+      "grad_norm": 2.4255683870081017,
+      "learning_rate": 3.731791091966974e-06,
+      "loss": 0.8092,
+      "num_input_tokens_seen": 33382920,
+      "step": 1595
+    },
+    {
+      "epoch": 0.19190765345998917,
+      "grad_norm": 2.680562863715639,
+      "learning_rate": 3.7314012996434826e-06,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 33401370,
+      "step": 1596
+    },
+    {
+      "epoch": 0.19202789635062828,
+      "grad_norm": 1.9848195910386617,
+      "learning_rate": 3.7310112446677907e-06,
+      "loss": 0.8062,
+      "num_input_tokens_seen": 33419000,
+      "step": 1597
+    },
+    {
+      "epoch": 0.19214813924126736,
+      "grad_norm": 2.7456396578817506,
+      "learning_rate": 3.7306209270990695e-06,
+      "loss": 0.6846,
+      "num_input_tokens_seen": 33436725,
+      "step": 1598
+    },
+    {
+      "epoch": 0.19226838213190645,
+      "grad_norm": 3.2930485429403435,
+      "learning_rate": 3.730230346996529e-06,
+      "loss": 0.8623,
+      "num_input_tokens_seen": 33455985,
+      "step": 1599
+    },
+    {
+      "epoch": 0.19238862502254553,
+      "grad_norm": 11.831647017186912,
+      "learning_rate": 3.7298395044194206e-06,
+      "loss": 0.7112,
+      "num_input_tokens_seen": 33474515,
+      "step": 1600
+    },
+    {
+      "epoch": 0.19250886791318464,
+      "grad_norm": 2.237600805877356,
+      "learning_rate": 3.7294483994270356e-06,
+      "loss": 0.9328,
+      "num_input_tokens_seen": 33492560,
+      "step": 1601
+    },
+    {
+      "epoch": 0.19262911080382372,
+      "grad_norm": 2.1683943949036655,
+      "learning_rate": 3.7290570320787033e-06,
+      "loss": 0.7714,
+      "num_input_tokens_seen": 33511860,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1927493536944628,
+      "grad_norm": 1.99225558453428,
+      "learning_rate": 3.728665402433793e-06,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 33530150,
+      "step": 1603
+    },
+    {
+      "epoch": 0.19286959658510192,
+      "grad_norm": 3.7084352496584523,
+      "learning_rate": 3.7282735105517164e-06,
+      "loss": 0.8591,
+      "num_input_tokens_seen": 33547995,
+      "step": 1604
+    },
+    {
+      "epoch": 0.192989839475741,
+      "grad_norm": 3.0591177755805847,
+      "learning_rate": 3.727881356491922e-06,
+      "loss": 0.6855,
+      "num_input_tokens_seen": 33566125,
+      "step": 1605
+    },
+    {
+      "epoch": 0.19311008236638008,
+      "grad_norm": 2.484198078876229,
+      "learning_rate": 3.7274889403139002e-06,
+      "loss": 0.7461,
+      "num_input_tokens_seen": 33583470,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1932303252570192,
+      "grad_norm": 2.55931531512891,
+      "learning_rate": 3.727096262077179e-06,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 33602185,
+      "step": 1607
+    },
+    {
+      "epoch": 0.19335056814765827,
+      "grad_norm": 2.1256388651417333,
+      "learning_rate": 3.7267033218413285e-06,
+      "loss": 0.8565,
+      "num_input_tokens_seen": 33619700,
+      "step": 1608
+    },
+    {
+      "epoch": 0.19347081103829736,
+      "grad_norm": 2.1351915308048968,
+      "learning_rate": 3.726310119665957e-06,
+      "loss": 0.8084,
+      "num_input_tokens_seen": 33635755,
+      "step": 1609
+    },
+    {
+      "epoch": 0.19359105392893644,
+      "grad_norm": 1.7371156523740492,
+      "learning_rate": 3.725916655610713e-06,
+      "loss": 0.8533,
+      "num_input_tokens_seen": 33654805,
+      "step": 1610
+    },
+    {
+      "epoch": 0.19371129681957555,
+      "grad_norm": 3.3087696257322485,
+      "learning_rate": 3.725522929735284e-06,
+      "loss": 0.754,
+      "num_input_tokens_seen": 33671460,
+      "step": 1611
+    },
+    {
+      "epoch": 0.19383153971021463,
+      "grad_norm": 15.445245229828124,
+      "learning_rate": 3.725128942099399e-06,
+      "loss": 0.7426,
+      "num_input_tokens_seen": 33691580,
+      "step": 1612
+    },
+    {
+      "epoch": 0.19395178260085372,
+      "grad_norm": 1.7887488418049415,
+      "learning_rate": 3.7247346927628245e-06,
+      "loss": 0.8037,
+      "num_input_tokens_seen": 33711235,
+      "step": 1613
+    },
+    {
+      "epoch": 0.19407202549149283,
+      "grad_norm": 2.262937169285758,
+      "learning_rate": 3.7243401817853694e-06,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 33731645,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1941922683821319,
+      "grad_norm": 1.8610910742503932,
+      "learning_rate": 3.723945409226879e-06,
+      "loss": 0.729,
+      "num_input_tokens_seen": 33749855,
+      "step": 1615
+    },
+    {
+      "epoch": 0.194312511272771,
+      "grad_norm": 2.278757183184574,
+      "learning_rate": 3.723550375147241e-06,
+      "loss": 0.8015,
+      "num_input_tokens_seen": 33764350,
+      "step": 1616
+    },
+    {
+      "epoch": 0.19443275416341008,
+      "grad_norm": 2.312731662091948,
+      "learning_rate": 3.723155079606381e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 33784080,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1945529970540492,
+      "grad_norm": 1.7175504066804508,
+      "learning_rate": 3.722759522664266e-06,
+      "loss": 0.6448,
+      "num_input_tokens_seen": 33801100,
+      "step": 1618
+    },
+    {
+      "epoch": 0.19467323994468827,
+      "grad_norm": 1.8096838385150495,
+      "learning_rate": 3.7223637043809016e-06,
+      "loss": 0.8033,
+      "num_input_tokens_seen": 33819800,
+      "step": 1619
+    },
+    {
+      "epoch": 0.19479348283532735,
+      "grad_norm": 2.0151905275269693,
+      "learning_rate": 3.7219676248163322e-06,
+      "loss": 0.8569,
+      "num_input_tokens_seen": 33836685,
+      "step": 1620
+    },
+    {
+      "epoch": 0.19491372572596646,
+      "grad_norm": 1.993810459595368,
+      "learning_rate": 3.7215712840306428e-06,
+      "loss": 0.9196,
+      "num_input_tokens_seen": 33856215,
+      "step": 1621
+    },
+    {
+      "epoch": 0.19503396861660555,
+      "grad_norm": 2.0166029183667162,
+      "learning_rate": 3.721174682083959e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 33873030,
+      "step": 1622
+    },
+    {
+      "epoch": 0.19515421150724463,
+      "grad_norm": 1.7962274802802138,
+      "learning_rate": 3.7207778190364437e-06,
+      "loss": 0.8138,
+      "num_input_tokens_seen": 33891175,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1952744543978837,
+      "grad_norm": 2.062505413255298,
+      "learning_rate": 3.720380694948302e-06,
+      "loss": 0.7386,
+      "num_input_tokens_seen": 33913780,
+      "step": 1624
+    },
+    {
+      "epoch": 0.19539469728852282,
+      "grad_norm": 1.0062495249915844,
+      "learning_rate": 3.719983309879777e-06,
+      "loss": 0.7392,
+      "num_input_tokens_seen": 33973280,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1955149401791619,
+      "grad_norm": 1.968081626200981,
+      "learning_rate": 3.719585663891151e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 33990535,
+      "step": 1626
+    },
+    {
+      "epoch": 0.195635183069801,
+      "grad_norm": 2.145030036503394,
+      "learning_rate": 3.719187757042747e-06,
+      "loss": 0.7844,
+      "num_input_tokens_seen": 34008075,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1957554259604401,
+      "grad_norm": 0.8194014281998503,
+      "learning_rate": 3.7187895893949275e-06,
+      "loss": 0.5991,
+      "num_input_tokens_seen": 34074265,
+      "step": 1628
+    },
+    {
+      "epoch": 0.19587566885107918,
+      "grad_norm": 2.365389004919559,
+      "learning_rate": 3.7183911610080937e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 34090850,
+      "step": 1629
+    },
+    {
+      "epoch": 0.19599591174171827,
+      "grad_norm": 2.6153362527154163,
+      "learning_rate": 3.7179924719426872e-06,
+      "loss": 0.7498,
+      "num_input_tokens_seen": 34108465,
+      "step": 1630
+    },
+    {
+      "epoch": 0.19611615463235738,
+      "grad_norm": 7.721002012720539,
+      "learning_rate": 3.7175935222591885e-06,
+      "loss": 0.758,
+      "num_input_tokens_seen": 34127485,
+      "step": 1631
+    },
+    {
+      "epoch": 0.19623639752299646,
+      "grad_norm": 1.8716485317669629,
+      "learning_rate": 3.717194312018118e-06,
+      "loss": 0.7481,
+      "num_input_tokens_seen": 34146190,
+      "step": 1632
+    },
+    {
+      "epoch": 0.19635664041363554,
+      "grad_norm": 2.1600212081022145,
+      "learning_rate": 3.716794841280036e-06,
+      "loss": 0.7621,
+      "num_input_tokens_seen": 34164615,
+      "step": 1633
+    },
+    {
+      "epoch": 0.19647688330427462,
+      "grad_norm": 2.632772414073602,
+      "learning_rate": 3.7163951101055407e-06,
+      "loss": 0.7763,
+      "num_input_tokens_seen": 34182395,
+      "step": 1634
+    },
+    {
+      "epoch": 0.19659712619491373,
+      "grad_norm": 1.8967059228638845,
+      "learning_rate": 3.715995118555273e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 34202090,
+      "step": 1635
+    },
+    {
+      "epoch": 0.19671736908555282,
+      "grad_norm": 2.785715465416871,
+      "learning_rate": 3.7155948666899095e-06,
+      "loss": 0.855,
+      "num_input_tokens_seen": 34220670,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1968376119761919,
+      "grad_norm": 1.959296963404671,
+      "learning_rate": 3.715194354570169e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 34240395,
+      "step": 1637
+    },
+    {
+      "epoch": 0.196957854866831,
+      "grad_norm": 2.5559562882847415,
+      "learning_rate": 3.714793582256809e-06,
+      "loss": 0.828,
+      "num_input_tokens_seen": 34257180,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1970780977574701,
+      "grad_norm": 2.6840248804523643,
+      "learning_rate": 3.7143925498106253e-06,
+      "loss": 0.8388,
+      "num_input_tokens_seen": 34275440,
+      "step": 1639
+    },
+    {
+      "epoch": 0.19719834064810918,
+      "grad_norm": 2.040932250147423,
+      "learning_rate": 3.7139912572924558e-06,
+      "loss": 0.7885,
+      "num_input_tokens_seen": 34294190,
+      "step": 1640
+    },
+    {
+      "epoch": 0.19731858353874826,
+      "grad_norm": 2.6976264910270173,
+      "learning_rate": 3.7135897047631744e-06,
+      "loss": 0.8119,
+      "num_input_tokens_seen": 34311795,
+      "step": 1641
+    },
+    {
+      "epoch": 0.19743882642938737,
+      "grad_norm": 2.2230732259828865,
+      "learning_rate": 3.713187892283698e-06,
+      "loss": 0.7511,
+      "num_input_tokens_seen": 34331125,
+      "step": 1642
+    },
+    {
+      "epoch": 0.19755906932002645,
+      "grad_norm": 2.309092914572269,
+      "learning_rate": 3.7127858199149796e-06,
+      "loss": 0.8655,
+      "num_input_tokens_seen": 34346705,
+      "step": 1643
+    },
+    {
+      "epoch": 0.19767931221066554,
+      "grad_norm": 2.1362879798713603,
+      "learning_rate": 3.712383487718015e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 34364665,
+      "step": 1644
+    },
+    {
+      "epoch": 0.19779955510130465,
+      "grad_norm": 1.820599807234752,
+      "learning_rate": 3.7119808957538365e-06,
+      "loss": 0.8645,
+      "num_input_tokens_seen": 34383380,
+      "step": 1645
+    },
+    {
+      "epoch": 0.19791979799194373,
+      "grad_norm": 1.8362251314843832,
+      "learning_rate": 3.711578044083517e-06,
+      "loss": 0.8006,
+      "num_input_tokens_seen": 34399900,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1980400408825828,
+      "grad_norm": 1.85253181699229,
+      "learning_rate": 3.7111749327681694e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 34419655,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1981602837732219,
+      "grad_norm": 2.2160189791189353,
+      "learning_rate": 3.7107715618689455e-06,
+      "loss": 0.862,
+      "num_input_tokens_seen": 34438350,
+      "step": 1648
+    },
+    {
+      "epoch": 0.198280526663861,
+      "grad_norm": 1.431735504671451,
+      "learning_rate": 3.710367931447035e-06,
+      "loss": 0.8286,
+      "num_input_tokens_seen": 34459850,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1984007695545001,
+      "grad_norm": 2.3948659806251933,
+      "learning_rate": 3.7099640415636695e-06,
+      "loss": 0.8616,
+      "num_input_tokens_seen": 34479205,
+      "step": 1650
+    },
+    {
+      "epoch": 0.19852101244513917,
+      "grad_norm": 1.7392946334656993,
+      "learning_rate": 3.7095598922801187e-06,
+      "loss": 0.722,
+      "num_input_tokens_seen": 34501000,
+      "step": 1651
+    },
+    {
+      "epoch": 0.19864125533577828,
+      "grad_norm": 1.993913314530379,
+      "learning_rate": 3.7091554836576914e-06,
+      "loss": 0.7495,
+      "num_input_tokens_seen": 34517395,
+      "step": 1652
+    },
+    {
+      "epoch": 0.19876149822641737,
+      "grad_norm": 1.983106055896671,
+      "learning_rate": 3.708750815757736e-06,
+      "loss": 0.8284,
+      "num_input_tokens_seen": 34537885,
+      "step": 1653
+    },
+    {
+      "epoch": 0.19888174111705645,
+      "grad_norm": 2.492828382722311,
+      "learning_rate": 3.7083458886416407e-06,
+      "loss": 0.7329,
+      "num_input_tokens_seen": 34556800,
+      "step": 1654
+    },
+    {
+      "epoch": 0.19900198400769553,
+      "grad_norm": 2.5221885356656517,
+      "learning_rate": 3.707940702370832e-06,
+      "loss": 0.8774,
+      "num_input_tokens_seen": 34577365,
+      "step": 1655
+    },
+    {
+      "epoch": 0.19912222689833464,
+      "grad_norm": 0.7797848497755522,
+      "learning_rate": 3.707535257006777e-06,
+      "loss": 0.5964,
+      "num_input_tokens_seen": 34642710,
+      "step": 1656
+    },
+    {
+      "epoch": 0.19924246978897373,
+      "grad_norm": 2.2654916235977716,
+      "learning_rate": 3.707129552610981e-06,
+      "loss": 0.8856,
+      "num_input_tokens_seen": 34661080,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1993627126796128,
+      "grad_norm": 1.9222574941794814,
+      "learning_rate": 3.70672358924499e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 34680040,
+      "step": 1658
+    },
+    {
+      "epoch": 0.19948295557025192,
+      "grad_norm": 2.080149965547982,
+      "learning_rate": 3.706317366970386e-06,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 34700760,
+      "step": 1659
+    },
+    {
+      "epoch": 0.199603198460891,
+      "grad_norm": 2.11917690030436,
+      "learning_rate": 3.705910885848795e-06,
+      "loss": 0.8374,
+      "num_input_tokens_seen": 34718855,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19972344135153008,
+      "grad_norm": 1.9858654369525628,
+      "learning_rate": 3.705504145941879e-06,
+      "loss": 0.8424,
+      "num_input_tokens_seen": 34736745,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1998436842421692,
+      "grad_norm": 1.9287523000063362,
+      "learning_rate": 3.7050971473113403e-06,
+      "loss": 0.7866,
+      "num_input_tokens_seen": 34756240,
+      "step": 1662
+    },
+    {
+      "epoch": 0.19996392713280828,
+      "grad_norm": 1.7340725728749529,
+      "learning_rate": 3.7046898900189196e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 34780295,
+      "step": 1663
+    },
+    {
+      "epoch": 0.20008417002344736,
+      "grad_norm": 3.0674734332445617,
+      "learning_rate": 3.704282374126398e-06,
+      "loss": 0.8377,
+      "num_input_tokens_seen": 34799695,
+      "step": 1664
+    },
+    {
+      "epoch": 0.20020441291408644,
+      "grad_norm": 1.6831879873041657,
+      "learning_rate": 3.703874599695595e-06,
+      "loss": 0.8753,
+      "num_input_tokens_seen": 34818760,
+      "step": 1665
+    },
+    {
+      "epoch": 0.20032465580472555,
+      "grad_norm": 2.67746229574641,
+      "learning_rate": 3.703466566788371e-06,
+      "loss": 0.7199,
+      "num_input_tokens_seen": 34837610,
+      "step": 1666
+    },
+    {
+      "epoch": 0.20044489869536464,
+      "grad_norm": 2.040016308116333,
+      "learning_rate": 3.703058275466622e-06,
+      "loss": 0.7426,
+      "num_input_tokens_seen": 34856565,
+      "step": 1667
+    },
+    {
+      "epoch": 0.20056514158600372,
+      "grad_norm": 1.688512910100869,
+      "learning_rate": 3.7026497257922877e-06,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 34876595,
+      "step": 1668
+    },
+    {
+      "epoch": 0.20068538447664283,
+      "grad_norm": 1.619946769116153,
+      "learning_rate": 3.7022409178273436e-06,
+      "loss": 0.8427,
+      "num_input_tokens_seen": 34897295,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2008056273672819,
+      "grad_norm": 1.7765693512836875,
+      "learning_rate": 3.7018318516338054e-06,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 34916175,
+      "step": 1670
+    },
+    {
+      "epoch": 0.200925870257921,
+      "grad_norm": 2.190236093410836,
+      "learning_rate": 3.7014225272737284e-06,
+      "loss": 0.8159,
+      "num_input_tokens_seen": 34935120,
+      "step": 1671
+    },
+    {
+      "epoch": 0.20104611314856008,
+      "grad_norm": 2.3242927807428377,
+      "learning_rate": 3.7010129448092067e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 34951955,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2011663560391992,
+      "grad_norm": 1.9562228173520317,
+      "learning_rate": 3.700603104302374e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 34971485,
+      "step": 1673
+    },
+    {
+      "epoch": 0.20128659892983827,
+      "grad_norm": 0.8988801713496993,
+      "learning_rate": 3.7001930058154027e-06,
+      "loss": 0.583,
+      "num_input_tokens_seen": 35036165,
+      "step": 1674
+    },
+    {
+      "epoch": 0.20140684182047736,
+      "grad_norm": 2.7150023880566065,
+      "learning_rate": 3.6997826494105037e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 35056330,
+      "step": 1675
+    },
+    {
+      "epoch": 0.20152708471111647,
+      "grad_norm": 2.1370815749037835,
+      "learning_rate": 3.6993720351499286e-06,
+      "loss": 0.697,
+      "num_input_tokens_seen": 35077175,
+      "step": 1676
+    },
+    {
+      "epoch": 0.20164732760175555,
+      "grad_norm": 1.9898782038150444,
+      "learning_rate": 3.6989611630959666e-06,
+      "loss": 0.7743,
+      "num_input_tokens_seen": 35095450,
+      "step": 1677
+    },
+    {
+      "epoch": 0.20176757049239463,
+      "grad_norm": 0.9131897228849142,
+      "learning_rate": 3.6985500333109474e-06,
+      "loss": 0.6156,
+      "num_input_tokens_seen": 35163500,
+      "step": 1678
+    },
+    {
+      "epoch": 0.20188781338303372,
+      "grad_norm": 2.425556843619448,
+      "learning_rate": 3.6981386458572385e-06,
+      "loss": 0.7614,
+      "num_input_tokens_seen": 35181195,
+      "step": 1679
+    },
+    {
+      "epoch": 0.20200805627367283,
+      "grad_norm": 2.9925600203932308,
+      "learning_rate": 3.6977270007972468e-06,
+      "loss": 0.7666,
+      "num_input_tokens_seen": 35198450,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2021282991643119,
+      "grad_norm": 5.495708211916301,
+      "learning_rate": 3.6973150981934196e-06,
+      "loss": 0.7163,
+      "num_input_tokens_seen": 35219400,
+      "step": 1681
+    },
+    {
+      "epoch": 0.202248542054951,
+      "grad_norm": 2.638097405345924,
+      "learning_rate": 3.6969029381082415e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 35235115,
+      "step": 1682
+    },
+    {
+      "epoch": 0.2023687849455901,
+      "grad_norm": 1.8448858296220345,
+      "learning_rate": 3.6964905206042365e-06,
+      "loss": 0.7978,
+      "num_input_tokens_seen": 35253525,
+      "step": 1683
+    },
+    {
+      "epoch": 0.20248902783622919,
+      "grad_norm": 1.817013406531175,
+      "learning_rate": 3.696077845743968e-06,
+      "loss": 0.802,
+      "num_input_tokens_seen": 35272835,
+      "step": 1684
+    },
+    {
+      "epoch": 0.20260927072686827,
+      "grad_norm": 2.3667986526461675,
+      "learning_rate": 3.69566491359004e-06,
+      "loss": 0.7302,
+      "num_input_tokens_seen": 35289200,
+      "step": 1685
+    },
+    {
+      "epoch": 0.20272951361750738,
+      "grad_norm": 1.8917840455382113,
+      "learning_rate": 3.695251724205092e-06,
+      "loss": 0.706,
+      "num_input_tokens_seen": 35313280,
+      "step": 1686
+    },
+    {
+      "epoch": 0.20284975650814646,
+      "grad_norm": 1.96974395180062,
+      "learning_rate": 3.6948382776518054e-06,
+      "loss": 0.8509,
+      "num_input_tokens_seen": 35333705,
+      "step": 1687
+    },
+    {
+      "epoch": 0.20296999939878554,
+      "grad_norm": 2.0377540080513974,
+      "learning_rate": 3.6944245739929e-06,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 35349585,
+      "step": 1688
+    },
+    {
+      "epoch": 0.20309024228942463,
+      "grad_norm": 2.0466242480256893,
+      "learning_rate": 3.694010613291133e-06,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 35366490,
+      "step": 1689
+    },
+    {
+      "epoch": 0.20321048518006374,
+      "grad_norm": 1.9592121639298246,
+      "learning_rate": 3.6935963956093037e-06,
+      "loss": 0.8798,
+      "num_input_tokens_seen": 35386295,
+      "step": 1690
+    },
+    {
+      "epoch": 0.20333072807070282,
+      "grad_norm": 1.6865200037692185,
+      "learning_rate": 3.6931819210102474e-06,
+      "loss": 0.6836,
+      "num_input_tokens_seen": 35405410,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2034509709613419,
+      "grad_norm": 2.1367374524467926,
+      "learning_rate": 3.6927671895568402e-06,
+      "loss": 0.8421,
+      "num_input_tokens_seen": 35424190,
+      "step": 1692
+    },
+    {
+      "epoch": 0.20357121385198101,
+      "grad_norm": 1.9639575355473116,
+      "learning_rate": 3.692352201311996e-06,
+      "loss": 0.8704,
+      "num_input_tokens_seen": 35442760,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2036914567426201,
+      "grad_norm": 2.1035098711192806,
+      "learning_rate": 3.6919369563386687e-06,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 35462280,
+      "step": 1694
+    },
+    {
+      "epoch": 0.20381169963325918,
+      "grad_norm": 2.2386580809874848,
+      "learning_rate": 3.69152145469985e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 35479045,
+      "step": 1695
+    },
+    {
+      "epoch": 0.20393194252389826,
+      "grad_norm": 2.5156604878031694,
+      "learning_rate": 3.691105696458572e-06,
+      "loss": 0.8119,
+      "num_input_tokens_seen": 35496060,
+      "step": 1696
+    },
+    {
+      "epoch": 0.20405218541453737,
+      "grad_norm": 4.767966355245232,
+      "learning_rate": 3.690689681677904e-06,
+      "loss": 0.6739,
+      "num_input_tokens_seen": 35514250,
+      "step": 1697
+    },
+    {
+      "epoch": 0.20417242830517646,
+      "grad_norm": 1.88629218774707,
+      "learning_rate": 3.690273410420956e-06,
+      "loss": 0.8777,
+      "num_input_tokens_seen": 35533735,
+      "step": 1698
+    },
+    {
+      "epoch": 0.20429267119581554,
+      "grad_norm": 2.515516650736705,
+      "learning_rate": 3.689856882750875e-06,
+      "loss": 0.7702,
+      "num_input_tokens_seen": 35548655,
+      "step": 1699
+    },
+    {
+      "epoch": 0.20441291408645465,
+      "grad_norm": 1.696633191333543,
+      "learning_rate": 3.6894400987308486e-06,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 35565895,
+      "step": 1700
+    },
+    {
+      "epoch": 0.20453315697709373,
+      "grad_norm": 2.621583851450548,
+      "learning_rate": 3.6890230584241024e-06,
+      "loss": 0.85,
+      "num_input_tokens_seen": 35582545,
+      "step": 1701
+    },
+    {
+      "epoch": 0.20465339986773282,
+      "grad_norm": 0.9313295863936389,
+      "learning_rate": 3.6886057618939016e-06,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 35645085,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2047736427583719,
+      "grad_norm": 2.166878862803526,
+      "learning_rate": 3.6881882092035492e-06,
+      "loss": 0.6989,
+      "num_input_tokens_seen": 35666190,
+      "step": 1703
+    },
+    {
+      "epoch": 0.204893885649011,
+      "grad_norm": 1.155973037261191,
+      "learning_rate": 3.6877704004163873e-06,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 35726315,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2050141285396501,
+      "grad_norm": 2.209310694153699,
+      "learning_rate": 3.687352335595798e-06,
+      "loss": 0.7788,
+      "num_input_tokens_seen": 35745035,
+      "step": 1705
+    },
+    {
+      "epoch": 0.20513437143028918,
+      "grad_norm": 1.0940077982506569,
+      "learning_rate": 3.686934014805201e-06,
+      "loss": 0.7194,
+      "num_input_tokens_seen": 35795385,
+      "step": 1706
+    },
+    {
+      "epoch": 0.20525461432092829,
+      "grad_norm": 2.9484250928601945,
+      "learning_rate": 3.6865154381080552e-06,
+      "loss": 0.8058,
+      "num_input_tokens_seen": 35815790,
+      "step": 1707
+    },
+    {
+      "epoch": 0.20537485721156737,
+      "grad_norm": 3.0038241820001104,
+      "learning_rate": 3.6860966055678585e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 35831865,
+      "step": 1708
+    },
+    {
+      "epoch": 0.20549510010220645,
+      "grad_norm": 1.7218405523027227,
+      "learning_rate": 3.685677517248147e-06,
+      "loss": 0.8546,
+      "num_input_tokens_seen": 35850475,
+      "step": 1709
+    },
+    {
+      "epoch": 0.20561534299284553,
+      "grad_norm": 1.8806139356743412,
+      "learning_rate": 3.6852581732124967e-06,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 35867540,
+      "step": 1710
+    },
+    {
+      "epoch": 0.20573558588348465,
+      "grad_norm": 2.285485242152226,
+      "learning_rate": 3.6848385735245213e-06,
+      "loss": 0.7523,
+      "num_input_tokens_seen": 35886350,
+      "step": 1711
+    },
+    {
+      "epoch": 0.20585582877412373,
+      "grad_norm": 1.779203810836933,
+      "learning_rate": 3.6844187182478734e-06,
+      "loss": 0.8597,
+      "num_input_tokens_seen": 35906925,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2059760716647628,
+      "grad_norm": 1.7618907288865344,
+      "learning_rate": 3.683998607446246e-06,
+      "loss": 0.7518,
+      "num_input_tokens_seen": 35925295,
+      "step": 1713
+    },
+    {
+      "epoch": 0.20609631455540192,
+      "grad_norm": 2.122945836973529,
+      "learning_rate": 3.6835782411833686e-06,
+      "loss": 0.742,
+      "num_input_tokens_seen": 35944535,
+      "step": 1714
+    },
+    {
+      "epoch": 0.206216557446041,
+      "grad_norm": 1.8805161860115227,
+      "learning_rate": 3.68315761952301e-06,
+      "loss": 0.7445,
+      "num_input_tokens_seen": 35961485,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2063368003366801,
+      "grad_norm": 2.031283104399946,
+      "learning_rate": 3.6827367425289797e-06,
+      "loss": 0.8286,
+      "num_input_tokens_seen": 35980980,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2064570432273192,
+      "grad_norm": 2.621724774069149,
+      "learning_rate": 3.6823156102651225e-06,
+      "loss": 0.7222,
+      "num_input_tokens_seen": 35998855,
+      "step": 1717
+    },
+    {
+      "epoch": 0.20657728611795828,
+      "grad_norm": 1.870451445254049,
+      "learning_rate": 3.6818942227953257e-06,
+      "loss": 0.7155,
+      "num_input_tokens_seen": 36019120,
+      "step": 1718
+    },
+    {
+      "epoch": 0.20669752900859736,
+      "grad_norm": 2.202148694033474,
+      "learning_rate": 3.681472580183512e-06,
+      "loss": 0.6899,
+      "num_input_tokens_seen": 36037490,
+      "step": 1719
+    },
+    {
+      "epoch": 0.20681777189923645,
+      "grad_norm": 2.2518439241820363,
+      "learning_rate": 3.6810506824936455e-06,
+      "loss": 0.86,
+      "num_input_tokens_seen": 36055290,
+      "step": 1720
+    },
+    {
+      "epoch": 0.20693801478987556,
+      "grad_norm": 1.1709678226815556,
+      "learning_rate": 3.680628529789726e-06,
+      "loss": 0.6612,
+      "num_input_tokens_seen": 36107420,
+      "step": 1721
+    },
+    {
+      "epoch": 0.20705825768051464,
+      "grad_norm": 1.9557401094809275,
+      "learning_rate": 3.680206122135796e-06,
+      "loss": 0.8575,
+      "num_input_tokens_seen": 36127745,
+      "step": 1722
+    },
+    {
+      "epoch": 0.20717850057115372,
+      "grad_norm": 1.6784370235543111,
+      "learning_rate": 3.6797834595959323e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 36147365,
+      "step": 1723
+    },
+    {
+      "epoch": 0.20729874346179283,
+      "grad_norm": 2.401998375912347,
+      "learning_rate": 3.679360542234254e-06,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 36166430,
+      "step": 1724
+    },
+    {
+      "epoch": 0.20741898635243192,
+      "grad_norm": 1.624799421370761,
+      "learning_rate": 3.678937370114916e-06,
+      "loss": 0.7169,
+      "num_input_tokens_seen": 36185955,
+      "step": 1725
+    },
+    {
+      "epoch": 0.207539229243071,
+      "grad_norm": 1.9325096176461283,
+      "learning_rate": 3.678513943302114e-06,
+      "loss": 0.7833,
+      "num_input_tokens_seen": 36202450,
+      "step": 1726
+    },
+    {
+      "epoch": 0.20765947213371008,
+      "grad_norm": 1.6361413769553386,
+      "learning_rate": 3.6780902618600816e-06,
+      "loss": 0.8473,
+      "num_input_tokens_seen": 36221900,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2077797150243492,
+      "grad_norm": 2.518197382896608,
+      "learning_rate": 3.6776663258530906e-06,
+      "loss": 0.7795,
+      "num_input_tokens_seen": 36240270,
+      "step": 1728
+    },
+    {
+      "epoch": 0.20789995791498828,
+      "grad_norm": 1.8278081803733086,
+      "learning_rate": 3.6772421353454516e-06,
+      "loss": 0.7095,
+      "num_input_tokens_seen": 36258585,
+      "step": 1729
+    },
+    {
+      "epoch": 0.20802020080562736,
+      "grad_norm": 1.901420879136906,
+      "learning_rate": 3.6768176904015153e-06,
+      "loss": 0.8723,
+      "num_input_tokens_seen": 36278110,
+      "step": 1730
+    },
+    {
+      "epoch": 0.20814044369626647,
+      "grad_norm": 1.9760401118248234,
+      "learning_rate": 3.6763929910856674e-06,
+      "loss": 0.5992,
+      "num_input_tokens_seen": 36296280,
+      "step": 1731
+    },
+    {
+      "epoch": 0.20826068658690555,
+      "grad_norm": 2.837331355067723,
+      "learning_rate": 3.6759680374623365e-06,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 36313915,
+      "step": 1732
+    },
+    {
+      "epoch": 0.20838092947754464,
+      "grad_norm": 2.6931341563366087,
+      "learning_rate": 3.675542829595986e-06,
+      "loss": 0.751,
+      "num_input_tokens_seen": 36333300,
+      "step": 1733
+    },
+    {
+      "epoch": 0.20850117236818372,
+      "grad_norm": 1.5158485298869635,
+      "learning_rate": 3.6751173675511213e-06,
+      "loss": 0.7885,
+      "num_input_tokens_seen": 36355065,
+      "step": 1734
+    },
+    {
+      "epoch": 0.20862141525882283,
+      "grad_norm": 2.079115273145643,
+      "learning_rate": 3.674691651392283e-06,
+      "loss": 0.8725,
+      "num_input_tokens_seen": 36372455,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2087416581494619,
+      "grad_norm": 2.1540501285476856,
+      "learning_rate": 3.674265681184053e-06,
+      "loss": 0.7638,
+      "num_input_tokens_seen": 36395435,
+      "step": 1736
+    },
+    {
+      "epoch": 0.208861901040101,
+      "grad_norm": 1.7858104004973228,
+      "learning_rate": 3.6738394569910504e-06,
+      "loss": 0.8643,
+      "num_input_tokens_seen": 36415695,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2089821439307401,
+      "grad_norm": 2.1969425779859773,
+      "learning_rate": 3.6734129788779333e-06,
+      "loss": 0.827,
+      "num_input_tokens_seen": 36434590,
+      "step": 1738
+    },
+    {
+      "epoch": 0.2091023868213792,
+      "grad_norm": 1.7744586532390394,
+      "learning_rate": 3.6729862469093976e-06,
+      "loss": 0.8965,
+      "num_input_tokens_seen": 36453405,
+      "step": 1739
+    },
+    {
+      "epoch": 0.20922262971201827,
+      "grad_norm": 2.2876353711637107,
+      "learning_rate": 3.6725592611501782e-06,
+      "loss": 0.8223,
+      "num_input_tokens_seen": 36471800,
+      "step": 1740
+    },
+    {
+      "epoch": 0.20934287260265738,
+      "grad_norm": 1.7983044959309142,
+      "learning_rate": 3.672132021665049e-06,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 36492135,
+      "step": 1741
+    },
+    {
+      "epoch": 0.20946311549329646,
+      "grad_norm": 2.2044283306396917,
+      "learning_rate": 3.6717045285188215e-06,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 36509550,
+      "step": 1742
+    },
+    {
+      "epoch": 0.20958335838393555,
+      "grad_norm": 2.2578066303733966,
+      "learning_rate": 3.671276781776346e-06,
+      "loss": 0.8527,
+      "num_input_tokens_seen": 36527925,
+      "step": 1743
+    },
+    {
+      "epoch": 0.20970360127457463,
+      "grad_norm": 2.676591364821894,
+      "learning_rate": 3.6708487815025124e-06,
+      "loss": 0.6656,
+      "num_input_tokens_seen": 36548225,
+      "step": 1744
+    },
+    {
+      "epoch": 0.20982384416521374,
+      "grad_norm": 2.0946013542465525,
+      "learning_rate": 3.6704205277622463e-06,
+      "loss": 0.7485,
+      "num_input_tokens_seen": 36566385,
+      "step": 1745
+    },
+    {
+      "epoch": 0.20994408705585282,
+      "grad_norm": 1.665079232889727,
+      "learning_rate": 3.6699920206205146e-06,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 36586845,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2100643299464919,
+      "grad_norm": 1.676139984296906,
+      "learning_rate": 3.669563260142321e-06,
+      "loss": 0.8226,
+      "num_input_tokens_seen": 36605455,
+      "step": 1747
+    },
+    {
+      "epoch": 0.21018457283713102,
+      "grad_norm": 2.4656115651985586,
+      "learning_rate": 3.6691342463927083e-06,
+      "loss": 0.8423,
+      "num_input_tokens_seen": 36624170,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2103048157277701,
+      "grad_norm": 2.165749739334828,
+      "learning_rate": 3.6687049794367574e-06,
+      "loss": 0.8116,
+      "num_input_tokens_seen": 36643985,
+      "step": 1749
+    },
+    {
+      "epoch": 0.21042505861840918,
+      "grad_norm": 1.9593749771324847,
+      "learning_rate": 3.668275459339588e-06,
+      "loss": 0.7849,
+      "num_input_tokens_seen": 36662185,
+      "step": 1750
+    },
+    {
+      "epoch": 0.21054530150904827,
+      "grad_norm": 1.8050103002391567,
+      "learning_rate": 3.667845686166358e-06,
+      "loss": 0.807,
+      "num_input_tokens_seen": 36678830,
+      "step": 1751
+    },
+    {
+      "epoch": 0.21066554439968738,
+      "grad_norm": 1.6123897525637478,
+      "learning_rate": 3.6674156599822634e-06,
+      "loss": 0.8518,
+      "num_input_tokens_seen": 36694345,
+      "step": 1752
+    },
+    {
+      "epoch": 0.21078578729032646,
+      "grad_norm": 5.378396038318722,
+      "learning_rate": 3.666985380852539e-06,
+      "loss": 0.8073,
+      "num_input_tokens_seen": 36713070,
+      "step": 1753
+    },
+    {
+      "epoch": 0.21090603018096554,
+      "grad_norm": 2.6517511505227205,
+      "learning_rate": 3.6665548488424576e-06,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 36731550,
+      "step": 1754
+    },
+    {
+      "epoch": 0.21102627307160465,
+      "grad_norm": 1.762766041219322,
+      "learning_rate": 3.6661240640173307e-06,
+      "loss": 0.8715,
+      "num_input_tokens_seen": 36752740,
+      "step": 1755
+    },
+    {
+      "epoch": 0.21114651596224374,
+      "grad_norm": 0.9534343848868646,
+      "learning_rate": 3.665693026442508e-06,
+      "loss": 0.6392,
+      "num_input_tokens_seen": 36816505,
+      "step": 1756
+    },
+    {
+      "epoch": 0.21126675885288282,
+      "grad_norm": 1.8700503001592552,
+      "learning_rate": 3.665261736183378e-06,
+      "loss": 0.7586,
+      "num_input_tokens_seen": 36836260,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2113870017435219,
+      "grad_norm": 3.9147321659655487,
+      "learning_rate": 3.664830193305366e-06,
+      "loss": 0.8845,
+      "num_input_tokens_seen": 36853755,
+      "step": 1758
+    },
+    {
+      "epoch": 0.211507244634161,
+      "grad_norm": 14.051091440765614,
+      "learning_rate": 3.6643983978739373e-06,
+      "loss": 0.7575,
+      "num_input_tokens_seen": 36870090,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2116274875248001,
+      "grad_norm": 2.024336694811304,
+      "learning_rate": 3.6639663499545958e-06,
+      "loss": 0.8135,
+      "num_input_tokens_seen": 36889990,
+      "step": 1760
+    },
+    {
+      "epoch": 0.21174773041543918,
+      "grad_norm": 0.772616748576978,
+      "learning_rate": 3.6635340496128816e-06,
+      "loss": 0.6151,
+      "num_input_tokens_seen": 36946640,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2118679733060783,
+      "grad_norm": 1.5823800804774302,
+      "learning_rate": 3.6631014969143747e-06,
+      "loss": 0.917,
+      "num_input_tokens_seen": 36966050,
+      "step": 1762
+    },
+    {
+      "epoch": 0.21198821619671737,
+      "grad_norm": 3.036715949957443,
+      "learning_rate": 3.6626686919246925e-06,
+      "loss": 0.8766,
+      "num_input_tokens_seen": 36986820,
+      "step": 1763
+    },
+    {
+      "epoch": 0.21210845908735645,
+      "grad_norm": 2.0885069023325116,
+      "learning_rate": 3.6622356347094927e-06,
+      "loss": 0.7151,
+      "num_input_tokens_seen": 37008105,
+      "step": 1764
+    },
+    {
+      "epoch": 0.21222870197799554,
+      "grad_norm": 2.5122354961197035,
+      "learning_rate": 3.6618023253344684e-06,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 37026685,
+      "step": 1765
+    },
+    {
+      "epoch": 0.21234894486863465,
+      "grad_norm": 1.8589895640069298,
+      "learning_rate": 3.6613687638653522e-06,
+      "loss": 0.8277,
+      "num_input_tokens_seen": 37044575,
+      "step": 1766
+    },
+    {
+      "epoch": 0.21246918775927373,
+      "grad_norm": 1.7594423186919539,
+      "learning_rate": 3.660934950367916e-06,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 37063540,
+      "step": 1767
+    },
+    {
+      "epoch": 0.21258943064991281,
+      "grad_norm": 1.7019040773505065,
+      "learning_rate": 3.660500884907968e-06,
+      "loss": 0.8281,
+      "num_input_tokens_seen": 37084000,
+      "step": 1768
+    },
+    {
+      "epoch": 0.21270967354055192,
+      "grad_norm": 0.8558150428626566,
+      "learning_rate": 3.660066567551356e-06,
+      "loss": 0.625,
+      "num_input_tokens_seen": 37143865,
+      "step": 1769
+    },
+    {
+      "epoch": 0.212829916431191,
+      "grad_norm": 2.3487661698096005,
+      "learning_rate": 3.6596319983639657e-06,
+      "loss": 0.8221,
+      "num_input_tokens_seen": 37162165,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2129501593218301,
+      "grad_norm": 1.5846460651320358,
+      "learning_rate": 3.659197177411721e-06,
+      "loss": 0.855,
+      "num_input_tokens_seen": 37184860,
+      "step": 1771
+    },
+    {
+      "epoch": 0.2130704022124692,
+      "grad_norm": 2.1984613137248874,
+      "learning_rate": 3.6587621047605833e-06,
+      "loss": 0.8099,
+      "num_input_tokens_seen": 37201750,
+      "step": 1772
+    },
+    {
+      "epoch": 0.21319064510310828,
+      "grad_norm": 2.3392578497592074,
+      "learning_rate": 3.6583267804765542e-06,
+      "loss": 0.8674,
+      "num_input_tokens_seen": 37215805,
+      "step": 1773
+    },
+    {
+      "epoch": 0.21331088799374737,
+      "grad_norm": 2.6227986588778824,
+      "learning_rate": 3.65789120462567e-06,
+      "loss": 0.8494,
+      "num_input_tokens_seen": 37234045,
+      "step": 1774
+    },
+    {
+      "epoch": 0.21343113088438645,
+      "grad_norm": 2.361543801373876,
+      "learning_rate": 3.6574553772740083e-06,
+      "loss": 0.7623,
+      "num_input_tokens_seen": 37251695,
+      "step": 1775
+    },
+    {
+      "epoch": 0.21355137377502556,
+      "grad_norm": 0.8780642632998993,
+      "learning_rate": 3.657019298487684e-06,
+      "loss": 0.6593,
+      "num_input_tokens_seen": 37316425,
+      "step": 1776
+    },
+    {
+      "epoch": 0.21367161666566464,
+      "grad_norm": 1.675448065795171,
+      "learning_rate": 3.6565829683328495e-06,
+      "loss": 0.8283,
+      "num_input_tokens_seen": 37338770,
+      "step": 1777
+    },
+    {
+      "epoch": 0.21379185955630373,
+      "grad_norm": 2.0976154727276493,
+      "learning_rate": 3.656146386875696e-06,
+      "loss": 0.8529,
+      "num_input_tokens_seen": 37357190,
+      "step": 1778
+    },
+    {
+      "epoch": 0.21391210244694284,
+      "grad_norm": 2.32082506889008,
+      "learning_rate": 3.6557095541824527e-06,
+      "loss": 0.7725,
+      "num_input_tokens_seen": 37377250,
+      "step": 1779
+    },
+    {
+      "epoch": 0.21403234533758192,
+      "grad_norm": 1.7766199493748616,
+      "learning_rate": 3.6552724703193855e-06,
+      "loss": 0.8421,
+      "num_input_tokens_seen": 37394160,
+      "step": 1780
+    },
+    {
+      "epoch": 0.214152588228221,
+      "grad_norm": 0.7985597271306865,
+      "learning_rate": 3.654835135352801e-06,
+      "loss": 0.5794,
+      "num_input_tokens_seen": 37448690,
+      "step": 1781
+    },
+    {
+      "epoch": 0.21427283111886009,
+      "grad_norm": 1.7653736663925563,
+      "learning_rate": 3.6543975493490424e-06,
+      "loss": 0.8677,
+      "num_input_tokens_seen": 37465785,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2143930740094992,
+      "grad_norm": 1.962107813728839,
+      "learning_rate": 3.653959712374491e-06,
+      "loss": 0.7514,
+      "num_input_tokens_seen": 37483610,
+      "step": 1783
+    },
+    {
+      "epoch": 0.21451331690013828,
+      "grad_norm": 1.6642107113663052,
+      "learning_rate": 3.6535216244955663e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 37503225,
+      "step": 1784
+    },
+    {
+      "epoch": 0.21463355979077736,
+      "grad_norm": 1.8214602724421463,
+      "learning_rate": 3.6530832857787253e-06,
+      "loss": 0.7053,
+      "num_input_tokens_seen": 37524315,
+      "step": 1785
+    },
+    {
+      "epoch": 0.21475380268141647,
+      "grad_norm": 2.176438180297038,
+      "learning_rate": 3.6526446962904653e-06,
+      "loss": 0.7991,
+      "num_input_tokens_seen": 37542750,
+      "step": 1786
+    },
+    {
+      "epoch": 0.21487404557205556,
+      "grad_norm": 1.7588275994460607,
+      "learning_rate": 3.652205856097318e-06,
+      "loss": 0.741,
+      "num_input_tokens_seen": 37565655,
+      "step": 1787
+    },
+    {
+      "epoch": 0.21499428846269464,
+      "grad_norm": 2.1164002231632706,
+      "learning_rate": 3.651766765265856e-06,
+      "loss": 0.7869,
+      "num_input_tokens_seen": 37582385,
+      "step": 1788
+    },
+    {
+      "epoch": 0.21511453135333372,
+      "grad_norm": 2.24463223110277,
+      "learning_rate": 3.65132742386269e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 37597325,
+      "step": 1789
+    },
+    {
+      "epoch": 0.21523477424397283,
+      "grad_norm": 1.715490416839966,
+      "learning_rate": 3.6508878319544656e-06,
+      "loss": 0.844,
+      "num_input_tokens_seen": 37617260,
+      "step": 1790
+    },
+    {
+      "epoch": 0.21535501713461191,
+      "grad_norm": 2.447114568916636,
+      "learning_rate": 3.65044798960787e-06,
+      "loss": 0.8061,
+      "num_input_tokens_seen": 37635320,
+      "step": 1791
+    },
+    {
+      "epoch": 0.215475260025251,
+      "grad_norm": 1.8517650271010826,
+      "learning_rate": 3.650007896889627e-06,
+      "loss": 0.7735,
+      "num_input_tokens_seen": 37653620,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2155955029158901,
+      "grad_norm": 1.8689050831744136,
+      "learning_rate": 3.6495675538664974e-06,
+      "loss": 0.8053,
+      "num_input_tokens_seen": 37672355,
+      "step": 1793
+    },
+    {
+      "epoch": 0.2157157458065292,
+      "grad_norm": 1.6525833162112833,
+      "learning_rate": 3.649126960605282e-06,
+      "loss": 0.8183,
+      "num_input_tokens_seen": 37693060,
+      "step": 1794
+    },
+    {
+      "epoch": 0.21583598869716827,
+      "grad_norm": 2.279537363859075,
+      "learning_rate": 3.6486861171728174e-06,
+      "loss": 0.8315,
+      "num_input_tokens_seen": 37711175,
+      "step": 1795
+    },
+    {
+      "epoch": 0.21595623158780738,
+      "grad_norm": 1.7906862766897846,
+      "learning_rate": 3.64824502363598e-06,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 37732750,
+      "step": 1796
+    },
+    {
+      "epoch": 0.21607647447844647,
+      "grad_norm": 3.01915372612889,
+      "learning_rate": 3.647803680061683e-06,
+      "loss": 0.7701,
+      "num_input_tokens_seen": 37752885,
+      "step": 1797
+    },
+    {
+      "epoch": 0.21619671736908555,
+      "grad_norm": 2.8847506398591807,
+      "learning_rate": 3.6473620865168776e-06,
+      "loss": 0.7598,
+      "num_input_tokens_seen": 37769475,
+      "step": 1798
+    },
+    {
+      "epoch": 0.21631696025972463,
+      "grad_norm": 2.1950726556323614,
+      "learning_rate": 3.646920243068554e-06,
+      "loss": 0.8177,
+      "num_input_tokens_seen": 37787090,
+      "step": 1799
+    },
+    {
+      "epoch": 0.21643720315036374,
+      "grad_norm": 1.6332063099557796,
+      "learning_rate": 3.6464781497837384e-06,
+      "loss": 0.737,
+      "num_input_tokens_seen": 37808785,
+      "step": 1800
+    },
+    {
+      "epoch": 0.21655744604100283,
+      "grad_norm": 1.6423910968289752,
+      "learning_rate": 3.6460358067294965e-06,
+      "loss": 0.7295,
+      "num_input_tokens_seen": 37829735,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2166776889316419,
+      "grad_norm": 1.9761402097314489,
+      "learning_rate": 3.645593213972932e-06,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 37848360,
+      "step": 1802
+    },
+    {
+      "epoch": 0.21679793182228102,
+      "grad_norm": 3.1337263419242958,
+      "learning_rate": 3.6451503715811852e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 37866390,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2169181747129201,
+      "grad_norm": 2.3076125154929867,
+      "learning_rate": 3.6447072796214345e-06,
+      "loss": 0.7987,
+      "num_input_tokens_seen": 37884675,
+      "step": 1804
+    },
+    {
+      "epoch": 0.21703841760355919,
+      "grad_norm": 0.9544705238557111,
+      "learning_rate": 3.644263938160898e-06,
+      "loss": 0.6662,
+      "num_input_tokens_seen": 37940360,
+      "step": 1805
+    },
+    {
+      "epoch": 0.21715866049419827,
+      "grad_norm": 1.9577700934723972,
+      "learning_rate": 3.6438203472668293e-06,
+      "loss": 0.7089,
+      "num_input_tokens_seen": 37959725,
+      "step": 1806
+    },
+    {
+      "epoch": 0.21727890338483738,
+      "grad_norm": 2.0163355174307718,
+      "learning_rate": 3.6433765070065206e-06,
+      "loss": 0.8171,
+      "num_input_tokens_seen": 37977235,
+      "step": 1807
+    },
+    {
+      "epoch": 0.21739914627547646,
+      "grad_norm": 2.1502965419176037,
+      "learning_rate": 3.6429324174473025e-06,
+      "loss": 0.8748,
+      "num_input_tokens_seen": 37990495,
+      "step": 1808
+    },
+    {
+      "epoch": 0.21751938916611555,
+      "grad_norm": 1.9979477753045656,
+      "learning_rate": 3.6424880786565425e-06,
+      "loss": 0.8512,
+      "num_input_tokens_seen": 38006360,
+      "step": 1809
+    },
+    {
+      "epoch": 0.21763963205675466,
+      "grad_norm": 2.098481125344564,
+      "learning_rate": 3.642043490701648e-06,
+      "loss": 0.7949,
+      "num_input_tokens_seen": 38025770,
+      "step": 1810
+    },
+    {
+      "epoch": 0.21775987494739374,
+      "grad_norm": 1.6266061774061915,
+      "learning_rate": 3.6415986536500606e-06,
+      "loss": 0.8127,
+      "num_input_tokens_seen": 38043820,
+      "step": 1811
+    },
+    {
+      "epoch": 0.21788011783803282,
+      "grad_norm": 1.825934314096968,
+      "learning_rate": 3.641153567569263e-06,
+      "loss": 0.8105,
+      "num_input_tokens_seen": 38061855,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2180003607286719,
+      "grad_norm": 2.0152629563030557,
+      "learning_rate": 3.640708232526774e-06,
+      "loss": 0.9426,
+      "num_input_tokens_seen": 38080230,
+      "step": 1813
+    },
+    {
+      "epoch": 0.21812060361931102,
+      "grad_norm": 2.0198996935154465,
+      "learning_rate": 3.6402626485901504e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 38099045,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2182408465099501,
+      "grad_norm": 1.8633676708640283,
+      "learning_rate": 3.639816815826988e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 38118090,
+      "step": 1815
+    },
+    {
+      "epoch": 0.21836108940058918,
+      "grad_norm": 3.03055645298261,
+      "learning_rate": 3.6393707343049176e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 38138140,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2184813322912283,
+      "grad_norm": 2.6308854968969317,
+      "learning_rate": 3.6389244040916104e-06,
+      "loss": 0.7352,
+      "num_input_tokens_seen": 38156935,
+      "step": 1817
+    },
+    {
+      "epoch": 0.21860157518186737,
+      "grad_norm": 1.96231934566628,
+      "learning_rate": 3.6384778252547747e-06,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 38172535,
+      "step": 1818
+    },
+    {
+      "epoch": 0.21872181807250646,
+      "grad_norm": 2.16040105604102,
+      "learning_rate": 3.638030997862155e-06,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 38191190,
+      "step": 1819
+    },
+    {
+      "epoch": 0.21884206096314554,
+      "grad_norm": 0.8546912444371211,
+      "learning_rate": 3.6375839219815356e-06,
+      "loss": 0.6249,
+      "num_input_tokens_seen": 38248710,
+      "step": 1820
+    },
+    {
+      "epoch": 0.21896230385378465,
+      "grad_norm": 2.0572362342852375,
+      "learning_rate": 3.637136597680737e-06,
+      "loss": 0.8227,
+      "num_input_tokens_seen": 38268825,
+      "step": 1821
+    },
+    {
+      "epoch": 0.21908254674442373,
+      "grad_norm": 2.011467520872459,
+      "learning_rate": 3.6366890250276185e-06,
+      "loss": 0.8345,
+      "num_input_tokens_seen": 38289500,
+      "step": 1822
+    },
+    {
+      "epoch": 0.21920278963506282,
+      "grad_norm": 2.1812186305445453,
+      "learning_rate": 3.6362412040900764e-06,
+      "loss": 0.8991,
+      "num_input_tokens_seen": 38309010,
+      "step": 1823
+    },
+    {
+      "epoch": 0.21932303252570193,
+      "grad_norm": 1.9498735884910756,
+      "learning_rate": 3.635793134936044e-06,
+      "loss": 0.7969,
+      "num_input_tokens_seen": 38329740,
+      "step": 1824
+    },
+    {
+      "epoch": 0.219443275416341,
+      "grad_norm": 1.8313355744695072,
+      "learning_rate": 3.635344817633494e-06,
+      "loss": 0.727,
+      "num_input_tokens_seen": 38348775,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2195635183069801,
+      "grad_norm": 2.2279761128753965,
+      "learning_rate": 3.634896252250436e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 38365260,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2196837611976192,
+      "grad_norm": 1.7546277055684447,
+      "learning_rate": 3.6344474388549157e-06,
+      "loss": 0.8228,
+      "num_input_tokens_seen": 38384635,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2198040040882583,
+      "grad_norm": 2.305943209207984,
+      "learning_rate": 3.6339983775150183e-06,
+      "loss": 0.8083,
+      "num_input_tokens_seen": 38400915,
+      "step": 1828
+    },
+    {
+      "epoch": 0.21992424697889737,
+      "grad_norm": 2.4417597642599165,
+      "learning_rate": 3.6335490682988664e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 38416245,
+      "step": 1829
+    },
+    {
+      "epoch": 0.22004448986953645,
+      "grad_norm": 2.1356834544559655,
+      "learning_rate": 3.63309951127462e-06,
+      "loss": 0.8269,
+      "num_input_tokens_seen": 38432875,
+      "step": 1830
+    },
+    {
+      "epoch": 0.22016473276017556,
+      "grad_norm": 2.0113820904986315,
+      "learning_rate": 3.6326497065104757e-06,
+      "loss": 0.7496,
+      "num_input_tokens_seen": 38453060,
+      "step": 1831
+    },
+    {
+      "epoch": 0.22028497565081465,
+      "grad_norm": 2.6705658766359264,
+      "learning_rate": 3.6321996540746693e-06,
+      "loss": 0.7782,
+      "num_input_tokens_seen": 38471855,
+      "step": 1832
+    },
+    {
+      "epoch": 0.22040521854145373,
+      "grad_norm": 1.8164125356603502,
+      "learning_rate": 3.631749354035473e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 38494990,
+      "step": 1833
+    },
+    {
+      "epoch": 0.22052546143209284,
+      "grad_norm": 2.2040712353406158,
+      "learning_rate": 3.6312988064611976e-06,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 38513020,
+      "step": 1834
+    },
+    {
+      "epoch": 0.22064570432273192,
+      "grad_norm": 1.8801019026834391,
+      "learning_rate": 3.6308480114201896e-06,
+      "loss": 0.7994,
+      "num_input_tokens_seen": 38534660,
+      "step": 1835
+    },
+    {
+      "epoch": 0.220765947213371,
+      "grad_norm": 1.7848474606803468,
+      "learning_rate": 3.630396968980835e-06,
+      "loss": 0.7617,
+      "num_input_tokens_seen": 38552255,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2208861901040101,
+      "grad_norm": 2.6160608806691155,
+      "learning_rate": 3.6299456792115575e-06,
+      "loss": 0.844,
+      "num_input_tokens_seen": 38573230,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2210064329946492,
+      "grad_norm": 1.9115691628978044,
+      "learning_rate": 3.629494142180815e-06,
+      "loss": 0.8073,
+      "num_input_tokens_seen": 38591695,
+      "step": 1838
+    },
+    {
+      "epoch": 0.22112667588528828,
+      "grad_norm": 2.2216648717181644,
+      "learning_rate": 3.6290423579571075e-06,
+      "loss": 0.8489,
+      "num_input_tokens_seen": 38607955,
+      "step": 1839
+    },
+    {
+      "epoch": 0.22124691877592736,
+      "grad_norm": 1.631429787222244,
+      "learning_rate": 3.6285903266089694e-06,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 38626950,
+      "step": 1840
+    },
+    {
+      "epoch": 0.22136716166656648,
+      "grad_norm": 1.912494853652078,
+      "learning_rate": 3.628138048204974e-06,
+      "loss": 0.7678,
+      "num_input_tokens_seen": 38647355,
+      "step": 1841
+    },
+    {
+      "epoch": 0.22148740455720556,
+      "grad_norm": 1.9831575817336016,
+      "learning_rate": 3.6276855228137304e-06,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 38665280,
+      "step": 1842
+    },
+    {
+      "epoch": 0.22160764744784464,
+      "grad_norm": 2.3080394577680328,
+      "learning_rate": 3.6272327505038874e-06,
+      "loss": 0.8124,
+      "num_input_tokens_seen": 38681465,
+      "step": 1843
+    },
+    {
+      "epoch": 0.22172789033848372,
+      "grad_norm": 1.7988845554334703,
+      "learning_rate": 3.6267797313441304e-06,
+      "loss": 0.7793,
+      "num_input_tokens_seen": 38700975,
+      "step": 1844
+    },
+    {
+      "epoch": 0.22184813322912283,
+      "grad_norm": 2.5296873394124937,
+      "learning_rate": 3.6263264654031814e-06,
+      "loss": 0.8511,
+      "num_input_tokens_seen": 38717595,
+      "step": 1845
+    },
+    {
+      "epoch": 0.22196837611976192,
+      "grad_norm": 0.7340154602236587,
+      "learning_rate": 3.6258729527498008e-06,
+      "loss": 0.6068,
+      "num_input_tokens_seen": 38778160,
+      "step": 1846
+    },
+    {
+      "epoch": 0.222088619010401,
+      "grad_norm": 2.1475370237623843,
+      "learning_rate": 3.6254191934527854e-06,
+      "loss": 0.6347,
+      "num_input_tokens_seen": 38797235,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2222088619010401,
+      "grad_norm": 1.8191542802363727,
+      "learning_rate": 3.6249651875809715e-06,
+      "loss": 0.6392,
+      "num_input_tokens_seen": 38816835,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2223291047916792,
+      "grad_norm": 1.8674553736196544,
+      "learning_rate": 3.62451093520323e-06,
+      "loss": 0.8902,
+      "num_input_tokens_seen": 38834460,
+      "step": 1849
+    },
+    {
+      "epoch": 0.22244934768231828,
+      "grad_norm": 2.0623193064801244,
+      "learning_rate": 3.6240564363884714e-06,
+      "loss": 0.8996,
+      "num_input_tokens_seen": 38854125,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2225695905729574,
+      "grad_norm": 1.908793957900236,
+      "learning_rate": 3.6236016912056425e-06,
+      "loss": 0.7039,
+      "num_input_tokens_seen": 38872920,
+      "step": 1851
+    },
+    {
+      "epoch": 0.22268983346359647,
+      "grad_norm": 1.9006776964486918,
+      "learning_rate": 3.623146699723729e-06,
+      "loss": 0.8097,
+      "num_input_tokens_seen": 38892100,
+      "step": 1852
+    },
+    {
+      "epoch": 0.22281007635423555,
+      "grad_norm": 1.7176681396304596,
+      "learning_rate": 3.6226914620117507e-06,
+      "loss": 0.78,
+      "num_input_tokens_seen": 38910440,
+      "step": 1853
+    },
+    {
+      "epoch": 0.22293031924487464,
+      "grad_norm": 3.7533601845539586,
+      "learning_rate": 3.622235978138768e-06,
+      "loss": 0.8017,
+      "num_input_tokens_seen": 38927785,
+      "step": 1854
+    },
+    {
+      "epoch": 0.22305056213551375,
+      "grad_norm": 2.50729892045481,
+      "learning_rate": 3.621780248173877e-06,
+      "loss": 0.8164,
+      "num_input_tokens_seen": 38945705,
+      "step": 1855
+    },
+    {
+      "epoch": 0.22317080502615283,
+      "grad_norm": 1.1892572803908956,
+      "learning_rate": 3.6213242721862125e-06,
+      "loss": 0.65,
+      "num_input_tokens_seen": 39003880,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2232910479167919,
+      "grad_norm": 1.9978649573635496,
+      "learning_rate": 3.6208680502449444e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 39024080,
+      "step": 1857
+    },
+    {
+      "epoch": 0.22341129080743102,
+      "grad_norm": 2.5519758207140133,
+      "learning_rate": 3.6204115824192817e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 39041275,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2235315336980701,
+      "grad_norm": 2.221701684658989,
+      "learning_rate": 3.619954868778471e-06,
+      "loss": 0.774,
+      "num_input_tokens_seen": 39057690,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2236517765887092,
+      "grad_norm": 1.9258829886510123,
+      "learning_rate": 3.6194979093917944e-06,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 39076825,
+      "step": 1860
+    },
+    {
+      "epoch": 0.22377201947934827,
+      "grad_norm": 2.5631935892351754,
+      "learning_rate": 3.6190407043285724e-06,
+      "loss": 0.8754,
+      "num_input_tokens_seen": 39094280,
+      "step": 1861
+    },
+    {
+      "epoch": 0.22389226236998738,
+      "grad_norm": 1.9336247433549711,
+      "learning_rate": 3.618583253658163e-06,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 39114100,
+      "step": 1862
+    },
+    {
+      "epoch": 0.22401250526062647,
+      "grad_norm": 1.812798208582827,
+      "learning_rate": 3.618125557449961e-06,
+      "loss": 0.8671,
+      "num_input_tokens_seen": 39131875,
+      "step": 1863
+    },
+    {
+      "epoch": 0.22413274815126555,
+      "grad_norm": 1.9761938795379543,
+      "learning_rate": 3.6176676157733983e-06,
+      "loss": 0.8285,
+      "num_input_tokens_seen": 39146605,
+      "step": 1864
+    },
+    {
+      "epoch": 0.22425299104190466,
+      "grad_norm": 2.0903878119532586,
+      "learning_rate": 3.6172094286979443e-06,
+      "loss": 0.7448,
+      "num_input_tokens_seen": 39163695,
+      "step": 1865
+    },
+    {
+      "epoch": 0.22437323393254374,
+      "grad_norm": 1.3381082508435713,
+      "learning_rate": 3.6167509962931064e-06,
+      "loss": 0.8006,
+      "num_input_tokens_seen": 39189115,
+      "step": 1866
+    },
+    {
+      "epoch": 0.22449347682318282,
+      "grad_norm": 2.3726250221171066,
+      "learning_rate": 3.6162923186284276e-06,
+      "loss": 0.7662,
+      "num_input_tokens_seen": 39204795,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2246137197138219,
+      "grad_norm": 2.4969097510737392,
+      "learning_rate": 3.6158333957734888e-06,
+      "loss": 0.859,
+      "num_input_tokens_seen": 39223105,
+      "step": 1868
+    },
+    {
+      "epoch": 0.22473396260446102,
+      "grad_norm": 2.1247521939150404,
+      "learning_rate": 3.6153742277979088e-06,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 39240255,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2248542054951001,
+      "grad_norm": 2.168341137545459,
+      "learning_rate": 3.6149148147713434e-06,
+      "loss": 0.7852,
+      "num_input_tokens_seen": 39258210,
+      "step": 1870
+    },
+    {
+      "epoch": 0.22497444838573918,
+      "grad_norm": 1.7610091189798507,
+      "learning_rate": 3.614455156763484e-06,
+      "loss": 0.8603,
+      "num_input_tokens_seen": 39276235,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2250946912763783,
+      "grad_norm": 2.0717717211434974,
+      "learning_rate": 3.613995253844061e-06,
+      "loss": 0.7109,
+      "num_input_tokens_seen": 39293635,
+      "step": 1872
+    },
+    {
+      "epoch": 0.22521493416701738,
+      "grad_norm": 2.408855250361453,
+      "learning_rate": 3.6135351060828405e-06,
+      "loss": 0.8072,
+      "num_input_tokens_seen": 39313830,
+      "step": 1873
+    },
+    {
+      "epoch": 0.22533517705765646,
+      "grad_norm": 2.869815449067798,
+      "learning_rate": 3.6130747135496285e-06,
+      "loss": 0.6872,
+      "num_input_tokens_seen": 39332550,
+      "step": 1874
+    },
+    {
+      "epoch": 0.22545541994829554,
+      "grad_norm": 1.7999228418991604,
+      "learning_rate": 3.6126140763142646e-06,
+      "loss": 0.6594,
+      "num_input_tokens_seen": 39357300,
+      "step": 1875
+    },
+    {
+      "epoch": 0.22557566283893465,
+      "grad_norm": 2.7960108686994563,
+      "learning_rate": 3.6121531944466275e-06,
+      "loss": 0.847,
+      "num_input_tokens_seen": 39374345,
+      "step": 1876
+    },
+    {
+      "epoch": 0.22569590572957374,
+      "grad_norm": 1.8832018494830414,
+      "learning_rate": 3.611692068016633e-06,
+      "loss": 0.7783,
+      "num_input_tokens_seen": 39390395,
+      "step": 1877
+    },
+    {
+      "epoch": 0.22581614862021282,
+      "grad_norm": 2.1249052589571793,
+      "learning_rate": 3.611230697094233e-06,
+      "loss": 0.7417,
+      "num_input_tokens_seen": 39406815,
+      "step": 1878
+    },
+    {
+      "epoch": 0.22593639151085193,
+      "grad_norm": 1.913001581318875,
+      "learning_rate": 3.610769081749417e-06,
+      "loss": 0.8721,
+      "num_input_tokens_seen": 39426755,
+      "step": 1879
+    },
+    {
+      "epoch": 0.226056634401491,
+      "grad_norm": 2.610814296566517,
+      "learning_rate": 3.6103072220522117e-06,
+      "loss": 0.705,
+      "num_input_tokens_seen": 39442005,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2261768772921301,
+      "grad_norm": 1.8198415178625762,
+      "learning_rate": 3.609845118072682e-06,
+      "loss": 0.9099,
+      "num_input_tokens_seen": 39460395,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2262971201827692,
+      "grad_norm": 2.9192030319729048,
+      "learning_rate": 3.6093827698809276e-06,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 39479215,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2264173630734083,
+      "grad_norm": 2.5983784499890774,
+      "learning_rate": 3.6089201775470864e-06,
+      "loss": 0.8402,
+      "num_input_tokens_seen": 39494390,
+      "step": 1883
+    },
+    {
+      "epoch": 0.22653760596404737,
+      "grad_norm": 1.4230749118343458,
+      "learning_rate": 3.6084573411413334e-06,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 39513505,
+      "step": 1884
+    },
+    {
+      "epoch": 0.22665784885468646,
+      "grad_norm": 2.8641327338007176,
+      "learning_rate": 3.607994260733881e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 39532465,
+      "step": 1885
+    },
+    {
+      "epoch": 0.22677809174532557,
+      "grad_norm": 1.5481582678222425,
+      "learning_rate": 3.6075309363949776e-06,
+      "loss": 0.7425,
+      "num_input_tokens_seen": 39551355,
+      "step": 1886
+    },
+    {
+      "epoch": 0.22689833463596465,
+      "grad_norm": 3.1816065893436445,
+      "learning_rate": 3.607067368194909e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 39569440,
+      "step": 1887
+    },
+    {
+      "epoch": 0.22701857752660373,
+      "grad_norm": 1.9406530019100705,
+      "learning_rate": 3.606603556203999e-06,
+      "loss": 0.8073,
+      "num_input_tokens_seen": 39594105,
+      "step": 1888
+    },
+    {
+      "epoch": 0.22713882041724284,
+      "grad_norm": 1.8902317384700433,
+      "learning_rate": 3.6061395004926066e-06,
+      "loss": 0.8358,
+      "num_input_tokens_seen": 39612760,
+      "step": 1889
+    },
+    {
+      "epoch": 0.22725906330788193,
+      "grad_norm": 2.0761888540522486,
+      "learning_rate": 3.6056752011311285e-06,
+      "loss": 0.8437,
+      "num_input_tokens_seen": 39630940,
+      "step": 1890
+    },
+    {
+      "epoch": 0.227379306198521,
+      "grad_norm": 3.1060045921748887,
+      "learning_rate": 3.60521065819e-06,
+      "loss": 0.7979,
+      "num_input_tokens_seen": 39647970,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2274995490891601,
+      "grad_norm": 1.7324612644895736,
+      "learning_rate": 3.60474587173969e-06,
+      "loss": 0.8657,
+      "num_input_tokens_seen": 39666175,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2276197919797992,
+      "grad_norm": 1.9323524270200998,
+      "learning_rate": 3.6042808418507084e-06,
+      "loss": 0.8293,
+      "num_input_tokens_seen": 39683580,
+      "step": 1893
+    },
+    {
+      "epoch": 0.22774003487043828,
+      "grad_norm": 1.941570105420506,
+      "learning_rate": 3.6038155685935976e-06,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 39699870,
+      "step": 1894
+    },
+    {
+      "epoch": 0.22786027776107737,
+      "grad_norm": 2.4576588742409817,
+      "learning_rate": 3.60335005203894e-06,
+      "loss": 0.6998,
+      "num_input_tokens_seen": 39716260,
+      "step": 1895
+    },
+    {
+      "epoch": 0.22798052065171648,
+      "grad_norm": 0.9042930779975276,
+      "learning_rate": 3.6028842922573553e-06,
+      "loss": 0.6839,
+      "num_input_tokens_seen": 39780125,
+      "step": 1896
+    },
+    {
+      "epoch": 0.22810076354235556,
+      "grad_norm": 0.8898505458941157,
+      "learning_rate": 3.602418289319497e-06,
+      "loss": 0.6598,
+      "num_input_tokens_seen": 39838400,
+      "step": 1897
+    },
+    {
+      "epoch": 0.22822100643299464,
+      "grad_norm": 1.8621562627362105,
+      "learning_rate": 3.601952043296059e-06,
+      "loss": 0.727,
+      "num_input_tokens_seen": 39858115,
+      "step": 1898
+    },
+    {
+      "epoch": 0.22834124932363373,
+      "grad_norm": 1.9052214302192667,
+      "learning_rate": 3.6014855542577696e-06,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 39875045,
+      "step": 1899
+    },
+    {
+      "epoch": 0.22846149221427284,
+      "grad_norm": 2.547643582378809,
+      "learning_rate": 3.6010188222753943e-06,
+      "loss": 0.8392,
+      "num_input_tokens_seen": 39895535,
+      "step": 1900
+    },
+    {
+      "epoch": 0.22858173510491192,
+      "grad_norm": 0.9733291840628823,
+      "learning_rate": 3.6005518474197372e-06,
+      "loss": 0.6705,
+      "num_input_tokens_seen": 39947300,
+      "step": 1901
+    },
+    {
+      "epoch": 0.228701977995551,
+      "grad_norm": 2.0508446792495802,
+      "learning_rate": 3.6000846297616373e-06,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 39965320,
+      "step": 1902
+    },
+    {
+      "epoch": 0.22882222088619011,
+      "grad_norm": 2.3883241773732675,
+      "learning_rate": 3.5996171693719717e-06,
+      "loss": 0.7304,
+      "num_input_tokens_seen": 39981135,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2289424637768292,
+      "grad_norm": 0.908203707794976,
+      "learning_rate": 3.5991494663216528e-06,
+      "loss": 0.6674,
+      "num_input_tokens_seen": 40043840,
+      "step": 1904
+    },
+    {
+      "epoch": 0.22906270666746828,
+      "grad_norm": 2.7397214990218233,
+      "learning_rate": 3.5986815206816314e-06,
+      "loss": 0.8742,
+      "num_input_tokens_seen": 40062380,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2291829495581074,
+      "grad_norm": 1.7228749480474173,
+      "learning_rate": 3.598213332522895e-06,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 40082130,
+      "step": 1906
+    },
+    {
+      "epoch": 0.22930319244874647,
+      "grad_norm": 1.946403900680886,
+      "learning_rate": 3.597744901916466e-06,
+      "loss": 0.7714,
+      "num_input_tokens_seen": 40103135,
+      "step": 1907
+    },
+    {
+      "epoch": 0.22942343533938556,
+      "grad_norm": 1.871246327206096,
+      "learning_rate": 3.5972762289334058e-06,
+      "loss": 0.7642,
+      "num_input_tokens_seen": 40122485,
+      "step": 1908
+    },
+    {
+      "epoch": 0.22954367823002464,
+      "grad_norm": 2.073440700065059,
+      "learning_rate": 3.5968073136448116e-06,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 40140225,
+      "step": 1909
+    },
+    {
+      "epoch": 0.22966392112066375,
+      "grad_norm": 2.0128507389733117,
+      "learning_rate": 3.596338156121818e-06,
+      "loss": 0.9019,
+      "num_input_tokens_seen": 40158830,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22978416401130283,
+      "grad_norm": 0.7991249344780432,
+      "learning_rate": 3.595868756435595e-06,
+      "loss": 0.6227,
+      "num_input_tokens_seen": 40226230,
+      "step": 1911
+    },
+    {
+      "epoch": 0.22990440690194192,
+      "grad_norm": 2.152751563852422,
+      "learning_rate": 3.5953991146573504e-06,
+      "loss": 0.7911,
+      "num_input_tokens_seen": 40244595,
+      "step": 1912
+    },
+    {
+      "epoch": 0.23002464979258103,
+      "grad_norm": 2.382317929602197,
+      "learning_rate": 3.5949292308583294e-06,
+      "loss": 0.8227,
+      "num_input_tokens_seen": 40257560,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2301448926832201,
+      "grad_norm": 2.14384655963776,
+      "learning_rate": 3.5944591051098113e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 40276460,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2302651355738592,
+      "grad_norm": 2.059664286123235,
+      "learning_rate": 3.593988737483115e-06,
+      "loss": 0.8117,
+      "num_input_tokens_seen": 40296120,
+      "step": 1915
+    },
+    {
+      "epoch": 0.23038537846449827,
+      "grad_norm": 2.2813009452218607,
+      "learning_rate": 3.5935181280495947e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 40314420,
+      "step": 1916
+    },
+    {
+      "epoch": 0.23050562135513739,
+      "grad_norm": 0.9350073377708853,
+      "learning_rate": 3.5930472768806412e-06,
+      "loss": 0.5852,
+      "num_input_tokens_seen": 40372810,
+      "step": 1917
+    },
+    {
+      "epoch": 0.23062586424577647,
+      "grad_norm": 1.9011411373269642,
+      "learning_rate": 3.5925761840476826e-06,
+      "loss": 0.7668,
+      "num_input_tokens_seen": 40391140,
+      "step": 1918
+    },
+    {
+      "epoch": 0.23074610713641555,
+      "grad_norm": 1.9402603982195683,
+      "learning_rate": 3.592104849622183e-06,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 40413115,
+      "step": 1919
+    },
+    {
+      "epoch": 0.23086635002705466,
+      "grad_norm": 1.5434925609479349,
+      "learning_rate": 3.591633273675644e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 40435070,
+      "step": 1920
+    },
+    {
+      "epoch": 0.23098659291769374,
+      "grad_norm": 1.0316945798435038,
+      "learning_rate": 3.591161456279602e-06,
+      "loss": 0.624,
+      "num_input_tokens_seen": 40480335,
+      "step": 1921
+    },
+    {
+      "epoch": 0.23110683580833283,
+      "grad_norm": 1.5438785500942025,
+      "learning_rate": 3.590689397505633e-06,
+      "loss": 0.7928,
+      "num_input_tokens_seen": 40500965,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2312270786989719,
+      "grad_norm": 1.9166282684196445,
+      "learning_rate": 3.590217097425347e-06,
+      "loss": 0.8615,
+      "num_input_tokens_seen": 40520585,
+      "step": 1923
+    },
+    {
+      "epoch": 0.23134732158961102,
+      "grad_norm": 2.359851919679069,
+      "learning_rate": 3.589744556110391e-06,
+      "loss": 0.7084,
+      "num_input_tokens_seen": 40538295,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2314675644802501,
+      "grad_norm": 1.6994741035968677,
+      "learning_rate": 3.5892717736324504e-06,
+      "loss": 0.8419,
+      "num_input_tokens_seen": 40560840,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2315878073708892,
+      "grad_norm": 1.9013900870818812,
+      "learning_rate": 3.5887987500632447e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 40578565,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2317080502615283,
+      "grad_norm": 2.1498371085491463,
+      "learning_rate": 3.5883254854745325e-06,
+      "loss": 0.8286,
+      "num_input_tokens_seen": 40596675,
+      "step": 1927
+    },
+    {
+      "epoch": 0.23182829315216738,
+      "grad_norm": 2.032336523439975,
+      "learning_rate": 3.587851979938107e-06,
+      "loss": 0.7539,
+      "num_input_tokens_seen": 40613285,
+      "step": 1928
+    },
+    {
+      "epoch": 0.23194853604280646,
+      "grad_norm": 3.3820340120023413,
+      "learning_rate": 3.5873782335257985e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 40631170,
+      "step": 1929
+    },
+    {
+      "epoch": 0.23206877893344555,
+      "grad_norm": 2.1114283118630945,
+      "learning_rate": 3.5869042463094744e-06,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 40648605,
+      "step": 1930
+    },
+    {
+      "epoch": 0.23218902182408466,
+      "grad_norm": 2.309807461470747,
+      "learning_rate": 3.586430018361038e-06,
+      "loss": 0.7675,
+      "num_input_tokens_seen": 40668095,
+      "step": 1931
+    },
+    {
+      "epoch": 0.23230926471472374,
+      "grad_norm": 2.0155603281029335,
+      "learning_rate": 3.5859555497524283e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 40685050,
+      "step": 1932
+    },
+    {
+      "epoch": 0.23242950760536282,
+      "grad_norm": 2.16417221299743,
+      "learning_rate": 3.5854808405556237e-06,
+      "loss": 0.9129,
+      "num_input_tokens_seen": 40704005,
+      "step": 1933
+    },
+    {
+      "epoch": 0.23254975049600193,
+      "grad_norm": 2.4675576852434182,
+      "learning_rate": 3.585005890842635e-06,
+      "loss": 0.7536,
+      "num_input_tokens_seen": 40722275,
+      "step": 1934
+    },
+    {
+      "epoch": 0.23266999338664102,
+      "grad_norm": 1.8692484351588736,
+      "learning_rate": 3.584530700685514e-06,
+      "loss": 0.8464,
+      "num_input_tokens_seen": 40742255,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2327902362772801,
+      "grad_norm": 2.3222816585076345,
+      "learning_rate": 3.584055270156345e-06,
+      "loss": 0.8953,
+      "num_input_tokens_seen": 40758175,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2329104791679192,
+      "grad_norm": 2.6789768470412003,
+      "learning_rate": 3.5835795993272513e-06,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 40776180,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2330307220585583,
+      "grad_norm": 1.9214433835965385,
+      "learning_rate": 3.583103688270391e-06,
+      "loss": 0.7092,
+      "num_input_tokens_seen": 40795680,
+      "step": 1938
+    },
+    {
+      "epoch": 0.23315096494919738,
+      "grad_norm": 2.04890629295157,
+      "learning_rate": 3.58262753705796e-06,
+      "loss": 0.8824,
+      "num_input_tokens_seen": 40810290,
+      "step": 1939
+    },
+    {
+      "epoch": 0.23327120783983646,
+      "grad_norm": 0.7816616211251448,
+      "learning_rate": 3.5821511457621902e-06,
+      "loss": 0.5702,
+      "num_input_tokens_seen": 40867310,
+      "step": 1940
+    },
+    {
+      "epoch": 0.23339145073047557,
+      "grad_norm": 4.122146312504833,
+      "learning_rate": 3.5816745144553497e-06,
+      "loss": 0.8075,
+      "num_input_tokens_seen": 40882350,
+      "step": 1941
+    },
+    {
+      "epoch": 0.23351169362111465,
+      "grad_norm": 2.0814030411958226,
+      "learning_rate": 3.5811976432097424e-06,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 40899740,
+      "step": 1942
+    },
+    {
+      "epoch": 0.23363193651175373,
+      "grad_norm": 2.078498202941445,
+      "learning_rate": 3.58072053209771e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 40916015,
+      "step": 1943
+    },
+    {
+      "epoch": 0.23375217940239285,
+      "grad_norm": 2.231386415973234,
+      "learning_rate": 3.5802431811916296e-06,
+      "loss": 0.7906,
+      "num_input_tokens_seen": 40932345,
+      "step": 1944
+    },
+    {
+      "epoch": 0.23387242229303193,
+      "grad_norm": 2.0200699406527702,
+      "learning_rate": 3.579765590563916e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 40951465,
+      "step": 1945
+    },
+    {
+      "epoch": 0.233992665183671,
+      "grad_norm": 2.374967798747567,
+      "learning_rate": 3.579287760287017e-06,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 40971935,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2341129080743101,
+      "grad_norm": 1.7044282873928702,
+      "learning_rate": 3.5788096904334214e-06,
+      "loss": 0.722,
+      "num_input_tokens_seen": 40993365,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2342331509649492,
+      "grad_norm": 2.611296468352227,
+      "learning_rate": 3.578331381075651e-06,
+      "loss": 0.8012,
+      "num_input_tokens_seen": 41013585,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2343533938555883,
+      "grad_norm": 2.1889266988398375,
+      "learning_rate": 3.5778528322862646e-06,
+      "loss": 0.6941,
+      "num_input_tokens_seen": 41032125,
+      "step": 1949
+    },
+    {
+      "epoch": 0.23447363674622737,
+      "grad_norm": 1.9381579271819027,
+      "learning_rate": 3.577374044137858e-06,
+      "loss": 0.8577,
+      "num_input_tokens_seen": 41052600,
+      "step": 1950
+    },
+    {
+      "epoch": 0.23459387963686648,
+      "grad_norm": 2.1886096039608622,
+      "learning_rate": 3.5768950167030633e-06,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 41077020,
+      "step": 1951
+    },
+    {
+      "epoch": 0.23471412252750556,
+      "grad_norm": 1.8588150875321676,
+      "learning_rate": 3.576415750054548e-06,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 41096860,
+      "step": 1952
+    },
+    {
+      "epoch": 0.23483436541814465,
+      "grad_norm": 1.8334784066444243,
+      "learning_rate": 3.5759362442650172e-06,
+      "loss": 0.8447,
+      "num_input_tokens_seen": 41113330,
+      "step": 1953
+    },
+    {
+      "epoch": 0.23495460830878373,
+      "grad_norm": 2.378774743979848,
+      "learning_rate": 3.5754564994072113e-06,
+      "loss": 0.8479,
+      "num_input_tokens_seen": 41131890,
+      "step": 1954
+    },
+    {
+      "epoch": 0.23507485119942284,
+      "grad_norm": 2.534952939312894,
+      "learning_rate": 3.5749765155539067e-06,
+      "loss": 0.6115,
+      "num_input_tokens_seen": 41152095,
+      "step": 1955
+    },
+    {
+      "epoch": 0.23519509409006192,
+      "grad_norm": 2.562524259204442,
+      "learning_rate": 3.574496292777917e-06,
+      "loss": 0.9201,
+      "num_input_tokens_seen": 41170025,
+      "step": 1956
+    },
+    {
+      "epoch": 0.235315336980701,
+      "grad_norm": 1.834295416569206,
+      "learning_rate": 3.574015831152092e-06,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 41190160,
+      "step": 1957
+    },
+    {
+      "epoch": 0.23543557987134012,
+      "grad_norm": 3.2013452052103997,
+      "learning_rate": 3.5735351307493166e-06,
+      "loss": 0.8434,
+      "num_input_tokens_seen": 41207830,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2355558227619792,
+      "grad_norm": 1.6488427306440858,
+      "learning_rate": 3.5730541916425127e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 41229030,
+      "step": 1959
+    },
+    {
+      "epoch": 0.23567606565261828,
+      "grad_norm": 2.0596323723126404,
+      "learning_rate": 3.572573013904639e-06,
+      "loss": 0.8533,
+      "num_input_tokens_seen": 41248660,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2357963085432574,
+      "grad_norm": 1.9912405631338357,
+      "learning_rate": 3.572091597608689e-06,
+      "loss": 0.9165,
+      "num_input_tokens_seen": 41266505,
+      "step": 1961
+    },
+    {
+      "epoch": 0.23591655143389648,
+      "grad_norm": 2.295139618794565,
+      "learning_rate": 3.571609942827694e-06,
+      "loss": 0.7327,
+      "num_input_tokens_seen": 41285340,
+      "step": 1962
+    },
+    {
+      "epoch": 0.23603679432453556,
+      "grad_norm": 1.8240603745795938,
+      "learning_rate": 3.57112804963472e-06,
+      "loss": 0.8712,
+      "num_input_tokens_seen": 41303275,
+      "step": 1963
+    },
+    {
+      "epoch": 0.23615703721517464,
+      "grad_norm": 1.8664697395555647,
+      "learning_rate": 3.57064591810287e-06,
+      "loss": 0.7622,
+      "num_input_tokens_seen": 41320495,
+      "step": 1964
+    },
+    {
+      "epoch": 0.23627728010581375,
+      "grad_norm": 2.51726234462876,
+      "learning_rate": 3.570163548305284e-06,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 41339145,
+      "step": 1965
+    },
+    {
+      "epoch": 0.23639752299645284,
+      "grad_norm": 2.6233812697566448,
+      "learning_rate": 3.569680940315135e-06,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 41355265,
+      "step": 1966
+    },
+    {
+      "epoch": 0.23651776588709192,
+      "grad_norm": 1.8418930640849012,
+      "learning_rate": 3.5691980942056356e-06,
+      "loss": 0.8124,
+      "num_input_tokens_seen": 41374355,
+      "step": 1967
+    },
+    {
+      "epoch": 0.23663800877773103,
+      "grad_norm": 1.7873392222538373,
+      "learning_rate": 3.5687150100500332e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 41393775,
+      "step": 1968
+    },
+    {
+      "epoch": 0.2367582516683701,
+      "grad_norm": 1.9145089141168465,
+      "learning_rate": 3.568231687921611e-06,
+      "loss": 0.7407,
+      "num_input_tokens_seen": 41413670,
+      "step": 1969
+    },
+    {
+      "epoch": 0.2368784945590092,
+      "grad_norm": 1.5665581282558378,
+      "learning_rate": 3.5677481278936883e-06,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 41432970,
+      "step": 1970
+    },
+    {
+      "epoch": 0.23699873744964828,
+      "grad_norm": 0.8248098712175799,
+      "learning_rate": 3.5672643300396214e-06,
+      "loss": 0.5901,
+      "num_input_tokens_seen": 41501835,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2371189803402874,
+      "grad_norm": 2.6703042533564743,
+      "learning_rate": 3.566780294432802e-06,
+      "loss": 0.6846,
+      "num_input_tokens_seen": 41518730,
+      "step": 1972
+    },
+    {
+      "epoch": 0.23723922323092647,
+      "grad_norm": 2.716035205999358,
+      "learning_rate": 3.566296021146657e-06,
+      "loss": 0.7399,
+      "num_input_tokens_seen": 41537830,
+      "step": 1973
+    },
+    {
+      "epoch": 0.23735946612156555,
+      "grad_norm": 1.8389660109622008,
+      "learning_rate": 3.565811510254652e-06,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 41558430,
+      "step": 1974
+    },
+    {
+      "epoch": 0.23747970901220466,
+      "grad_norm": 0.7631641606922571,
+      "learning_rate": 3.5653267618302845e-06,
+      "loss": 0.5905,
+      "num_input_tokens_seen": 41625730,
+      "step": 1975
+    },
+    {
+      "epoch": 0.23759995190284375,
+      "grad_norm": 2.6548712039004365,
+      "learning_rate": 3.564841775947093e-06,
+      "loss": 0.855,
+      "num_input_tokens_seen": 41646340,
+      "step": 1976
+    },
+    {
+      "epoch": 0.23772019479348283,
+      "grad_norm": 2.0541591140911724,
+      "learning_rate": 3.5643565526786475e-06,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 41666000,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2378404376841219,
+      "grad_norm": 1.5183100645913723,
+      "learning_rate": 3.5638710920985574e-06,
+      "loss": 0.7697,
+      "num_input_tokens_seen": 41687180,
+      "step": 1978
+    },
+    {
+      "epoch": 0.23796068057476102,
+      "grad_norm": 2.2118660110579227,
+      "learning_rate": 3.563385394280465e-06,
+      "loss": 0.8153,
+      "num_input_tokens_seen": 41705225,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2380809234654001,
+      "grad_norm": 2.085147084130721,
+      "learning_rate": 3.5628994592980527e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 41722850,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2382011663560392,
+      "grad_norm": 1.868855569274737,
+      "learning_rate": 3.562413287225034e-06,
+      "loss": 0.7002,
+      "num_input_tokens_seen": 41740680,
+      "step": 1981
+    },
+    {
+      "epoch": 0.2383214092466783,
+      "grad_norm": 2.3100230194182365,
+      "learning_rate": 3.5619268781351623e-06,
+      "loss": 0.8874,
+      "num_input_tokens_seen": 41758470,
+      "step": 1982
+    },
+    {
+      "epoch": 0.23844165213731738,
+      "grad_norm": 1.8363520100131123,
+      "learning_rate": 3.5614402321022256e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 41776020,
+      "step": 1983
+    },
+    {
+      "epoch": 0.23856189502795647,
+      "grad_norm": 1.8507315502145791,
+      "learning_rate": 3.5609533492000463e-06,
+      "loss": 0.8678,
+      "num_input_tokens_seen": 41794630,
+      "step": 1984
+    },
+    {
+      "epoch": 0.23868213791859555,
+      "grad_norm": 2.299590355263758,
+      "learning_rate": 3.560466229502485e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 41813695,
+      "step": 1985
+    },
+    {
+      "epoch": 0.23880238080923466,
+      "grad_norm": 2.186898759084893,
+      "learning_rate": 3.5599788730834384e-06,
+      "loss": 0.8909,
+      "num_input_tokens_seen": 41831375,
+      "step": 1986
+    },
+    {
+      "epoch": 0.23892262369987374,
+      "grad_norm": 2.3807540324637033,
+      "learning_rate": 3.559491280016836e-06,
+      "loss": 0.8024,
+      "num_input_tokens_seen": 41849040,
+      "step": 1987
+    },
+    {
+      "epoch": 0.23904286659051283,
+      "grad_norm": 1.924623316523647,
+      "learning_rate": 3.5590034503766465e-06,
+      "loss": 0.715,
+      "num_input_tokens_seen": 41868425,
+      "step": 1988
+    },
+    {
+      "epoch": 0.23916310948115194,
+      "grad_norm": 2.274216111557667,
+      "learning_rate": 3.558515384236874e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 41885575,
+      "step": 1989
+    },
+    {
+      "epoch": 0.23928335237179102,
+      "grad_norm": 1.812759918663933,
+      "learning_rate": 3.558027081671556e-06,
+      "loss": 0.8336,
+      "num_input_tokens_seen": 41902280,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2394035952624301,
+      "grad_norm": 2.152535117884136,
+      "learning_rate": 3.557538542754769e-06,
+      "loss": 0.6907,
+      "num_input_tokens_seen": 41921695,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2395238381530692,
+      "grad_norm": 1.943737888704646,
+      "learning_rate": 3.557049767560623e-06,
+      "loss": 0.6659,
+      "num_input_tokens_seen": 41940330,
+      "step": 1992
+    },
+    {
+      "epoch": 0.2396440810437083,
+      "grad_norm": 1.9323410978081577,
+      "learning_rate": 3.5565607561632655e-06,
+      "loss": 0.8579,
+      "num_input_tokens_seen": 41958890,
+      "step": 1993
+    },
+    {
+      "epoch": 0.23976432393434738,
+      "grad_norm": 2.3876692544241798,
+      "learning_rate": 3.5560715086368787e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 41976480,
+      "step": 1994
+    },
+    {
+      "epoch": 0.23988456682498646,
+      "grad_norm": 1.8914006405504875,
+      "learning_rate": 3.5555820250556816e-06,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 41993400,
+      "step": 1995
+    },
+    {
+      "epoch": 0.24000480971562557,
+      "grad_norm": 2.3573687831034613,
+      "learning_rate": 3.5550923054939278e-06,
+      "loss": 0.6909,
+      "num_input_tokens_seen": 42012575,
+      "step": 1996
+    },
+    {
+      "epoch": 0.24012505260626466,
+      "grad_norm": 1.9824455732222286,
+      "learning_rate": 3.554602350025908e-06,
+      "loss": 0.7427,
+      "num_input_tokens_seen": 42033390,
+      "step": 1997
+    },
+    {
+      "epoch": 0.24024529549690374,
+      "grad_norm": 2.180520785259545,
+      "learning_rate": 3.5541121587259477e-06,
+      "loss": 0.8042,
+      "num_input_tokens_seen": 42050945,
+      "step": 1998
+    },
+    {
+      "epoch": 0.24036553838754285,
+      "grad_norm": 0.8148133872401061,
+      "learning_rate": 3.553621731668408e-06,
+      "loss": 0.5895,
+      "num_input_tokens_seen": 42113875,
+      "step": 1999
+    },
+    {
+      "epoch": 0.24048578127818193,
+      "grad_norm": 1.6484687125744413,
+      "learning_rate": 3.553131068927688e-06,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 42132000,
+      "step": 2000
+    },
+    {
+      "epoch": 0.24060602416882101,
+      "grad_norm": 1.5957627459394856,
+      "learning_rate": 3.552640170578219e-06,
+      "loss": 0.8017,
+      "num_input_tokens_seen": 42151970,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2407262670594601,
+      "grad_norm": 1.8658965983823987,
+      "learning_rate": 3.5521490366944703e-06,
+      "loss": 0.77,
+      "num_input_tokens_seen": 42169340,
+      "step": 2002
+    },
+    {
+      "epoch": 0.2408465099500992,
+      "grad_norm": 2.0297332442341562,
+      "learning_rate": 3.5516576673509474e-06,
+      "loss": 0.7948,
+      "num_input_tokens_seen": 42187060,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2409667528407383,
+      "grad_norm": 1.6637318474426894,
+      "learning_rate": 3.5511660626221896e-06,
+      "loss": 0.8544,
+      "num_input_tokens_seen": 42207420,
+      "step": 2004
+    },
+    {
+      "epoch": 0.24108699573137737,
+      "grad_norm": 2.418551717755083,
+      "learning_rate": 3.5506742225827744e-06,
+      "loss": 0.8789,
+      "num_input_tokens_seen": 42223995,
+      "step": 2005
+    },
+    {
+      "epoch": 0.24120723862201648,
+      "grad_norm": 2.1766767127385402,
+      "learning_rate": 3.5501821473073116e-06,
+      "loss": 0.8983,
+      "num_input_tokens_seen": 42240300,
+      "step": 2006
+    },
+    {
+      "epoch": 0.24132748151265557,
+      "grad_norm": 2.2011696445638007,
+      "learning_rate": 3.54968983687045e-06,
+      "loss": 0.8575,
+      "num_input_tokens_seen": 42256890,
+      "step": 2007
+    },
+    {
+      "epoch": 0.24144772440329465,
+      "grad_norm": 2.3247580656665994,
+      "learning_rate": 3.5491972913468717e-06,
+      "loss": 0.898,
+      "num_input_tokens_seen": 42273135,
+      "step": 2008
+    },
+    {
+      "epoch": 0.24156796729393373,
+      "grad_norm": 2.1757664170620674,
+      "learning_rate": 3.548704510811297e-06,
+      "loss": 0.791,
+      "num_input_tokens_seen": 42292050,
+      "step": 2009
+    },
+    {
+      "epoch": 0.24168821018457284,
+      "grad_norm": 2.282454304731145,
+      "learning_rate": 3.5482114953384787e-06,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 42311000,
+      "step": 2010
+    },
+    {
+      "epoch": 0.24180845307521193,
+      "grad_norm": 2.9543059651519874,
+      "learning_rate": 3.5477182450032077e-06,
+      "loss": 0.8391,
+      "num_input_tokens_seen": 42329320,
+      "step": 2011
+    },
+    {
+      "epoch": 0.241928695965851,
+      "grad_norm": 1.9156207666066267,
+      "learning_rate": 3.5472247598803097e-06,
+      "loss": 0.8301,
+      "num_input_tokens_seen": 42348385,
+      "step": 2012
+    },
+    {
+      "epoch": 0.24204893885649012,
+      "grad_norm": 2.2988084307208756,
+      "learning_rate": 3.546731040044645e-06,
+      "loss": 0.8533,
+      "num_input_tokens_seen": 42363275,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2421691817471292,
+      "grad_norm": 2.313373638790436,
+      "learning_rate": 3.546237085571112e-06,
+      "loss": 0.7484,
+      "num_input_tokens_seen": 42381430,
+      "step": 2014
+    },
+    {
+      "epoch": 0.24228942463776829,
+      "grad_norm": 2.0930367563196532,
+      "learning_rate": 3.5457428965346425e-06,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 42400090,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2424096675284074,
+      "grad_norm": 1.5929314870914237,
+      "learning_rate": 3.545248473010205e-06,
+      "loss": 0.7415,
+      "num_input_tokens_seen": 42422615,
+      "step": 2016
+    },
+    {
+      "epoch": 0.24252991041904648,
+      "grad_norm": 1.9203239408810884,
+      "learning_rate": 3.544753815072802e-06,
+      "loss": 0.8637,
+      "num_input_tokens_seen": 42440990,
+      "step": 2017
+    },
+    {
+      "epoch": 0.24265015330968556,
+      "grad_norm": 1.98105279902387,
+      "learning_rate": 3.544258922797474e-06,
+      "loss": 0.8844,
+      "num_input_tokens_seen": 42458830,
+      "step": 2018
+    },
+    {
+      "epoch": 0.24277039620032465,
+      "grad_norm": 1.6324632982825704,
+      "learning_rate": 3.543763796259295e-06,
+      "loss": 0.7763,
+      "num_input_tokens_seen": 42478505,
+      "step": 2019
+    },
+    {
+      "epoch": 0.24289063909096376,
+      "grad_norm": 1.7098462790156814,
+      "learning_rate": 3.5432684355333754e-06,
+      "loss": 0.9031,
+      "num_input_tokens_seen": 42496880,
+      "step": 2020
+    },
+    {
+      "epoch": 0.24301088198160284,
+      "grad_norm": 2.0373781607688772,
+      "learning_rate": 3.5427728406948613e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 42515715,
+      "step": 2021
+    },
+    {
+      "epoch": 0.24313112487224192,
+      "grad_norm": 0.7904769503625179,
+      "learning_rate": 3.5422770118189336e-06,
+      "loss": 0.5995,
+      "num_input_tokens_seen": 42579270,
+      "step": 2022
+    },
+    {
+      "epoch": 0.24325136776288103,
+      "grad_norm": 2.4704779280396734,
+      "learning_rate": 3.54178094898081e-06,
+      "loss": 0.738,
+      "num_input_tokens_seen": 42600600,
+      "step": 2023
+    },
+    {
+      "epoch": 0.24337161065352012,
+      "grad_norm": 1.850751660477363,
+      "learning_rate": 3.5412846522557422e-06,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 42621210,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2434918535441592,
+      "grad_norm": 2.5380020233235903,
+      "learning_rate": 3.540788121719018e-06,
+      "loss": 0.739,
+      "num_input_tokens_seen": 42639350,
+      "step": 2025
+    },
+    {
+      "epoch": 0.24361209643479828,
+      "grad_norm": 1.7385055779837548,
+      "learning_rate": 3.5402913574459604e-06,
+      "loss": 0.8224,
+      "num_input_tokens_seen": 42658975,
+      "step": 2026
+    },
+    {
+      "epoch": 0.2437323393254374,
+      "grad_norm": 1.646546972204318,
+      "learning_rate": 3.5397943595119297e-06,
+      "loss": 0.8484,
+      "num_input_tokens_seen": 42680115,
+      "step": 2027
+    },
+    {
+      "epoch": 0.24385258221607647,
+      "grad_norm": 2.7384630475939957,
+      "learning_rate": 3.5392971279923177e-06,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 42698055,
+      "step": 2028
+    },
+    {
+      "epoch": 0.24397282510671556,
+      "grad_norm": 2.1581138909219804,
+      "learning_rate": 3.5387996629625557e-06,
+      "loss": 0.8293,
+      "num_input_tokens_seen": 42715365,
+      "step": 2029
+    },
+    {
+      "epoch": 0.24409306799735467,
+      "grad_norm": 0.9004886639915187,
+      "learning_rate": 3.5383019644981083e-06,
+      "loss": 0.5971,
+      "num_input_tokens_seen": 42778780,
+      "step": 2030
+    },
+    {
+      "epoch": 0.24421331088799375,
+      "grad_norm": 2.453630244589166,
+      "learning_rate": 3.5378040326744763e-06,
+      "loss": 0.7286,
+      "num_input_tokens_seen": 42797985,
+      "step": 2031
+    },
+    {
+      "epoch": 0.24433355377863283,
+      "grad_norm": 3.5193544304400852,
+      "learning_rate": 3.5373058675671946e-06,
+      "loss": 0.8585,
+      "num_input_tokens_seen": 42815710,
+      "step": 2032
+    },
+    {
+      "epoch": 0.24445379666927192,
+      "grad_norm": 1.8330589959800159,
+      "learning_rate": 3.536807469251836e-06,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 42834585,
+      "step": 2033
+    },
+    {
+      "epoch": 0.24457403955991103,
+      "grad_norm": 1.884339570486781,
+      "learning_rate": 3.5363088378040055e-06,
+      "loss": 0.8145,
+      "num_input_tokens_seen": 42853195,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2446942824505501,
+      "grad_norm": 0.8071806796159787,
+      "learning_rate": 3.5358099732993463e-06,
+      "loss": 0.6803,
+      "num_input_tokens_seen": 42912025,
+      "step": 2035
+    },
+    {
+      "epoch": 0.2448145253411892,
+      "grad_norm": 2.0116741979293393,
+      "learning_rate": 3.5353108758135345e-06,
+      "loss": 0.8927,
+      "num_input_tokens_seen": 42930140,
+      "step": 2036
+    },
+    {
+      "epoch": 0.2449347682318283,
+      "grad_norm": 1.7226274994237825,
+      "learning_rate": 3.5348115454222843e-06,
+      "loss": 0.8095,
+      "num_input_tokens_seen": 42952445,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2450550111224674,
+      "grad_norm": 1.890449788576488,
+      "learning_rate": 3.5343119822013425e-06,
+      "loss": 0.8568,
+      "num_input_tokens_seen": 42971275,
+      "step": 2038
+    },
+    {
+      "epoch": 0.24517525401310647,
+      "grad_norm": 1.8181680870509245,
+      "learning_rate": 3.533812186226493e-06,
+      "loss": 0.766,
+      "num_input_tokens_seen": 42991705,
+      "step": 2039
+    },
+    {
+      "epoch": 0.24529549690374555,
+      "grad_norm": 1.856803025547529,
+      "learning_rate": 3.5333121575735545e-06,
+      "loss": 0.7609,
+      "num_input_tokens_seen": 43011065,
+      "step": 2040
+    },
+    {
+      "epoch": 0.24541573979438466,
+      "grad_norm": 1.9609392417679323,
+      "learning_rate": 3.532811896318381e-06,
+      "loss": 0.7505,
+      "num_input_tokens_seen": 43032855,
+      "step": 2041
+    },
+    {
+      "epoch": 0.24553598268502375,
+      "grad_norm": 2.286515650532021,
+      "learning_rate": 3.5323114025368615e-06,
+      "loss": 0.8074,
+      "num_input_tokens_seen": 43047640,
+      "step": 2042
+    },
+    {
+      "epoch": 0.24565622557566283,
+      "grad_norm": 2.914654354511826,
+      "learning_rate": 3.53181067630492e-06,
+      "loss": 0.8096,
+      "num_input_tokens_seen": 43064830,
+      "step": 2043
+    },
+    {
+      "epoch": 0.24577646846630194,
+      "grad_norm": 1.8052415755406959,
+      "learning_rate": 3.5313097176985175e-06,
+      "loss": 0.7582,
+      "num_input_tokens_seen": 43082860,
+      "step": 2044
+    },
+    {
+      "epoch": 0.24589671135694102,
+      "grad_norm": 1.9229918510804753,
+      "learning_rate": 3.5308085267936482e-06,
+      "loss": 0.8005,
+      "num_input_tokens_seen": 43100295,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2460169542475801,
+      "grad_norm": 1.8163909885262082,
+      "learning_rate": 3.530307103666342e-06,
+      "loss": 0.898,
+      "num_input_tokens_seen": 43119095,
+      "step": 2046
+    },
+    {
+      "epoch": 0.24613719713821922,
+      "grad_norm": 1.602627473312504,
+      "learning_rate": 3.5298054483926658e-06,
+      "loss": 0.8,
+      "num_input_tokens_seen": 43139510,
+      "step": 2047
+    },
+    {
+      "epoch": 0.2462574400288583,
+      "grad_norm": 2.415646835562988,
+      "learning_rate": 3.5293035610487187e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 43158595,
+      "step": 2048
+    },
+    {
+      "epoch": 0.24637768291949738,
+      "grad_norm": 0.7493648449419144,
+      "learning_rate": 3.5288014417106374e-06,
+      "loss": 0.6413,
+      "num_input_tokens_seen": 43224335,
+      "step": 2049
+    },
+    {
+      "epoch": 0.24649792581013646,
+      "grad_norm": 1.8568123621801542,
+      "learning_rate": 3.528299090454593e-06,
+      "loss": 0.7474,
+      "num_input_tokens_seen": 43244590,
+      "step": 2050
+    },
+    {
+      "epoch": 0.24661816870077558,
+      "grad_norm": 2.407486119507478,
+      "learning_rate": 3.527796507356792e-06,
+      "loss": 0.8165,
+      "num_input_tokens_seen": 43258200,
+      "step": 2051
+    },
+    {
+      "epoch": 0.24673841159141466,
+      "grad_norm": 2.7678091843740513,
+      "learning_rate": 3.527293692493475e-06,
+      "loss": 0.9039,
+      "num_input_tokens_seen": 43273785,
+      "step": 2052
+    },
+    {
+      "epoch": 0.24685865448205374,
+      "grad_norm": 2.840227984479663,
+      "learning_rate": 3.52679064594092e-06,
+      "loss": 0.7365,
+      "num_input_tokens_seen": 43290845,
+      "step": 2053
+    },
+    {
+      "epoch": 0.24697889737269285,
+      "grad_norm": 3.3607301221165264,
+      "learning_rate": 3.5262873677754375e-06,
+      "loss": 0.7391,
+      "num_input_tokens_seen": 43308570,
+      "step": 2054
+    },
+    {
+      "epoch": 0.24709914026333193,
+      "grad_norm": 1.6237964565075589,
+      "learning_rate": 3.5257838580733745e-06,
+      "loss": 0.8015,
+      "num_input_tokens_seen": 43327895,
+      "step": 2055
+    },
+    {
+      "epoch": 0.24721938315397102,
+      "grad_norm": 1.9121722110074375,
+      "learning_rate": 3.5252801169111138e-06,
+      "loss": 0.8675,
+      "num_input_tokens_seen": 43345280,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2473396260446101,
+      "grad_norm": 1.6931106581847013,
+      "learning_rate": 3.524776144365072e-06,
+      "loss": 0.7968,
+      "num_input_tokens_seen": 43363455,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2474598689352492,
+      "grad_norm": 1.6304278620037065,
+      "learning_rate": 3.5242719405117012e-06,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 43382980,
+      "step": 2058
+    },
+    {
+      "epoch": 0.2475801118258883,
+      "grad_norm": 3.8068261928314855,
+      "learning_rate": 3.5237675054274893e-06,
+      "loss": 0.7508,
+      "num_input_tokens_seen": 43401900,
+      "step": 2059
+    },
+    {
+      "epoch": 0.24770035471652738,
+      "grad_norm": 2.305675794792676,
+      "learning_rate": 3.5232628391889584e-06,
+      "loss": 0.7935,
+      "num_input_tokens_seen": 43419910,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2478205976071665,
+      "grad_norm": 2.6892557416943705,
+      "learning_rate": 3.522757941872666e-06,
+      "loss": 0.6339,
+      "num_input_tokens_seen": 43437785,
+      "step": 2061
+    },
+    {
+      "epoch": 0.24794084049780557,
+      "grad_norm": 1.609201304057219,
+      "learning_rate": 3.5222528135552042e-06,
+      "loss": 0.8237,
+      "num_input_tokens_seen": 43458965,
+      "step": 2062
+    },
+    {
+      "epoch": 0.24806108338844465,
+      "grad_norm": 1.8515480458279074,
+      "learning_rate": 3.5217474543132007e-06,
+      "loss": 0.8056,
+      "num_input_tokens_seen": 43477365,
+      "step": 2063
+    },
+    {
+      "epoch": 0.24818132627908374,
+      "grad_norm": 2.3621856143985607,
+      "learning_rate": 3.521241864223319e-06,
+      "loss": 0.6693,
+      "num_input_tokens_seen": 43496045,
+      "step": 2064
+    },
+    {
+      "epoch": 0.24830156916972285,
+      "grad_norm": 0.8180592572425286,
+      "learning_rate": 3.5207360433622552e-06,
+      "loss": 0.6341,
+      "num_input_tokens_seen": 43557765,
+      "step": 2065
+    },
+    {
+      "epoch": 0.24842181206036193,
+      "grad_norm": 1.5793706840060322,
+      "learning_rate": 3.5202299918067437e-06,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 43581080,
+      "step": 2066
+    },
+    {
+      "epoch": 0.248542054951001,
+      "grad_norm": 2.2629269704749038,
+      "learning_rate": 3.519723709633551e-06,
+      "loss": 0.6948,
+      "num_input_tokens_seen": 43599560,
+      "step": 2067
+    },
+    {
+      "epoch": 0.24866229784164012,
+      "grad_norm": 1.880290825264862,
+      "learning_rate": 3.519217196919479e-06,
+      "loss": 0.8276,
+      "num_input_tokens_seen": 43618265,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2487825407322792,
+      "grad_norm": 1.7902536325574798,
+      "learning_rate": 3.5187104537413664e-06,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 43637185,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2489027836229183,
+      "grad_norm": 2.0169731780268365,
+      "learning_rate": 3.518203480176086e-06,
+      "loss": 0.6741,
+      "num_input_tokens_seen": 43655835,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2490230265135574,
+      "grad_norm": 1.6276325073230646,
+      "learning_rate": 3.517696276300545e-06,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 43677095,
+      "step": 2071
+    },
+    {
+      "epoch": 0.24914326940419648,
+      "grad_norm": 2.3480156156059078,
+      "learning_rate": 3.517188842191685e-06,
+      "loss": 0.6989,
+      "num_input_tokens_seen": 43694965,
+      "step": 2072
+    },
+    {
+      "epoch": 0.24926351229483557,
+      "grad_norm": 2.033355337487096,
+      "learning_rate": 3.5166811779264837e-06,
+      "loss": 0.7504,
+      "num_input_tokens_seen": 43715005,
+      "step": 2073
+    },
+    {
+      "epoch": 0.24938375518547465,
+      "grad_norm": 1.8799020063600251,
+      "learning_rate": 3.5161732835819545e-06,
+      "loss": 0.7792,
+      "num_input_tokens_seen": 43734035,
+      "step": 2074
+    },
+    {
+      "epoch": 0.24950399807611376,
+      "grad_norm": 1.8782366078855302,
+      "learning_rate": 3.515665159235143e-06,
+      "loss": 0.8252,
+      "num_input_tokens_seen": 43752640,
+      "step": 2075
+    },
+    {
+      "epoch": 0.24962424096675284,
+      "grad_norm": 1.7172366300940103,
+      "learning_rate": 3.5151568049631318e-06,
+      "loss": 0.7514,
+      "num_input_tokens_seen": 43771075,
+      "step": 2076
+    },
+    {
+      "epoch": 0.24974448385739192,
+      "grad_norm": 1.535927721949452,
+      "learning_rate": 3.514648220843038e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 43792625,
+      "step": 2077
+    },
+    {
+      "epoch": 0.24986472674803104,
+      "grad_norm": 2.409551539587793,
+      "learning_rate": 3.514139406952014e-06,
+      "loss": 0.6723,
+      "num_input_tokens_seen": 43814370,
+      "step": 2078
+    },
+    {
+      "epoch": 0.24998496963867012,
+      "grad_norm": 1.6476746401070432,
+      "learning_rate": 3.5136303633672454e-06,
+      "loss": 0.8353,
+      "num_input_tokens_seen": 43834220,
+      "step": 2079
+    },
+    {
+      "epoch": 0.25010521252930923,
+      "grad_norm": 1.7154318910086155,
+      "learning_rate": 3.5131210901659544e-06,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 43855695,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2502254554199483,
+      "grad_norm": 4.150297074338366,
+      "learning_rate": 3.5126115874253967e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 43874970,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2503456983105874,
+      "grad_norm": 2.6762236521400276,
+      "learning_rate": 3.5121018552228644e-06,
+      "loss": 0.7993,
+      "num_input_tokens_seen": 43893195,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2504659412012265,
+      "grad_norm": 2.159700302791591,
+      "learning_rate": 3.5115918936356827e-06,
+      "loss": 0.7593,
+      "num_input_tokens_seen": 43909670,
+      "step": 2083
+    },
+    {
+      "epoch": 0.25058618409186556,
+      "grad_norm": 1.9639538826739673,
+      "learning_rate": 3.5110817027412123e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 43928480,
+      "step": 2084
+    },
+    {
+      "epoch": 0.25070642698250467,
+      "grad_norm": 2.1870331538462207,
+      "learning_rate": 3.5105712826168493e-06,
+      "loss": 0.6893,
+      "num_input_tokens_seen": 43947850,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2508266698731437,
+      "grad_norm": 1.7433921199255855,
+      "learning_rate": 3.5100606333400235e-06,
+      "loss": 0.7001,
+      "num_input_tokens_seen": 43964705,
+      "step": 2086
+    },
+    {
+      "epoch": 0.25094691276378284,
+      "grad_norm": 2.107617501671543,
+      "learning_rate": 3.5095497549882006e-06,
+      "loss": 0.7707,
+      "num_input_tokens_seen": 43982870,
+      "step": 2087
+    },
+    {
+      "epoch": 0.25106715565442195,
+      "grad_norm": 3.1158067235897913,
+      "learning_rate": 3.5090386476388796e-06,
+      "loss": 0.7208,
+      "num_input_tokens_seen": 44003380,
+      "step": 2088
+    },
+    {
+      "epoch": 0.251187398545061,
+      "grad_norm": 2.085876155159922,
+      "learning_rate": 3.5085273113695965e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 44027670,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2513076414357001,
+      "grad_norm": 1.9147780392185196,
+      "learning_rate": 3.508015746257919e-06,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 44046430,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2514278843263392,
+      "grad_norm": 2.010849773456429,
+      "learning_rate": 3.5075039523814518e-06,
+      "loss": 0.8265,
+      "num_input_tokens_seen": 44065340,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2515481272169783,
+      "grad_norm": 2.477381487997849,
+      "learning_rate": 3.5069919298178335e-06,
+      "loss": 0.8226,
+      "num_input_tokens_seen": 44081780,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2516683701076174,
+      "grad_norm": 1.6739047191476184,
+      "learning_rate": 3.506479678644738e-06,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 44101895,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2517886129982565,
+      "grad_norm": 2.639004105867083,
+      "learning_rate": 3.505967198939873e-06,
+      "loss": 0.7356,
+      "num_input_tokens_seen": 44118655,
+      "step": 2094
+    },
+    {
+      "epoch": 0.25190885588889556,
+      "grad_norm": 1.8571230490588497,
+      "learning_rate": 3.5054544907809813e-06,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 44138875,
+      "step": 2095
+    },
+    {
+      "epoch": 0.25202909877953467,
+      "grad_norm": 1.9237551894281173,
+      "learning_rate": 3.5049415542458397e-06,
+      "loss": 0.7946,
+      "num_input_tokens_seen": 44157500,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2521493416701738,
+      "grad_norm": 1.701850570795859,
+      "learning_rate": 3.504428389412262e-06,
+      "loss": 0.8305,
+      "num_input_tokens_seen": 44178030,
+      "step": 2097
+    },
+    {
+      "epoch": 0.25226958456081283,
+      "grad_norm": 2.471204438608791,
+      "learning_rate": 3.5039149963580927e-06,
+      "loss": 0.7345,
+      "num_input_tokens_seen": 44197770,
+      "step": 2098
+    },
+    {
+      "epoch": 0.25238982745145194,
+      "grad_norm": 2.3895430131241664,
+      "learning_rate": 3.503401375161215e-06,
+      "loss": 0.6947,
+      "num_input_tokens_seen": 44217235,
+      "step": 2099
+    },
+    {
+      "epoch": 0.252510070342091,
+      "grad_norm": 2.3929994729936674,
+      "learning_rate": 3.502887525899544e-06,
+      "loss": 0.8345,
+      "num_input_tokens_seen": 44235935,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2526303132327301,
+      "grad_norm": 1.7154447329042763,
+      "learning_rate": 3.50237344865103e-06,
+      "loss": 0.8208,
+      "num_input_tokens_seen": 44256655,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2527505561233692,
+      "grad_norm": 2.188366363478398,
+      "learning_rate": 3.501859143493658e-06,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 44277005,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2528707990140083,
+      "grad_norm": 0.9268280160926616,
+      "learning_rate": 3.5013446105054484e-06,
+      "loss": 0.6457,
+      "num_input_tokens_seen": 44329645,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2529910419046474,
+      "grad_norm": 2.148727841953047,
+      "learning_rate": 3.5008298497644555e-06,
+      "loss": 0.7508,
+      "num_input_tokens_seen": 44348410,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2531112847952865,
+      "grad_norm": 1.7830399982444798,
+      "learning_rate": 3.500314861348767e-06,
+      "loss": 0.8726,
+      "num_input_tokens_seen": 44368765,
+      "step": 2105
+    },
+    {
+      "epoch": 0.25323152768592555,
+      "grad_norm": 1.8745524804015687,
+      "learning_rate": 3.499799645336507e-06,
+      "loss": 0.7678,
+      "num_input_tokens_seen": 44385380,
+      "step": 2106
+    },
+    {
+      "epoch": 0.25335177057656466,
+      "grad_norm": 1.4415840588501692,
+      "learning_rate": 3.4992842018058336e-06,
+      "loss": 0.8633,
+      "num_input_tokens_seen": 44408000,
+      "step": 2107
+    },
+    {
+      "epoch": 0.25347201346720377,
+      "grad_norm": 2.9254912526342935,
+      "learning_rate": 3.4987685308349384e-06,
+      "loss": 0.8665,
+      "num_input_tokens_seen": 44425450,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2535922563578428,
+      "grad_norm": 2.270618612635359,
+      "learning_rate": 3.4982526325020497e-06,
+      "loss": 0.615,
+      "num_input_tokens_seen": 44442140,
+      "step": 2109
+    },
+    {
+      "epoch": 0.25371249924848194,
+      "grad_norm": 2.4502802225620157,
+      "learning_rate": 3.497736506885427e-06,
+      "loss": 0.819,
+      "num_input_tokens_seen": 44457480,
+      "step": 2110
+    },
+    {
+      "epoch": 0.25383274213912105,
+      "grad_norm": 1.7968232571779221,
+      "learning_rate": 3.4972201540633676e-06,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 44476555,
+      "step": 2111
+    },
+    {
+      "epoch": 0.2539529850297601,
+      "grad_norm": 1.8596786372081398,
+      "learning_rate": 3.4967035741142008e-06,
+      "loss": 0.8454,
+      "num_input_tokens_seen": 44495095,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2540732279203992,
+      "grad_norm": 1.940512766138527,
+      "learning_rate": 3.4961867671162917e-06,
+      "loss": 0.815,
+      "num_input_tokens_seen": 44514745,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2541934708110383,
+      "grad_norm": 2.5143612349076188,
+      "learning_rate": 3.4956697331480402e-06,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 44533035,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2543137137016774,
+      "grad_norm": 1.5744780532127773,
+      "learning_rate": 3.495152472287879e-06,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 44553465,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2544339565923165,
+      "grad_norm": 34.26379690078498,
+      "learning_rate": 3.4946349846142766e-06,
+      "loss": 0.7322,
+      "num_input_tokens_seen": 44572325,
+      "step": 2116
+    },
+    {
+      "epoch": 0.25455419948295555,
+      "grad_norm": 2.1448054262698215,
+      "learning_rate": 3.4941172702057353e-06,
+      "loss": 0.7509,
+      "num_input_tokens_seen": 44592105,
+      "step": 2117
+    },
+    {
+      "epoch": 0.25467444237359466,
+      "grad_norm": 1.8188379838192985,
+      "learning_rate": 3.4935993291407924e-06,
+      "loss": 0.7985,
+      "num_input_tokens_seen": 44610650,
+      "step": 2118
+    },
+    {
+      "epoch": 0.25479468526423377,
+      "grad_norm": 2.334983122139885,
+      "learning_rate": 3.4930811614980183e-06,
+      "loss": 0.7053,
+      "num_input_tokens_seen": 44632065,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2549149281548728,
+      "grad_norm": 1.7362084582043864,
+      "learning_rate": 3.4925627673560198e-06,
+      "loss": 0.7891,
+      "num_input_tokens_seen": 44652445,
+      "step": 2120
+    },
+    {
+      "epoch": 0.25503517104551193,
+      "grad_norm": 1.6921461899369696,
+      "learning_rate": 3.4920441467934357e-06,
+      "loss": 0.8832,
+      "num_input_tokens_seen": 44672680,
+      "step": 2121
+    },
+    {
+      "epoch": 0.25515541393615104,
+      "grad_norm": 2.1800507449511457,
+      "learning_rate": 3.491525299888941e-06,
+      "loss": 0.8308,
+      "num_input_tokens_seen": 44691245,
+      "step": 2122
+    },
+    {
+      "epoch": 0.2552756568267901,
+      "grad_norm": 0.9948294463111278,
+      "learning_rate": 3.491006226721244e-06,
+      "loss": 0.6831,
+      "num_input_tokens_seen": 44755175,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2553958997174292,
+      "grad_norm": 2.555251923405769,
+      "learning_rate": 3.4904869273690882e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 44772785,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2555161426080683,
+      "grad_norm": 2.052406816960019,
+      "learning_rate": 3.4899674019112506e-06,
+      "loss": 0.8899,
+      "num_input_tokens_seen": 44791805,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2556363854987074,
+      "grad_norm": 1.6574644626872266,
+      "learning_rate": 3.4894476504265428e-06,
+      "loss": 0.6899,
+      "num_input_tokens_seen": 44815765,
+      "step": 2126
+    },
+    {
+      "epoch": 0.2557566283893465,
+      "grad_norm": 0.7862676542900094,
+      "learning_rate": 3.4889276729938104e-06,
+      "loss": 0.5734,
+      "num_input_tokens_seen": 44874015,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2558768712799856,
+      "grad_norm": 1.9774783479580482,
+      "learning_rate": 3.488407469691934e-06,
+      "loss": 0.8014,
+      "num_input_tokens_seen": 44894430,
+      "step": 2128
+    },
+    {
+      "epoch": 0.25599711417062465,
+      "grad_norm": 2.3910767444223384,
+      "learning_rate": 3.487887040599828e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 44913950,
+      "step": 2129
+    },
+    {
+      "epoch": 0.25611735706126376,
+      "grad_norm": 2.6924668493362702,
+      "learning_rate": 3.4873663857964407e-06,
+      "loss": 0.7505,
+      "num_input_tokens_seen": 44930885,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2562375999519028,
+      "grad_norm": 1.8992595283427995,
+      "learning_rate": 3.4868455053607556e-06,
+      "loss": 0.6676,
+      "num_input_tokens_seen": 44950220,
+      "step": 2131
+    },
+    {
+      "epoch": 0.2563578428425419,
+      "grad_norm": 2.37519934545032,
+      "learning_rate": 3.4863243993717887e-06,
+      "loss": 0.71,
+      "num_input_tokens_seen": 44969240,
+      "step": 2132
+    },
+    {
+      "epoch": 0.25647808573318104,
+      "grad_norm": 1.863093317516603,
+      "learning_rate": 3.485803067908593e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 44988470,
+      "step": 2133
+    },
+    {
+      "epoch": 0.2565983286238201,
+      "grad_norm": 1.7910847964400878,
+      "learning_rate": 3.485281511050253e-06,
+      "loss": 0.7882,
+      "num_input_tokens_seen": 45010325,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2567185715144592,
+      "grad_norm": 3.058592255023808,
+      "learning_rate": 3.484759728875889e-06,
+      "loss": 0.8919,
+      "num_input_tokens_seen": 45025410,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2568388144050983,
+      "grad_norm": 1.8227084894311985,
+      "learning_rate": 3.484237721464654e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 45043425,
+      "step": 2136
+    },
+    {
+      "epoch": 0.25695905729573737,
+      "grad_norm": 2.4415837073030637,
+      "learning_rate": 3.483715488895737e-06,
+      "loss": 0.6643,
+      "num_input_tokens_seen": 45063475,
+      "step": 2137
+    },
+    {
+      "epoch": 0.2570793001863765,
+      "grad_norm": 2.197391872815811,
+      "learning_rate": 3.48319303124836e-06,
+      "loss": 0.7885,
+      "num_input_tokens_seen": 45083575,
+      "step": 2138
+    },
+    {
+      "epoch": 0.2571995430770156,
+      "grad_norm": 2.405762954180929,
+      "learning_rate": 3.4826703486017798e-06,
+      "loss": 0.6764,
+      "num_input_tokens_seen": 45102920,
+      "step": 2139
+    },
+    {
+      "epoch": 0.25731978596765465,
+      "grad_norm": 1.630444781653425,
+      "learning_rate": 3.4821474410352862e-06,
+      "loss": 0.7602,
+      "num_input_tokens_seen": 45121300,
+      "step": 2140
+    },
+    {
+      "epoch": 0.25744002885829376,
+      "grad_norm": 0.968172411862686,
+      "learning_rate": 3.481624308628205e-06,
+      "loss": 0.6683,
+      "num_input_tokens_seen": 45182390,
+      "step": 2141
+    },
+    {
+      "epoch": 0.25756027174893287,
+      "grad_norm": 3.4037471300743953,
+      "learning_rate": 3.481100951459893e-06,
+      "loss": 0.9978,
+      "num_input_tokens_seen": 45195130,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2576805146395719,
+      "grad_norm": 1.621048231219969,
+      "learning_rate": 3.480577369609745e-06,
+      "loss": 0.7818,
+      "num_input_tokens_seen": 45215740,
+      "step": 2143
+    },
+    {
+      "epoch": 0.25780075753021103,
+      "grad_norm": 2.034880649045596,
+      "learning_rate": 3.4800535631571874e-06,
+      "loss": 0.8729,
+      "num_input_tokens_seen": 45230990,
+      "step": 2144
+    },
+    {
+      "epoch": 0.25792100042085014,
+      "grad_norm": 2.2247593353905457,
+      "learning_rate": 3.4795295321816804e-06,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 45249535,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2580412433114892,
+      "grad_norm": 2.0019763486429216,
+      "learning_rate": 3.47900527676272e-06,
+      "loss": 0.9097,
+      "num_input_tokens_seen": 45267590,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2581614862021283,
+      "grad_norm": 1.9973985384800115,
+      "learning_rate": 3.478480796979835e-06,
+      "loss": 0.88,
+      "num_input_tokens_seen": 45285195,
+      "step": 2147
+    },
+    {
+      "epoch": 0.25828172909276736,
+      "grad_norm": 1.5622650017445285,
+      "learning_rate": 3.477956092912589e-06,
+      "loss": 0.7697,
+      "num_input_tokens_seen": 45306460,
+      "step": 2148
+    },
+    {
+      "epoch": 0.2584019719834065,
+      "grad_norm": 0.7082370874638909,
+      "learning_rate": 3.4774311646405783e-06,
+      "loss": 0.5816,
+      "num_input_tokens_seen": 45376085,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2585222148740456,
+      "grad_norm": 1.963988692512629,
+      "learning_rate": 3.476906012243435e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 45394715,
+      "step": 2150
+    },
+    {
+      "epoch": 0.25864245776468464,
+      "grad_norm": 2.4758094485214253,
+      "learning_rate": 3.4763806358008235e-06,
+      "loss": 0.816,
+      "num_input_tokens_seen": 45415635,
+      "step": 2151
+    },
+    {
+      "epoch": 0.25876270065532375,
+      "grad_norm": 2.5172415041472256,
+      "learning_rate": 3.475855035392444e-06,
+      "loss": 0.8597,
+      "num_input_tokens_seen": 45430675,
+      "step": 2152
+    },
+    {
+      "epoch": 0.25888294354596286,
+      "grad_norm": 1.6722701755583247,
+      "learning_rate": 3.475329211098029e-06,
+      "loss": 0.7008,
+      "num_input_tokens_seen": 45453550,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2590031864366019,
+      "grad_norm": 1.6905553467223597,
+      "learning_rate": 3.474803162997345e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 45474000,
+      "step": 2154
+    },
+    {
+      "epoch": 0.25912342932724103,
+      "grad_norm": 0.8806568624122615,
+      "learning_rate": 3.4742768911701944e-06,
+      "loss": 0.5767,
+      "num_input_tokens_seen": 45536415,
+      "step": 2155
+    },
+    {
+      "epoch": 0.25924367221788014,
+      "grad_norm": 3.683229347470357,
+      "learning_rate": 3.4737503956964113e-06,
+      "loss": 0.6925,
+      "num_input_tokens_seen": 45548440,
+      "step": 2156
+    },
+    {
+      "epoch": 0.2593639151085192,
+      "grad_norm": 2.2002151234468927,
+      "learning_rate": 3.473223676655865e-06,
+      "loss": 0.6697,
+      "num_input_tokens_seen": 45566160,
+      "step": 2157
+    },
+    {
+      "epoch": 0.2594841579991583,
+      "grad_norm": 1.7902954293342122,
+      "learning_rate": 3.4726967341284585e-06,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 45583745,
+      "step": 2158
+    },
+    {
+      "epoch": 0.2596044008897974,
+      "grad_norm": 1.923754365866425,
+      "learning_rate": 3.4721695681941282e-06,
+      "loss": 0.7524,
+      "num_input_tokens_seen": 45602505,
+      "step": 2159
+    },
+    {
+      "epoch": 0.25972464378043647,
+      "grad_norm": 2.071586187874828,
+      "learning_rate": 3.471642178932845e-06,
+      "loss": 0.832,
+      "num_input_tokens_seen": 45620870,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2598448866710756,
+      "grad_norm": 1.8995520874012908,
+      "learning_rate": 3.471114566424613e-06,
+      "loss": 0.8949,
+      "num_input_tokens_seen": 45639050,
+      "step": 2161
+    },
+    {
+      "epoch": 0.25996512956171464,
+      "grad_norm": 1.946944032000019,
+      "learning_rate": 3.4705867307494715e-06,
+      "loss": 0.7633,
+      "num_input_tokens_seen": 45657840,
+      "step": 2162
+    },
+    {
+      "epoch": 0.26008537245235375,
+      "grad_norm": 2.3324971376385024,
+      "learning_rate": 3.470058671987492e-06,
+      "loss": 0.8426,
+      "num_input_tokens_seen": 45675825,
+      "step": 2163
+    },
+    {
+      "epoch": 0.26020561534299286,
+      "grad_norm": 2.0286072220221314,
+      "learning_rate": 3.4695303902187805e-06,
+      "loss": 0.8294,
+      "num_input_tokens_seen": 45695100,
+      "step": 2164
+    },
+    {
+      "epoch": 0.2603258582336319,
+      "grad_norm": 1.7980597260221758,
+      "learning_rate": 3.4690018855234775e-06,
+      "loss": 0.7817,
+      "num_input_tokens_seen": 45715540,
+      "step": 2165
+    },
+    {
+      "epoch": 0.260446101124271,
+      "grad_norm": 1.7662292637463965,
+      "learning_rate": 3.4684731579817568e-06,
+      "loss": 0.8059,
+      "num_input_tokens_seen": 45736250,
+      "step": 2166
+    },
+    {
+      "epoch": 0.26056634401491013,
+      "grad_norm": 1.6959706584943053,
+      "learning_rate": 3.4679442076738247e-06,
+      "loss": 0.7668,
+      "num_input_tokens_seen": 45755685,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2606865869055492,
+      "grad_norm": 4.693639325318077,
+      "learning_rate": 3.467415034679924e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 45775105,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2608068297961883,
+      "grad_norm": 2.1487949100743715,
+      "learning_rate": 3.4668856390803295e-06,
+      "loss": 0.7999,
+      "num_input_tokens_seen": 45792705,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2609270726868274,
+      "grad_norm": 1.9495231218974878,
+      "learning_rate": 3.4663560209553495e-06,
+      "loss": 0.892,
+      "num_input_tokens_seen": 45810490,
+      "step": 2170
+    },
+    {
+      "epoch": 0.26104731557746647,
+      "grad_norm": 1.7341364471547513,
+      "learning_rate": 3.4658261803853267e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 45828135,
+      "step": 2171
+    },
+    {
+      "epoch": 0.2611675584681056,
+      "grad_norm": 2.270725660903174,
+      "learning_rate": 3.4652961174506383e-06,
+      "loss": 0.8033,
+      "num_input_tokens_seen": 45847725,
+      "step": 2172
+    },
+    {
+      "epoch": 0.2612878013587447,
+      "grad_norm": 1.052839797496255,
+      "learning_rate": 3.464765832231694e-06,
+      "loss": 0.6039,
+      "num_input_tokens_seen": 45901610,
+      "step": 2173
+    },
+    {
+      "epoch": 0.26140804424938374,
+      "grad_norm": 1.6998793311022538,
+      "learning_rate": 3.4642353248089373e-06,
+      "loss": 0.703,
+      "num_input_tokens_seen": 45920090,
+      "step": 2174
+    },
+    {
+      "epoch": 0.26152828714002285,
+      "grad_norm": 1.8263511866976727,
+      "learning_rate": 3.463704595262846e-06,
+      "loss": 0.8006,
+      "num_input_tokens_seen": 45940690,
+      "step": 2175
+    },
+    {
+      "epoch": 0.26164853003066196,
+      "grad_norm": 1.6755540018575477,
+      "learning_rate": 3.463173643673931e-06,
+      "loss": 0.6986,
+      "num_input_tokens_seen": 45962935,
+      "step": 2176
+    },
+    {
+      "epoch": 0.261768772921301,
+      "grad_norm": 0.9759673344215334,
+      "learning_rate": 3.4626424701227387e-06,
+      "loss": 0.6617,
+      "num_input_tokens_seen": 46017715,
+      "step": 2177
+    },
+    {
+      "epoch": 0.26188901581194013,
+      "grad_norm": 0.875338679096256,
+      "learning_rate": 3.4621110746898452e-06,
+      "loss": 0.6194,
+      "num_input_tokens_seen": 46085295,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2620092587025792,
+      "grad_norm": 1.6743386784600551,
+      "learning_rate": 3.4615794574558654e-06,
+      "loss": 0.7411,
+      "num_input_tokens_seen": 46104025,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2621295015932183,
+      "grad_norm": 2.4061105282545574,
+      "learning_rate": 3.4610476185014436e-06,
+      "loss": 0.8492,
+      "num_input_tokens_seen": 46121005,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2622497444838574,
+      "grad_norm": 2.5406562160632533,
+      "learning_rate": 3.4605155579072597e-06,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 46140580,
+      "step": 2181
+    },
+    {
+      "epoch": 0.26236998737449646,
+      "grad_norm": 1.8753165110430303,
+      "learning_rate": 3.459983275754027e-06,
+      "loss": 0.7093,
+      "num_input_tokens_seen": 46159195,
+      "step": 2182
+    },
+    {
+      "epoch": 0.26249023026513557,
+      "grad_norm": 2.8662073220461406,
+      "learning_rate": 3.4594507721224918e-06,
+      "loss": 0.7887,
+      "num_input_tokens_seen": 46177565,
+      "step": 2183
+    },
+    {
+      "epoch": 0.2626104731557747,
+      "grad_norm": 2.0352399382495685,
+      "learning_rate": 3.4589180470934353e-06,
+      "loss": 0.8151,
+      "num_input_tokens_seen": 46197150,
+      "step": 2184
+    },
+    {
+      "epoch": 0.26273071604641374,
+      "grad_norm": 2.404176957693624,
+      "learning_rate": 3.4583851007476713e-06,
+      "loss": 0.7681,
+      "num_input_tokens_seen": 46215340,
+      "step": 2185
+    },
+    {
+      "epoch": 0.26285095893705285,
+      "grad_norm": 2.617946147317752,
+      "learning_rate": 3.4578519331660464e-06,
+      "loss": 0.6916,
+      "num_input_tokens_seen": 46232055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.26297120182769196,
+      "grad_norm": 1.8247657448747576,
+      "learning_rate": 3.4573185444294426e-06,
+      "loss": 0.8213,
+      "num_input_tokens_seen": 46250140,
+      "step": 2187
+    },
+    {
+      "epoch": 0.263091444718331,
+      "grad_norm": 7.264731591927456,
+      "learning_rate": 3.456784934618774e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 46271025,
+      "step": 2188
+    },
+    {
+      "epoch": 0.2632116876089701,
+      "grad_norm": 2.04998011172163,
+      "learning_rate": 3.4562511038149897e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 46286240,
+      "step": 2189
+    },
+    {
+      "epoch": 0.26333193049960923,
+      "grad_norm": 1.2344818286963726,
+      "learning_rate": 3.4557170520990705e-06,
+      "loss": 0.5989,
+      "num_input_tokens_seen": 46346635,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2634521733902483,
+      "grad_norm": 1.487749041657066,
+      "learning_rate": 3.455182779552032e-06,
+      "loss": 0.8583,
+      "num_input_tokens_seen": 46369240,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2635724162808874,
+      "grad_norm": 1.8305847418538992,
+      "learning_rate": 3.4546482862549226e-06,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 46389275,
+      "step": 2192
+    },
+    {
+      "epoch": 0.2636926591715265,
+      "grad_norm": 2.20253937856688,
+      "learning_rate": 3.454113572288825e-06,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 46405585,
+      "step": 2193
+    },
+    {
+      "epoch": 0.26381290206216557,
+      "grad_norm": 1.8257681357010076,
+      "learning_rate": 3.453578637734854e-06,
+      "loss": 0.8047,
+      "num_input_tokens_seen": 46426495,
+      "step": 2194
+    },
+    {
+      "epoch": 0.2639331449528047,
+      "grad_norm": 2.334486540073954,
+      "learning_rate": 3.4530434826741605e-06,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 46447155,
+      "step": 2195
+    },
+    {
+      "epoch": 0.26405338784344373,
+      "grad_norm": 1.893350803765931,
+      "learning_rate": 3.452508107187926e-06,
+      "loss": 0.6949,
+      "num_input_tokens_seen": 46470250,
+      "step": 2196
+    },
+    {
+      "epoch": 0.26417363073408284,
+      "grad_norm": 1.6633260682290516,
+      "learning_rate": 3.451972511357366e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 46489515,
+      "step": 2197
+    },
+    {
+      "epoch": 0.26429387362472195,
+      "grad_norm": 1.6705127026126372,
+      "learning_rate": 3.4514366952637296e-06,
+      "loss": 0.847,
+      "num_input_tokens_seen": 46508995,
+      "step": 2198
+    },
+    {
+      "epoch": 0.264414116515361,
+      "grad_norm": 0.8430303941519564,
+      "learning_rate": 3.450900658988302e-06,
+      "loss": 0.6352,
+      "num_input_tokens_seen": 46570265,
+      "step": 2199
+    },
+    {
+      "epoch": 0.2645343594060001,
+      "grad_norm": 1.9688282072250518,
+      "learning_rate": 3.450364402612397e-06,
+      "loss": 0.7795,
+      "num_input_tokens_seen": 46587140,
+      "step": 2200
+    },
+    {
+      "epoch": 0.26465460229663923,
+      "grad_norm": 2.0738004212785186,
+      "learning_rate": 3.449827926217366e-06,
+      "loss": 0.8291,
+      "num_input_tokens_seen": 46606295,
+      "step": 2201
+    },
+    {
+      "epoch": 0.2647748451872783,
+      "grad_norm": 2.16193287978649,
+      "learning_rate": 3.449291229884591e-06,
+      "loss": 0.801,
+      "num_input_tokens_seen": 46627255,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2648950880779174,
+      "grad_norm": 2.4941842073282823,
+      "learning_rate": 3.4487543136954887e-06,
+      "loss": 0.861,
+      "num_input_tokens_seen": 46646595,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2650153309685565,
+      "grad_norm": 1.9009527232872498,
+      "learning_rate": 3.448217177731509e-06,
+      "loss": 0.9067,
+      "num_input_tokens_seen": 46666800,
+      "step": 2204
+    },
+    {
+      "epoch": 0.26513557385919556,
+      "grad_norm": 1.9754661053663947,
+      "learning_rate": 3.4476798220741348e-06,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 46685400,
+      "step": 2205
+    },
+    {
+      "epoch": 0.26525581674983467,
+      "grad_norm": 1.6107570961172946,
+      "learning_rate": 3.4471422468048826e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 46703845,
+      "step": 2206
+    },
+    {
+      "epoch": 0.2653760596404738,
+      "grad_norm": 4.460285902430027,
+      "learning_rate": 3.4466044520053022e-06,
+      "loss": 0.7344,
+      "num_input_tokens_seen": 46722570,
+      "step": 2207
+    },
+    {
+      "epoch": 0.26549630253111284,
+      "grad_norm": 2.0064457896255057,
+      "learning_rate": 3.446066437756977e-06,
+      "loss": 0.6027,
+      "num_input_tokens_seen": 46741495,
+      "step": 2208
+    },
+    {
+      "epoch": 0.26561654542175195,
+      "grad_norm": 2.2774316458225368,
+      "learning_rate": 3.4455282041415224e-06,
+      "loss": 0.7534,
+      "num_input_tokens_seen": 46760425,
+      "step": 2209
+    },
+    {
+      "epoch": 0.265736788312391,
+      "grad_norm": 2.849207733363642,
+      "learning_rate": 3.4449897512405894e-06,
+      "loss": 0.8709,
+      "num_input_tokens_seen": 46779295,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2658570312030301,
+      "grad_norm": 1.889225322529489,
+      "learning_rate": 3.444451079135859e-06,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 46798525,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2659772740936692,
+      "grad_norm": 1.8949274588165255,
+      "learning_rate": 3.4439121879090485e-06,
+      "loss": 0.7347,
+      "num_input_tokens_seen": 46816025,
+      "step": 2212
+    },
+    {
+      "epoch": 0.2660975169843083,
+      "grad_norm": 1.860559261096679,
+      "learning_rate": 3.443373077641908e-06,
+      "loss": 0.8213,
+      "num_input_tokens_seen": 46834670,
+      "step": 2213
+    },
+    {
+      "epoch": 0.2662177598749474,
+      "grad_norm": 8.146773699127522,
+      "learning_rate": 3.4428337484162183e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 46855200,
+      "step": 2214
+    },
+    {
+      "epoch": 0.2663380027655865,
+      "grad_norm": 2.182734401590511,
+      "learning_rate": 3.4422942003137967e-06,
+      "loss": 0.8469,
+      "num_input_tokens_seen": 46872950,
+      "step": 2215
+    },
+    {
+      "epoch": 0.26645824565622556,
+      "grad_norm": 0.9227837422763205,
+      "learning_rate": 3.4417544334164916e-06,
+      "loss": 0.5806,
+      "num_input_tokens_seen": 46936815,
+      "step": 2216
+    },
+    {
+      "epoch": 0.26657848854686467,
+      "grad_norm": 1.6901293441811094,
+      "learning_rate": 3.4412144478061854e-06,
+      "loss": 0.7766,
+      "num_input_tokens_seen": 46958945,
+      "step": 2217
+    },
+    {
+      "epoch": 0.2666987314375038,
+      "grad_norm": 1.8534360895325481,
+      "learning_rate": 3.4406742435647925e-06,
+      "loss": 0.7498,
+      "num_input_tokens_seen": 46978730,
+      "step": 2218
+    },
+    {
+      "epoch": 0.26681897432814283,
+      "grad_norm": 2.932088770184242,
+      "learning_rate": 3.440133820774263e-06,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 46998260,
+      "step": 2219
+    },
+    {
+      "epoch": 0.26693921721878194,
+      "grad_norm": 2.739173324547336,
+      "learning_rate": 3.439593179516578e-06,
+      "loss": 0.8108,
+      "num_input_tokens_seen": 47017890,
+      "step": 2220
+    },
+    {
+      "epoch": 0.26705946010942105,
+      "grad_norm": 1.8765620181825116,
+      "learning_rate": 3.4390523198737524e-06,
+      "loss": 0.8069,
+      "num_input_tokens_seen": 47036770,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2671797030000601,
+      "grad_norm": 1.7216434298191865,
+      "learning_rate": 3.4385112419278333e-06,
+      "loss": 0.7278,
+      "num_input_tokens_seen": 47057715,
+      "step": 2222
+    },
+    {
+      "epoch": 0.2672999458906992,
+      "grad_norm": 0.8408885734508484,
+      "learning_rate": 3.437969945760903e-06,
+      "loss": 0.6955,
+      "num_input_tokens_seen": 47115260,
+      "step": 2223
+    },
+    {
+      "epoch": 0.26742018878133833,
+      "grad_norm": 1.8580465936065493,
+      "learning_rate": 3.4374284314550755e-06,
+      "loss": 0.8932,
+      "num_input_tokens_seen": 47134020,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2675404316719774,
+      "grad_norm": 2.1219419616278925,
+      "learning_rate": 3.436886699092498e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 47152255,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2676606745626165,
+      "grad_norm": 2.8806147659039283,
+      "learning_rate": 3.4363447487553502e-06,
+      "loss": 0.7135,
+      "num_input_tokens_seen": 47165290,
+      "step": 2226
+    },
+    {
+      "epoch": 0.26778091745325555,
+      "grad_norm": 2.2812779444416376,
+      "learning_rate": 3.4358025805258455e-06,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 47184715,
+      "step": 2227
+    },
+    {
+      "epoch": 0.26790116034389466,
+      "grad_norm": 3.3352759992147303,
+      "learning_rate": 3.435260194486232e-06,
+      "loss": 0.8294,
+      "num_input_tokens_seen": 47202405,
+      "step": 2228
+    },
+    {
+      "epoch": 0.2680214032345338,
+      "grad_norm": 2.047436263158215,
+      "learning_rate": 3.4347175907187875e-06,
+      "loss": 0.8114,
+      "num_input_tokens_seen": 47219115,
+      "step": 2229
+    },
+    {
+      "epoch": 0.26814164612517283,
+      "grad_norm": 1.9031646212382474,
+      "learning_rate": 3.4341747693058254e-06,
+      "loss": 0.869,
+      "num_input_tokens_seen": 47237310,
+      "step": 2230
+    },
+    {
+      "epoch": 0.26826188901581194,
+      "grad_norm": 1.815388871146807,
+      "learning_rate": 3.4336317303296916e-06,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 47258005,
+      "step": 2231
+    },
+    {
+      "epoch": 0.26838213190645105,
+      "grad_norm": 2.5098328407837944,
+      "learning_rate": 3.4330884738727635e-06,
+      "loss": 0.7513,
+      "num_input_tokens_seen": 47275900,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2685023747970901,
+      "grad_norm": 1.8592202056204403,
+      "learning_rate": 3.4325450000174535e-06,
+      "loss": 0.7088,
+      "num_input_tokens_seen": 47292260,
+      "step": 2233
+    },
+    {
+      "epoch": 0.2686226176877292,
+      "grad_norm": 3.6834760310652723,
+      "learning_rate": 3.4320013088462063e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 47309340,
+      "step": 2234
+    },
+    {
+      "epoch": 0.2687428605783683,
+      "grad_norm": 1.6239923813927342,
+      "learning_rate": 3.4314574004414987e-06,
+      "loss": 0.8172,
+      "num_input_tokens_seen": 47329455,
+      "step": 2235
+    },
+    {
+      "epoch": 0.2688631034690074,
+      "grad_norm": 1.094267471694779,
+      "learning_rate": 3.4309132748858424e-06,
+      "loss": 0.6914,
+      "num_input_tokens_seen": 47390165,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2689833463596465,
+      "grad_norm": 1.7073026828010218,
+      "learning_rate": 3.430368932261779e-06,
+      "loss": 0.8366,
+      "num_input_tokens_seen": 47410240,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2691035892502856,
+      "grad_norm": 2.2535789342823027,
+      "learning_rate": 3.429824372651886e-06,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 47428110,
+      "step": 2238
+    },
+    {
+      "epoch": 0.26922383214092466,
+      "grad_norm": 2.156970536661827,
+      "learning_rate": 3.4292795961387732e-06,
+      "loss": 0.8373,
+      "num_input_tokens_seen": 47445730,
+      "step": 2239
+    },
+    {
+      "epoch": 0.26934407503156377,
+      "grad_norm": 2.8140608201981485,
+      "learning_rate": 3.4287346028050818e-06,
+      "loss": 0.8706,
+      "num_input_tokens_seen": 47461520,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2694643179222028,
+      "grad_norm": 1.4800634493951534,
+      "learning_rate": 3.4281893927334866e-06,
+      "loss": 0.792,
+      "num_input_tokens_seen": 47481150,
+      "step": 2241
+    },
+    {
+      "epoch": 0.26958456081284193,
+      "grad_norm": 2.26960457523604,
+      "learning_rate": 3.4276439660066963e-06,
+      "loss": 0.7464,
+      "num_input_tokens_seen": 47500570,
+      "step": 2242
+    },
+    {
+      "epoch": 0.26970480370348104,
+      "grad_norm": 2.4117836098713807,
+      "learning_rate": 3.427098322707452e-06,
+      "loss": 0.8347,
+      "num_input_tokens_seen": 47516255,
+      "step": 2243
+    },
+    {
+      "epoch": 0.2698250465941201,
+      "grad_norm": 2.075858671628424,
+      "learning_rate": 3.426552462918526e-06,
+      "loss": 0.8938,
+      "num_input_tokens_seen": 47533910,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2699452894847592,
+      "grad_norm": 2.564195606133789,
+      "learning_rate": 3.426006386722726e-06,
+      "loss": 0.7322,
+      "num_input_tokens_seen": 47551690,
+      "step": 2245
+    },
+    {
+      "epoch": 0.2700655323753983,
+      "grad_norm": 2.3469277653166585,
+      "learning_rate": 3.425460094202891e-06,
+      "loss": 0.9055,
+      "num_input_tokens_seen": 47569285,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2701857752660374,
+      "grad_norm": 2.0713905260041887,
+      "learning_rate": 3.424913585441893e-06,
+      "loss": 0.8122,
+      "num_input_tokens_seen": 47586840,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2703060181566765,
+      "grad_norm": 2.133451806300382,
+      "learning_rate": 3.424366860522637e-06,
+      "loss": 0.8697,
+      "num_input_tokens_seen": 47603585,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2704262610473156,
+      "grad_norm": 2.8016176917746347,
+      "learning_rate": 3.423819919528061e-06,
+      "loss": 0.8336,
+      "num_input_tokens_seen": 47621390,
+      "step": 2249
+    },
+    {
+      "epoch": 0.27054650393795465,
+      "grad_norm": 1.7331452914084589,
+      "learning_rate": 3.4232727625411355e-06,
+      "loss": 0.7846,
+      "num_input_tokens_seen": 47640215,
+      "step": 2250
+    },
+    {
+      "epoch": 0.27066674682859376,
+      "grad_norm": 1.6610563637184759,
+      "learning_rate": 3.4227253896448626e-06,
+      "loss": 0.8614,
+      "num_input_tokens_seen": 47657795,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2707869897192329,
+      "grad_norm": 2.272445659605438,
+      "learning_rate": 3.42217780092228e-06,
+      "loss": 0.8053,
+      "num_input_tokens_seen": 47675855,
+      "step": 2252
+    },
+    {
+      "epoch": 0.27090723260987193,
+      "grad_norm": 0.8459225567920692,
+      "learning_rate": 3.4216299964564554e-06,
+      "loss": 0.6326,
+      "num_input_tokens_seen": 47734195,
+      "step": 2253
+    },
+    {
+      "epoch": 0.27102747550051104,
+      "grad_norm": 1.864517718352496,
+      "learning_rate": 3.421081976330491e-06,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 47752430,
+      "step": 2254
+    },
+    {
+      "epoch": 0.27114771839115015,
+      "grad_norm": 1.9946337360967077,
+      "learning_rate": 3.4205337406275207e-06,
+      "loss": 0.8606,
+      "num_input_tokens_seen": 47772270,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2712679612817892,
+      "grad_norm": 2.5037745810711116,
+      "learning_rate": 3.419985289430711e-06,
+      "loss": 0.756,
+      "num_input_tokens_seen": 47788740,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2713882041724283,
+      "grad_norm": 2.515454140819365,
+      "learning_rate": 3.419436622823262e-06,
+      "loss": 0.7843,
+      "num_input_tokens_seen": 47809180,
+      "step": 2257
+    },
+    {
+      "epoch": 0.27150844706306737,
+      "grad_norm": 1.6423051647653797,
+      "learning_rate": 3.4188877408884063e-06,
+      "loss": 0.7307,
+      "num_input_tokens_seen": 47829605,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2716286899537065,
+      "grad_norm": 3.011537137280228,
+      "learning_rate": 3.4183386437094084e-06,
+      "loss": 0.6504,
+      "num_input_tokens_seen": 47845990,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2717489328443456,
+      "grad_norm": 2.7906706769725194,
+      "learning_rate": 3.417789331369565e-06,
+      "loss": 0.8177,
+      "num_input_tokens_seen": 47861500,
+      "step": 2260
+    },
+    {
+      "epoch": 0.27186917573498465,
+      "grad_norm": 1.9057125778202275,
+      "learning_rate": 3.4172398039522088e-06,
+      "loss": 0.9029,
+      "num_input_tokens_seen": 47882505,
+      "step": 2261
+    },
+    {
+      "epoch": 0.27198941862562376,
+      "grad_norm": 1.9142659183243795,
+      "learning_rate": 3.4166900615407e-06,
+      "loss": 0.7845,
+      "num_input_tokens_seen": 47900140,
+      "step": 2262
+    },
+    {
+      "epoch": 0.27210966151626287,
+      "grad_norm": 2.1066841094433126,
+      "learning_rate": 3.416140104218436e-06,
+      "loss": 0.741,
+      "num_input_tokens_seen": 47919225,
+      "step": 2263
+    },
+    {
+      "epoch": 0.2722299044069019,
+      "grad_norm": 0.8934330366747231,
+      "learning_rate": 3.4155899320688437e-06,
+      "loss": 0.7417,
+      "num_input_tokens_seen": 47985020,
+      "step": 2264
+    },
+    {
+      "epoch": 0.27235014729754103,
+      "grad_norm": 2.0616460731994066,
+      "learning_rate": 3.415039545175384e-06,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 48000465,
+      "step": 2265
+    },
+    {
+      "epoch": 0.27247039018818014,
+      "grad_norm": 2.328166864699395,
+      "learning_rate": 3.414488943621551e-06,
+      "loss": 0.6555,
+      "num_input_tokens_seen": 48018850,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2725906330788192,
+      "grad_norm": 2.1410102835554357,
+      "learning_rate": 3.41393812749087e-06,
+      "loss": 0.7303,
+      "num_input_tokens_seen": 48036615,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2727108759694583,
+      "grad_norm": 2.8251924852307178,
+      "learning_rate": 3.4133870968668984e-06,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 48051135,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2728311188600974,
+      "grad_norm": 1.6637437816768221,
+      "learning_rate": 3.412835851833229e-06,
+      "loss": 0.7771,
+      "num_input_tokens_seen": 48073050,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2729513617507365,
+      "grad_norm": 1.862581230659382,
+      "learning_rate": 3.4122843924734834e-06,
+      "loss": 0.7712,
+      "num_input_tokens_seen": 48095070,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2730716046413756,
+      "grad_norm": 2.0034830680973386,
+      "learning_rate": 3.411732718871319e-06,
+      "loss": 0.8828,
+      "num_input_tokens_seen": 48110630,
+      "step": 2271
+    },
+    {
+      "epoch": 0.27319184753201464,
+      "grad_norm": 1.464338193958209,
+      "learning_rate": 3.4111808311104227e-06,
+      "loss": 0.7769,
+      "num_input_tokens_seen": 48132665,
+      "step": 2272
+    },
+    {
+      "epoch": 0.27331209042265375,
+      "grad_norm": 1.7927893769858916,
+      "learning_rate": 3.4106287292745174e-06,
+      "loss": 0.6859,
+      "num_input_tokens_seen": 48153905,
+      "step": 2273
+    },
+    {
+      "epoch": 0.27343233331329286,
+      "grad_norm": 2.043283388861086,
+      "learning_rate": 3.4100764134473546e-06,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 48172910,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2735525762039319,
+      "grad_norm": 2.6610707435458765,
+      "learning_rate": 3.4095238837127215e-06,
+      "loss": 0.8544,
+      "num_input_tokens_seen": 48191770,
+      "step": 2275
+    },
+    {
+      "epoch": 0.27367281909457103,
+      "grad_norm": 1.9454266750925413,
+      "learning_rate": 3.4089711401544355e-06,
+      "loss": 0.7939,
+      "num_input_tokens_seen": 48209085,
+      "step": 2276
+    },
+    {
+      "epoch": 0.27379306198521014,
+      "grad_norm": 2.151313775388218,
+      "learning_rate": 3.4084181828563486e-06,
+      "loss": 0.671,
+      "num_input_tokens_seen": 48225525,
+      "step": 2277
+    },
+    {
+      "epoch": 0.2739133048758492,
+      "grad_norm": 1.9683967155289295,
+      "learning_rate": 3.4078650119023423e-06,
+      "loss": 0.7022,
+      "num_input_tokens_seen": 48243560,
+      "step": 2278
+    },
+    {
+      "epoch": 0.2740335477664883,
+      "grad_norm": 2.0107365978161567,
+      "learning_rate": 3.4073116273763337e-06,
+      "loss": 0.7442,
+      "num_input_tokens_seen": 48257725,
+      "step": 2279
+    },
+    {
+      "epoch": 0.2741537906571274,
+      "grad_norm": 2.291958059631235,
+      "learning_rate": 3.40675802936227e-06,
+      "loss": 0.8057,
+      "num_input_tokens_seen": 48278230,
+      "step": 2280
+    },
+    {
+      "epoch": 0.27427403354776647,
+      "grad_norm": 1.9577781658485498,
+      "learning_rate": 3.4062042179441318e-06,
+      "loss": 0.7076,
+      "num_input_tokens_seen": 48298420,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2743942764384056,
+      "grad_norm": 1.8264070487480006,
+      "learning_rate": 3.4056501932059314e-06,
+      "loss": 0.8075,
+      "num_input_tokens_seen": 48316215,
+      "step": 2282
+    },
+    {
+      "epoch": 0.2745145193290447,
+      "grad_norm": 0.8532713687739912,
+      "learning_rate": 3.405095955231715e-06,
+      "loss": 0.6085,
+      "num_input_tokens_seen": 48367590,
+      "step": 2283
+    },
+    {
+      "epoch": 0.27463476221968375,
+      "grad_norm": 2.051499555812841,
+      "learning_rate": 3.4045415041055585e-06,
+      "loss": 0.9423,
+      "num_input_tokens_seen": 48382950,
+      "step": 2284
+    },
+    {
+      "epoch": 0.27475500511032286,
+      "grad_norm": 2.890748148251801,
+      "learning_rate": 3.403986839911573e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 48397310,
+      "step": 2285
+    },
+    {
+      "epoch": 0.27487524800096197,
+      "grad_norm": 1.8465120291999164,
+      "learning_rate": 3.4034319627339003e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 48413895,
+      "step": 2286
+    },
+    {
+      "epoch": 0.274995490891601,
+      "grad_norm": 2.4133902134408074,
+      "learning_rate": 3.402876872656715e-06,
+      "loss": 0.6887,
+      "num_input_tokens_seen": 48431935,
+      "step": 2287
+    },
+    {
+      "epoch": 0.27511573378224013,
+      "grad_norm": 2.9979038913290865,
+      "learning_rate": 3.402321569764223e-06,
+      "loss": 0.8912,
+      "num_input_tokens_seen": 48450960,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2752359766728792,
+      "grad_norm": 1.7818168492005624,
+      "learning_rate": 3.4017660541406635e-06,
+      "loss": 0.8285,
+      "num_input_tokens_seen": 48466745,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2753562195635183,
+      "grad_norm": 1.901667196276985,
+      "learning_rate": 3.4012103258703092e-06,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 48485220,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2754764624541574,
+      "grad_norm": 2.423705664961733,
+      "learning_rate": 3.4006543850374616e-06,
+      "loss": 0.8227,
+      "num_input_tokens_seen": 48499990,
+      "step": 2291
+    },
+    {
+      "epoch": 0.27559670534479647,
+      "grad_norm": 2.160714636846373,
+      "learning_rate": 3.4000982317264577e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 48516810,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2757169482354356,
+      "grad_norm": 1.9161741555199356,
+      "learning_rate": 3.3995418660216657e-06,
+      "loss": 0.8733,
+      "num_input_tokens_seen": 48533985,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2758371911260747,
+      "grad_norm": 2.316333046390821,
+      "learning_rate": 3.3989852880074848e-06,
+      "loss": 0.8086,
+      "num_input_tokens_seen": 48555135,
+      "step": 2294
+    },
+    {
+      "epoch": 0.27595743401671374,
+      "grad_norm": 0.9008967053694367,
+      "learning_rate": 3.398428497768348e-06,
+      "loss": 0.6403,
+      "num_input_tokens_seen": 48620025,
+      "step": 2295
+    },
+    {
+      "epoch": 0.27607767690735285,
+      "grad_norm": 1.8778933751956248,
+      "learning_rate": 3.3978714953887205e-06,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 48639500,
+      "step": 2296
+    },
+    {
+      "epoch": 0.27619791979799196,
+      "grad_norm": 6.191129429766541,
+      "learning_rate": 3.397314280953098e-06,
+      "loss": 0.8588,
+      "num_input_tokens_seen": 48660045,
+      "step": 2297
+    },
+    {
+      "epoch": 0.276318162688631,
+      "grad_norm": 2.001912299228255,
+      "learning_rate": 3.3967568545460108e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 48679305,
+      "step": 2298
+    },
+    {
+      "epoch": 0.27643840557927013,
+      "grad_norm": 1.7700210410343649,
+      "learning_rate": 3.396199216252019e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 48697650,
+      "step": 2299
+    },
+    {
+      "epoch": 0.27655864846990924,
+      "grad_norm": 2.1802939960234893,
+      "learning_rate": 3.3956413661557156e-06,
+      "loss": 0.7199,
+      "num_input_tokens_seen": 48717545,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2766788913605483,
+      "grad_norm": 2.81793463226604,
+      "learning_rate": 3.3950833043417273e-06,
+      "loss": 0.6572,
+      "num_input_tokens_seen": 48735410,
+      "step": 2301
+    },
+    {
+      "epoch": 0.2767991342511874,
+      "grad_norm": 3.328374460944523,
+      "learning_rate": 3.3945250308947105e-06,
+      "loss": 0.7133,
+      "num_input_tokens_seen": 48751435,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2769193771418265,
+      "grad_norm": 1.3003075002858855,
+      "learning_rate": 3.3939665458993556e-06,
+      "loss": 0.7,
+      "num_input_tokens_seen": 48805575,
+      "step": 2303
+    },
+    {
+      "epoch": 0.27703962003246557,
+      "grad_norm": 2.744390874938876,
+      "learning_rate": 3.393407849440384e-06,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 48824870,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2771598629231047,
+      "grad_norm": 3.6795682436540846,
+      "learning_rate": 3.3928489416025495e-06,
+      "loss": 0.8048,
+      "num_input_tokens_seen": 48845435,
+      "step": 2305
+    },
+    {
+      "epoch": 0.27728010581374374,
+      "grad_norm": 2.203084394611633,
+      "learning_rate": 3.392289822470638e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 48863135,
+      "step": 2306
+    },
+    {
+      "epoch": 0.27740034870438285,
+      "grad_norm": 2.140315436788674,
+      "learning_rate": 3.3917304921294674e-06,
+      "loss": 0.7575,
+      "num_input_tokens_seen": 48881020,
+      "step": 2307
+    },
+    {
+      "epoch": 0.27752059159502196,
+      "grad_norm": 1.654651789632888,
+      "learning_rate": 3.3911709506638876e-06,
+      "loss": 0.8068,
+      "num_input_tokens_seen": 48900050,
+      "step": 2308
+    },
+    {
+      "epoch": 0.277640834485661,
+      "grad_norm": 2.103030776941744,
+      "learning_rate": 3.390611198158781e-06,
+      "loss": 0.8044,
+      "num_input_tokens_seen": 48917645,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2777610773763001,
+      "grad_norm": 2.6317138466191543,
+      "learning_rate": 3.3900512346990612e-06,
+      "loss": 0.8928,
+      "num_input_tokens_seen": 48933355,
+      "step": 2310
+    },
+    {
+      "epoch": 0.27788132026693924,
+      "grad_norm": 2.1834016492938506,
+      "learning_rate": 3.389491060369674e-06,
+      "loss": 0.654,
+      "num_input_tokens_seen": 48958750,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2780015631575783,
+      "grad_norm": 2.0215184062906264,
+      "learning_rate": 3.388930675255598e-06,
+      "loss": 0.8906,
+      "num_input_tokens_seen": 48978320,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2781218060482174,
+      "grad_norm": 8.766957940392066,
+      "learning_rate": 3.388370079441843e-06,
+      "loss": 0.7838,
+      "num_input_tokens_seen": 48993555,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2782420489388565,
+      "grad_norm": 2.0819072496483115,
+      "learning_rate": 3.3878092730134505e-06,
+      "loss": 0.9243,
+      "num_input_tokens_seen": 49011260,
+      "step": 2314
+    },
+    {
+      "epoch": 0.27836229182949557,
+      "grad_norm": 1.7454940400959307,
+      "learning_rate": 3.3872482560554947e-06,
+      "loss": 0.8007,
+      "num_input_tokens_seen": 49029755,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2784825347201347,
+      "grad_norm": 0.8398309645351633,
+      "learning_rate": 3.386687028653082e-06,
+      "loss": 0.5924,
+      "num_input_tokens_seen": 49092320,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2786027776107738,
+      "grad_norm": 2.0516560227825167,
+      "learning_rate": 3.386125590891349e-06,
+      "loss": 0.8418,
+      "num_input_tokens_seen": 49108915,
+      "step": 2317
+    },
+    {
+      "epoch": 0.27872302050141284,
+      "grad_norm": 2.195437433171907,
+      "learning_rate": 3.3855639428554657e-06,
+      "loss": 0.8304,
+      "num_input_tokens_seen": 49126165,
+      "step": 2318
+    },
+    {
+      "epoch": 0.27884326339205195,
+      "grad_norm": 2.0714147772337754,
+      "learning_rate": 3.385002084630635e-06,
+      "loss": 0.8106,
+      "num_input_tokens_seen": 49144855,
+      "step": 2319
+    },
+    {
+      "epoch": 0.278963506282691,
+      "grad_norm": 2.5695153849915977,
+      "learning_rate": 3.384440016302088e-06,
+      "loss": 0.8394,
+      "num_input_tokens_seen": 49163250,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2790837491733301,
+      "grad_norm": 2.87457309783845,
+      "learning_rate": 3.3838777379550923e-06,
+      "loss": 0.6156,
+      "num_input_tokens_seen": 49182415,
+      "step": 2321
+    },
+    {
+      "epoch": 0.27920399206396923,
+      "grad_norm": 2.0630528174721885,
+      "learning_rate": 3.3833152496749434e-06,
+      "loss": 0.787,
+      "num_input_tokens_seen": 49200700,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2793242349546083,
+      "grad_norm": 2.948427561828374,
+      "learning_rate": 3.3827525515469715e-06,
+      "loss": 0.8547,
+      "num_input_tokens_seen": 49215325,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2794444778452474,
+      "grad_norm": 2.216959619145687,
+      "learning_rate": 3.3821896436565367e-06,
+      "loss": 0.7018,
+      "num_input_tokens_seen": 49234705,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2795647207358865,
+      "grad_norm": 1.9411968826961243,
+      "learning_rate": 3.381626526089032e-06,
+      "loss": 0.6943,
+      "num_input_tokens_seen": 49253990,
+      "step": 2325
+    },
+    {
+      "epoch": 0.27968496362652556,
+      "grad_norm": 2.456723916220405,
+      "learning_rate": 3.3810631989298815e-06,
+      "loss": 0.7844,
+      "num_input_tokens_seen": 49273320,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2798052065171647,
+      "grad_norm": 2.653837432607224,
+      "learning_rate": 3.3804996622645423e-06,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 49291040,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2799254494078038,
+      "grad_norm": 1.735249267645477,
+      "learning_rate": 3.3799359161785015e-06,
+      "loss": 0.882,
+      "num_input_tokens_seen": 49310410,
+      "step": 2328
+    },
+    {
+      "epoch": 0.28004569229844284,
+      "grad_norm": 1.6241011759720148,
+      "learning_rate": 3.3793719607572798e-06,
+      "loss": 0.8458,
+      "num_input_tokens_seen": 49331095,
+      "step": 2329
+    },
+    {
+      "epoch": 0.28016593518908195,
+      "grad_norm": 1.9786870178526337,
+      "learning_rate": 3.378807796086428e-06,
+      "loss": 0.7723,
+      "num_input_tokens_seen": 49353675,
+      "step": 2330
+    },
+    {
+      "epoch": 0.28028617807972106,
+      "grad_norm": 2.1043044457698215,
+      "learning_rate": 3.37824342225153e-06,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 49369815,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2804064209703601,
+      "grad_norm": 1.8449009584367904,
+      "learning_rate": 3.3776788393382006e-06,
+      "loss": 0.7674,
+      "num_input_tokens_seen": 49389015,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2805266638609992,
+      "grad_norm": 2.3814322280863753,
+      "learning_rate": 3.3771140474320872e-06,
+      "loss": 0.7667,
+      "num_input_tokens_seen": 49408685,
+      "step": 2333
+    },
+    {
+      "epoch": 0.28064690675163834,
+      "grad_norm": 2.1504838537722177,
+      "learning_rate": 3.3765490466188664e-06,
+      "loss": 0.7865,
+      "num_input_tokens_seen": 49425805,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2807671496422774,
+      "grad_norm": 2.6568371678436744,
+      "learning_rate": 3.375983836984251e-06,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 49443600,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2808873925329165,
+      "grad_norm": 2.1006114655073653,
+      "learning_rate": 3.3754184186139807e-06,
+      "loss": 0.7305,
+      "num_input_tokens_seen": 49462345,
+      "step": 2336
+    },
+    {
+      "epoch": 0.28100763542355556,
+      "grad_norm": 2.193732071946392,
+      "learning_rate": 3.374852791593831e-06,
+      "loss": 0.8277,
+      "num_input_tokens_seen": 49478265,
+      "step": 2337
+    },
+    {
+      "epoch": 0.28112787831419467,
+      "grad_norm": 4.2055493465037745,
+      "learning_rate": 3.3742869560096047e-06,
+      "loss": 0.5424,
+      "num_input_tokens_seen": 49496550,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2812481212048338,
+      "grad_norm": 1.9337409973651492,
+      "learning_rate": 3.3737209119471405e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 49512780,
+      "step": 2339
+    },
+    {
+      "epoch": 0.28136836409547283,
+      "grad_norm": 3.6805333674022838,
+      "learning_rate": 3.373154659492306e-06,
+      "loss": 0.6438,
+      "num_input_tokens_seen": 49530640,
+      "step": 2340
+    },
+    {
+      "epoch": 0.28148860698611194,
+      "grad_norm": 1.9683957509887322,
+      "learning_rate": 3.3725881987310016e-06,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 49547895,
+      "step": 2341
+    },
+    {
+      "epoch": 0.28160884987675106,
+      "grad_norm": 2.2372899926232854,
+      "learning_rate": 3.372021529749159e-06,
+      "loss": 0.8735,
+      "num_input_tokens_seen": 49566675,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2817290927673901,
+      "grad_norm": 1.7335200184222752,
+      "learning_rate": 3.3714546526327405e-06,
+      "loss": 0.9167,
+      "num_input_tokens_seen": 49584395,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2818493356580292,
+      "grad_norm": 2.1711987886453206,
+      "learning_rate": 3.3708875674677423e-06,
+      "loss": 0.8815,
+      "num_input_tokens_seen": 49602090,
+      "step": 2344
+    },
+    {
+      "epoch": 0.28196957854866833,
+      "grad_norm": 2.051798326515003,
+      "learning_rate": 3.37032027434019e-06,
+      "loss": 0.825,
+      "num_input_tokens_seen": 49621330,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2820898214393074,
+      "grad_norm": 1.8933125668368698,
+      "learning_rate": 3.369752773336141e-06,
+      "loss": 0.8329,
+      "num_input_tokens_seen": 49640530,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2822100643299465,
+      "grad_norm": 1.593694445478885,
+      "learning_rate": 3.3691850645416864e-06,
+      "loss": 0.7748,
+      "num_input_tokens_seen": 49659960,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2823303072205856,
+      "grad_norm": 4.588622696084842,
+      "learning_rate": 3.368617148042945e-06,
+      "loss": 0.8283,
+      "num_input_tokens_seen": 49677350,
+      "step": 2348
+    },
+    {
+      "epoch": 0.28245055011122466,
+      "grad_norm": 1.7289211220494785,
+      "learning_rate": 3.3680490239260707e-06,
+      "loss": 0.8434,
+      "num_input_tokens_seen": 49696065,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2825707930018638,
+      "grad_norm": 1.5039041006252167,
+      "learning_rate": 3.3674806922772476e-06,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 49716670,
+      "step": 2350
+    },
+    {
+      "epoch": 0.28269103589250283,
+      "grad_norm": 1.7468866401604508,
+      "learning_rate": 3.366912153182691e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 49737370,
+      "step": 2351
+    },
+    {
+      "epoch": 0.28281127878314194,
+      "grad_norm": 2.2261955112367935,
+      "learning_rate": 3.366343406728647e-06,
+      "loss": 0.8306,
+      "num_input_tokens_seen": 49756540,
+      "step": 2352
+    },
+    {
+      "epoch": 0.28293152167378105,
+      "grad_norm": 2.0896194097100445,
+      "learning_rate": 3.365774453001395e-06,
+      "loss": 0.6811,
+      "num_input_tokens_seen": 49775495,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2830517645644201,
+      "grad_norm": 2.1200228273944766,
+      "learning_rate": 3.3652052920872437e-06,
+      "loss": 0.7157,
+      "num_input_tokens_seen": 49798080,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2831720074550592,
+      "grad_norm": 2.9015622918028314,
+      "learning_rate": 3.3646359240725355e-06,
+      "loss": 0.8491,
+      "num_input_tokens_seen": 49816990,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2832922503456983,
+      "grad_norm": 2.0546911706134647,
+      "learning_rate": 3.364066349043643e-06,
+      "loss": 0.6794,
+      "num_input_tokens_seen": 49837915,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2834124932363374,
+      "grad_norm": 1.625026343171864,
+      "learning_rate": 3.3634965670869695e-06,
+      "loss": 0.8225,
+      "num_input_tokens_seen": 49854730,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2835327361269765,
+      "grad_norm": 2.0339618866187283,
+      "learning_rate": 3.3629265782889506e-06,
+      "loss": 0.7495,
+      "num_input_tokens_seen": 49876275,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2836529790176156,
+      "grad_norm": 2.331083866823628,
+      "learning_rate": 3.362356382736054e-06,
+      "loss": 0.7115,
+      "num_input_tokens_seen": 49896600,
+      "step": 2359
+    },
+    {
+      "epoch": 0.28377322190825466,
+      "grad_norm": 2.3154376182729304,
+      "learning_rate": 3.361785980514777e-06,
+      "loss": 0.9085,
+      "num_input_tokens_seen": 49912520,
+      "step": 2360
+    },
+    {
+      "epoch": 0.28389346479889377,
+      "grad_norm": 1.9159869804799416,
+      "learning_rate": 3.361215371711649e-06,
+      "loss": 0.7672,
+      "num_input_tokens_seen": 49931335,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2840137076895329,
+      "grad_norm": 1.7415243885138814,
+      "learning_rate": 3.3606445564132326e-06,
+      "loss": 0.8238,
+      "num_input_tokens_seen": 49948350,
+      "step": 2362
+    },
+    {
+      "epoch": 0.28413395058017193,
+      "grad_norm": 2.1226279951787825,
+      "learning_rate": 3.360073534706118e-06,
+      "loss": 0.8088,
+      "num_input_tokens_seen": 49965225,
+      "step": 2363
+    },
+    {
+      "epoch": 0.28425419347081105,
+      "grad_norm": 2.048902717683217,
+      "learning_rate": 3.35950230667693e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 49986640,
+      "step": 2364
+    },
+    {
+      "epoch": 0.28437443636145016,
+      "grad_norm": 2.180278951163515,
+      "learning_rate": 3.358930872412323e-06,
+      "loss": 0.8594,
+      "num_input_tokens_seen": 50003525,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2844946792520892,
+      "grad_norm": 1.6277711045186845,
+      "learning_rate": 3.3583592319989825e-06,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 50022615,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2846149221427283,
+      "grad_norm": 2.995159725165637,
+      "learning_rate": 3.357787385523627e-06,
+      "loss": 0.6801,
+      "num_input_tokens_seen": 50043740,
+      "step": 2367
+    },
+    {
+      "epoch": 0.2847351650333674,
+      "grad_norm": 1.8263151665286457,
+      "learning_rate": 3.3572153330730048e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 50064555,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2848554079240065,
+      "grad_norm": 0.8245052233677317,
+      "learning_rate": 3.3566430747338956e-06,
+      "loss": 0.6663,
+      "num_input_tokens_seen": 50119480,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2849756508146456,
+      "grad_norm": 2.037816991470415,
+      "learning_rate": 3.35607061059311e-06,
+      "loss": 0.8651,
+      "num_input_tokens_seen": 50134130,
+      "step": 2370
+    },
+    {
+      "epoch": 0.28509589370528465,
+      "grad_norm": 2.0180644576480113,
+      "learning_rate": 3.3554979407374917e-06,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 50155960,
+      "step": 2371
+    },
+    {
+      "epoch": 0.28521613659592376,
+      "grad_norm": 1.458449352781362,
+      "learning_rate": 3.354925065253913e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 50174775,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2853363794865629,
+      "grad_norm": 1.8653316341988153,
+      "learning_rate": 3.3543519842292794e-06,
+      "loss": 0.8135,
+      "num_input_tokens_seen": 50194150,
+      "step": 2373
+    },
+    {
+      "epoch": 0.28545662237720193,
+      "grad_norm": 1.9161792419834263,
+      "learning_rate": 3.353778697750527e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 50212275,
+      "step": 2374
+    },
+    {
+      "epoch": 0.28557686526784104,
+      "grad_norm": 2.31139332373254,
+      "learning_rate": 3.3532052059046224e-06,
+      "loss": 0.886,
+      "num_input_tokens_seen": 50231105,
+      "step": 2375
+    },
+    {
+      "epoch": 0.28569710815848015,
+      "grad_norm": 2.03694748948585,
+      "learning_rate": 3.3526315087785637e-06,
+      "loss": 0.7238,
+      "num_input_tokens_seen": 50251940,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2858173510491192,
+      "grad_norm": 1.9488155469221076,
+      "learning_rate": 3.3520576064593805e-06,
+      "loss": 0.8044,
+      "num_input_tokens_seen": 50271615,
+      "step": 2377
+    },
+    {
+      "epoch": 0.2859375939397583,
+      "grad_norm": 2.6390625272226833,
+      "learning_rate": 3.3514834990341337e-06,
+      "loss": 0.8176,
+      "num_input_tokens_seen": 50291660,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2860578368303974,
+      "grad_norm": 3.4807822414807865,
+      "learning_rate": 3.3509091865899144e-06,
+      "loss": 0.9207,
+      "num_input_tokens_seen": 50306570,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2861780797210365,
+      "grad_norm": 1.8557746554040935,
+      "learning_rate": 3.350334669213846e-06,
+      "loss": 0.6995,
+      "num_input_tokens_seen": 50323695,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2862983226116756,
+      "grad_norm": 2.205574963023131,
+      "learning_rate": 3.3497599469930816e-06,
+      "loss": 0.765,
+      "num_input_tokens_seen": 50341625,
+      "step": 2381
+    },
+    {
+      "epoch": 0.28641856550231465,
+      "grad_norm": 2.8725603321002007,
+      "learning_rate": 3.349185020014807e-06,
+      "loss": 0.8266,
+      "num_input_tokens_seen": 50358610,
+      "step": 2382
+    },
+    {
+      "epoch": 0.28653880839295376,
+      "grad_norm": 4.057371987433365,
+      "learning_rate": 3.348609888366237e-06,
+      "loss": 0.7447,
+      "num_input_tokens_seen": 50377345,
+      "step": 2383
+    },
+    {
+      "epoch": 0.28665905128359287,
+      "grad_norm": 2.129213760912391,
+      "learning_rate": 3.348034552134619e-06,
+      "loss": 0.6367,
+      "num_input_tokens_seen": 50396470,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2867792941742319,
+      "grad_norm": 1.8525430496598638,
+      "learning_rate": 3.3474590114072316e-06,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 50414190,
+      "step": 2385
+    },
+    {
+      "epoch": 0.28689953706487104,
+      "grad_norm": 2.111382287681849,
+      "learning_rate": 3.3468832662713836e-06,
+      "loss": 0.8302,
+      "num_input_tokens_seen": 50432155,
+      "step": 2386
+    },
+    {
+      "epoch": 0.28701977995551015,
+      "grad_norm": 2.2048254544299914,
+      "learning_rate": 3.346307316814415e-06,
+      "loss": 0.8395,
+      "num_input_tokens_seen": 50447045,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2871400228461492,
+      "grad_norm": 2.0161948321955636,
+      "learning_rate": 3.3457311631236965e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 50467750,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2872602657367883,
+      "grad_norm": 1.6782248205404353,
+      "learning_rate": 3.345154805286631e-06,
+      "loss": 0.841,
+      "num_input_tokens_seen": 50487730,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2873805086274274,
+      "grad_norm": 2.3343942253584467,
+      "learning_rate": 3.344578243390651e-06,
+      "loss": 0.7561,
+      "num_input_tokens_seen": 50503010,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2875007515180665,
+      "grad_norm": 2.4883317336903685,
+      "learning_rate": 3.3440014775232206e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 50520785,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2876209944087056,
+      "grad_norm": 1.8567321108212271,
+      "learning_rate": 3.343424507771834e-06,
+      "loss": 0.716,
+      "num_input_tokens_seen": 50538715,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2877412372993447,
+      "grad_norm": 1.7237803187885317,
+      "learning_rate": 3.342847334224018e-06,
+      "loss": 0.8664,
+      "num_input_tokens_seen": 50555835,
+      "step": 2393
+    },
+    {
+      "epoch": 0.28786148018998375,
+      "grad_norm": 0.8601963083552077,
+      "learning_rate": 3.342269956967329e-06,
+      "loss": 0.6685,
+      "num_input_tokens_seen": 50617460,
+      "step": 2394
+    },
+    {
+      "epoch": 0.28798172308062286,
+      "grad_norm": 2.756407830338849,
+      "learning_rate": 3.341692376089355e-06,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 50632735,
+      "step": 2395
+    },
+    {
+      "epoch": 0.288101965971262,
+      "grad_norm": 4.997115905569317,
+      "learning_rate": 3.3411145916777146e-06,
+      "loss": 0.8358,
+      "num_input_tokens_seen": 50646615,
+      "step": 2396
+    },
+    {
+      "epoch": 0.28822220886190103,
+      "grad_norm": 2.8799385138002234,
+      "learning_rate": 3.3405366038200566e-06,
+      "loss": 0.9075,
+      "num_input_tokens_seen": 50665270,
+      "step": 2397
+    },
+    {
+      "epoch": 0.28834245175254014,
+      "grad_norm": 3.1408771963981925,
+      "learning_rate": 3.3399584126040617e-06,
+      "loss": 0.8429,
+      "num_input_tokens_seen": 50684490,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2884626946431792,
+      "grad_norm": 1.7716305417768738,
+      "learning_rate": 3.339380018117441e-06,
+      "loss": 0.8947,
+      "num_input_tokens_seen": 50705045,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2885829375338183,
+      "grad_norm": 2.3509406794548253,
+      "learning_rate": 3.3388014204479366e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 50722570,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2887031804244574,
+      "grad_norm": 1.9856856668262237,
+      "learning_rate": 3.338222619683321e-06,
+      "loss": 0.9074,
+      "num_input_tokens_seen": 50742255,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2888234233150965,
+      "grad_norm": 2.0710029298337296,
+      "learning_rate": 3.337643615911398e-06,
+      "loss": 0.7392,
+      "num_input_tokens_seen": 50761600,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2889436662057356,
+      "grad_norm": 2.271643134441523,
+      "learning_rate": 3.3370644092200026e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 50778595,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2890639090963747,
+      "grad_norm": 1.8935822366119175,
+      "learning_rate": 3.3364849996969985e-06,
+      "loss": 0.7848,
+      "num_input_tokens_seen": 50798335,
+      "step": 2404
+    },
+    {
+      "epoch": 0.28918415198701375,
+      "grad_norm": 2.203306085437626,
+      "learning_rate": 3.335905387430283e-06,
+      "loss": 0.8455,
+      "num_input_tokens_seen": 50819490,
+      "step": 2405
+    },
+    {
+      "epoch": 0.28930439487765286,
+      "grad_norm": 2.4950887348299546,
+      "learning_rate": 3.335325572507782e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 50839710,
+      "step": 2406
+    },
+    {
+      "epoch": 0.28942463776829197,
+      "grad_norm": 1.6024700502263713,
+      "learning_rate": 3.3347455550174537e-06,
+      "loss": 0.7422,
+      "num_input_tokens_seen": 50858770,
+      "step": 2407
+    },
+    {
+      "epoch": 0.289544880658931,
+      "grad_norm": 2.1834084625158177,
+      "learning_rate": 3.3341653350472864e-06,
+      "loss": 0.684,
+      "num_input_tokens_seen": 50875320,
+      "step": 2408
+    },
+    {
+      "epoch": 0.28966512354957014,
+      "grad_norm": 2.7441756714514915,
+      "learning_rate": 3.333584912685298e-06,
+      "loss": 0.7012,
+      "num_input_tokens_seen": 50893660,
+      "step": 2409
+    },
+    {
+      "epoch": 0.28978536644020925,
+      "grad_norm": 0.9030873691811463,
+      "learning_rate": 3.3330042880195385e-06,
+      "loss": 0.5874,
+      "num_input_tokens_seen": 50947730,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2899056093308483,
+      "grad_norm": 2.7922675411835867,
+      "learning_rate": 3.3324234611380888e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 50966180,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2900258522214874,
+      "grad_norm": 1.7662829775399878,
+      "learning_rate": 3.3318424321290596e-06,
+      "loss": 0.8154,
+      "num_input_tokens_seen": 50985615,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2901460951121265,
+      "grad_norm": 0.8971180930924121,
+      "learning_rate": 3.3312612010805917e-06,
+      "loss": 0.6452,
+      "num_input_tokens_seen": 51044910,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2902663380027656,
+      "grad_norm": 1.8514742773590152,
+      "learning_rate": 3.330679768080858e-06,
+      "loss": 0.6922,
+      "num_input_tokens_seen": 51068515,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2903865808934047,
+      "grad_norm": 2.396960076832458,
+      "learning_rate": 3.3300981332180627e-06,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 51087440,
+      "step": 2415
+    },
+    {
+      "epoch": 0.29050682378404374,
+      "grad_norm": 2.0592319946698776,
+      "learning_rate": 3.3295162965804373e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 51105655,
+      "step": 2416
+    },
+    {
+      "epoch": 0.29062706667468285,
+      "grad_norm": 2.0730644262668476,
+      "learning_rate": 3.328934258256247e-06,
+      "loss": 0.7721,
+      "num_input_tokens_seen": 51123440,
+      "step": 2417
+    },
+    {
+      "epoch": 0.29074730956532197,
+      "grad_norm": 2.315997863697858,
+      "learning_rate": 3.3283520183337856e-06,
+      "loss": 0.6665,
+      "num_input_tokens_seen": 51142865,
+      "step": 2418
+    },
+    {
+      "epoch": 0.290867552455961,
+      "grad_norm": 1.6396967508811393,
+      "learning_rate": 3.32776957690138e-06,
+      "loss": 0.692,
+      "num_input_tokens_seen": 51162030,
+      "step": 2419
+    },
+    {
+      "epoch": 0.29098779534660013,
+      "grad_norm": 1.9017461653737526,
+      "learning_rate": 3.327186934047385e-06,
+      "loss": 0.7664,
+      "num_input_tokens_seen": 51180445,
+      "step": 2420
+    },
+    {
+      "epoch": 0.29110803823723924,
+      "grad_norm": 11.288210420598688,
+      "learning_rate": 3.3266040898601877e-06,
+      "loss": 0.6601,
+      "num_input_tokens_seen": 51198000,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2912282811278783,
+      "grad_norm": 1.8089073991980655,
+      "learning_rate": 3.3260210444282045e-06,
+      "loss": 0.7775,
+      "num_input_tokens_seen": 51215675,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2913485240185174,
+      "grad_norm": 2.067262619173759,
+      "learning_rate": 3.325437797839883e-06,
+      "loss": 0.7237,
+      "num_input_tokens_seen": 51233765,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2914687669091565,
+      "grad_norm": 3.9177865054011183,
+      "learning_rate": 3.3248543501837015e-06,
+      "loss": 0.7518,
+      "num_input_tokens_seen": 51250690,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2915890097997956,
+      "grad_norm": 2.1706685490162774,
+      "learning_rate": 3.3242707015481684e-06,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 51270345,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2917092526904347,
+      "grad_norm": 1.7188174916915615,
+      "learning_rate": 3.323686852021823e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 51287575,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2918294955810738,
+      "grad_norm": 2.288009234157425,
+      "learning_rate": 3.323102801693235e-06,
+      "loss": 0.7929,
+      "num_input_tokens_seen": 51306060,
+      "step": 2427
+    },
+    {
+      "epoch": 0.29194973847171285,
+      "grad_norm": 2.153413574201045,
+      "learning_rate": 3.3225185506510025e-06,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 51325090,
+      "step": 2428
+    },
+    {
+      "epoch": 0.29206998136235196,
+      "grad_norm": 1.8872514057508358,
+      "learning_rate": 3.3219340989837586e-06,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 51344800,
+      "step": 2429
+    },
+    {
+      "epoch": 0.292190224252991,
+      "grad_norm": 1.9173651755054195,
+      "learning_rate": 3.3213494467801625e-06,
+      "loss": 0.8064,
+      "num_input_tokens_seen": 51363695,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2923104671436301,
+      "grad_norm": 2.1108424370028094,
+      "learning_rate": 3.3207645941289063e-06,
+      "loss": 0.7172,
+      "num_input_tokens_seen": 51381760,
+      "step": 2431
+    },
+    {
+      "epoch": 0.29243071003426924,
+      "grad_norm": 3.879031149823903,
+      "learning_rate": 3.320179541118711e-06,
+      "loss": 0.8025,
+      "num_input_tokens_seen": 51403980,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2925509529249083,
+      "grad_norm": 1.0909535952449185,
+      "learning_rate": 3.3195942878383293e-06,
+      "loss": 0.6435,
+      "num_input_tokens_seen": 51459800,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2926711958155474,
+      "grad_norm": 1.7813342450566485,
+      "learning_rate": 3.319008834376543e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 51479210,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2927914387061865,
+      "grad_norm": 2.4602465599161265,
+      "learning_rate": 3.3184231808221654e-06,
+      "loss": 0.8752,
+      "num_input_tokens_seen": 51493255,
+      "step": 2435
+    },
+    {
+      "epoch": 0.29291168159682557,
+      "grad_norm": 2.400940382068107,
+      "learning_rate": 3.3178373272640394e-06,
+      "loss": 0.6267,
+      "num_input_tokens_seen": 51512070,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2930319244874647,
+      "grad_norm": 2.1780229300924576,
+      "learning_rate": 3.317251273791039e-06,
+      "loss": 0.8618,
+      "num_input_tokens_seen": 51529300,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2931521673781038,
+      "grad_norm": 2.2260427300901813,
+      "learning_rate": 3.316665020492067e-06,
+      "loss": 0.882,
+      "num_input_tokens_seen": 51550190,
+      "step": 2438
+    },
+    {
+      "epoch": 0.29327241026874284,
+      "grad_norm": 1.627963780846525,
+      "learning_rate": 3.316078567456059e-06,
+      "loss": 0.8149,
+      "num_input_tokens_seen": 51567750,
+      "step": 2439
+    },
+    {
+      "epoch": 0.29339265315938196,
+      "grad_norm": 1.860293357065652,
+      "learning_rate": 3.3154919147719786e-06,
+      "loss": 0.7649,
+      "num_input_tokens_seen": 51588485,
+      "step": 2440
+    },
+    {
+      "epoch": 0.29351289605002107,
+      "grad_norm": 1.8099545327764344,
+      "learning_rate": 3.3149050625288206e-06,
+      "loss": 0.8714,
+      "num_input_tokens_seen": 51607585,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2936331389406601,
+      "grad_norm": 1.740896376267365,
+      "learning_rate": 3.31431801081561e-06,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 51626240,
+      "step": 2442
+    },
+    {
+      "epoch": 0.29375338183129923,
+      "grad_norm": 1.0780926283131673,
+      "learning_rate": 3.313730759721402e-06,
+      "loss": 0.6853,
+      "num_input_tokens_seen": 51688890,
+      "step": 2443
+    },
+    {
+      "epoch": 0.29387362472193834,
+      "grad_norm": 2.05148341379875,
+      "learning_rate": 3.313143309335282e-06,
+      "loss": 0.8552,
+      "num_input_tokens_seen": 51707100,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2939938676125774,
+      "grad_norm": 1.7016368137711781,
+      "learning_rate": 3.3125556597463665e-06,
+      "loss": 0.8336,
+      "num_input_tokens_seen": 51726125,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2941141105032165,
+      "grad_norm": 1.4523775894637847,
+      "learning_rate": 3.311967811043801e-06,
+      "loss": 0.6535,
+      "num_input_tokens_seen": 51747765,
+      "step": 2446
+    },
+    {
+      "epoch": 0.29423435339385556,
+      "grad_norm": 2.612657832494984,
+      "learning_rate": 3.3113797633167617e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 51765780,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2943545962844947,
+      "grad_norm": 2.112544091494324,
+      "learning_rate": 3.310791516654455e-06,
+      "loss": 0.683,
+      "num_input_tokens_seen": 51782560,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2944748391751338,
+      "grad_norm": 2.1895601178464603,
+      "learning_rate": 3.3102030711461177e-06,
+      "loss": 0.7939,
+      "num_input_tokens_seen": 51801855,
+      "step": 2449
+    },
+    {
+      "epoch": 0.29459508206577284,
+      "grad_norm": 2.035623488088594,
+      "learning_rate": 3.3096144268810156e-06,
+      "loss": 0.6792,
+      "num_input_tokens_seen": 51820335,
+      "step": 2450
+    },
+    {
+      "epoch": 0.29471532495641195,
+      "grad_norm": 2.5482851314735293,
+      "learning_rate": 3.3090255839484462e-06,
+      "loss": 0.7168,
+      "num_input_tokens_seen": 51838050,
+      "step": 2451
+    },
+    {
+      "epoch": 0.29483556784705106,
+      "grad_norm": 1.7241812729777966,
+      "learning_rate": 3.3084365424377366e-06,
+      "loss": 0.8518,
+      "num_input_tokens_seen": 51856535,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2949558107376901,
+      "grad_norm": 0.7890756144280393,
+      "learning_rate": 3.307847302438245e-06,
+      "loss": 0.5799,
+      "num_input_tokens_seen": 51910235,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2950760536283292,
+      "grad_norm": 2.339193335608346,
+      "learning_rate": 3.307257864039356e-06,
+      "loss": 0.7758,
+      "num_input_tokens_seen": 51927290,
+      "step": 2454
+    },
+    {
+      "epoch": 0.29519629651896834,
+      "grad_norm": 1.7898455396707815,
+      "learning_rate": 3.306668227330489e-06,
+      "loss": 0.7877,
+      "num_input_tokens_seen": 51944655,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2953165394096074,
+      "grad_norm": 1.9943872322131644,
+      "learning_rate": 3.3060783924010904e-06,
+      "loss": 0.7724,
+      "num_input_tokens_seen": 51962300,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2954367823002465,
+      "grad_norm": 2.1246258453013955,
+      "learning_rate": 3.3054883593406387e-06,
+      "loss": 0.8452,
+      "num_input_tokens_seen": 51976770,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2955570251908856,
+      "grad_norm": 2.648904710154184,
+      "learning_rate": 3.3048981282386404e-06,
+      "loss": 0.6413,
+      "num_input_tokens_seen": 51997800,
+      "step": 2458
+    },
+    {
+      "epoch": 0.29567726808152467,
+      "grad_norm": 2.185750167176726,
+      "learning_rate": 3.304307699184634e-06,
+      "loss": 0.8218,
+      "num_input_tokens_seen": 52016110,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2957975109721638,
+      "grad_norm": 1.5885380689557118,
+      "learning_rate": 3.3037170722681866e-06,
+      "loss": 0.7912,
+      "num_input_tokens_seen": 52036665,
+      "step": 2460
+    },
+    {
+      "epoch": 0.29591775386280283,
+      "grad_norm": 1.9911603431610132,
+      "learning_rate": 3.3031262475788956e-06,
+      "loss": 0.6869,
+      "num_input_tokens_seen": 52053325,
+      "step": 2461
+    },
+    {
+      "epoch": 0.29603799675344195,
+      "grad_norm": 1.990018127730373,
+      "learning_rate": 3.3025352252063897e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 52071740,
+      "step": 2462
+    },
+    {
+      "epoch": 0.29615823964408106,
+      "grad_norm": 2.485793276782081,
+      "learning_rate": 3.3019440052403252e-06,
+      "loss": 0.7478,
+      "num_input_tokens_seen": 52091325,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2962784825347201,
+      "grad_norm": 2.474399644067521,
+      "learning_rate": 3.30135258777039e-06,
+      "loss": 0.7083,
+      "num_input_tokens_seen": 52110415,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2963987254253592,
+      "grad_norm": 2.1437613636519677,
+      "learning_rate": 3.3007609728863024e-06,
+      "loss": 0.6958,
+      "num_input_tokens_seen": 52128225,
+      "step": 2465
+    },
+    {
+      "epoch": 0.29651896831599833,
+      "grad_norm": 1.7764703039792589,
+      "learning_rate": 3.300169160677809e-06,
+      "loss": 0.7265,
+      "num_input_tokens_seen": 52151860,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2966392112066374,
+      "grad_norm": 2.4051179307053316,
+      "learning_rate": 3.299577151234688e-06,
+      "loss": 0.7712,
+      "num_input_tokens_seen": 52169930,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2967594540972765,
+      "grad_norm": 2.5196428611488666,
+      "learning_rate": 3.298984944646746e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 52188330,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2968796969879156,
+      "grad_norm": 1.9804512974836999,
+      "learning_rate": 3.298392541003822e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 52207455,
+      "step": 2469
+    },
+    {
+      "epoch": 0.29699993987855466,
+      "grad_norm": 1.6473123172237887,
+      "learning_rate": 3.2977999403957806e-06,
+      "loss": 0.8867,
+      "num_input_tokens_seen": 52225935,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2971201827691938,
+      "grad_norm": 2.152949661252587,
+      "learning_rate": 3.2972071429125207e-06,
+      "loss": 0.6688,
+      "num_input_tokens_seen": 52246875,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2972404256598329,
+      "grad_norm": 2.0907392195718306,
+      "learning_rate": 3.2966141486439682e-06,
+      "loss": 0.8814,
+      "num_input_tokens_seen": 52265785,
+      "step": 2472
+    },
+    {
+      "epoch": 0.29736066855047194,
+      "grad_norm": 3.0497037353033973,
+      "learning_rate": 3.29602095768008e-06,
+      "loss": 0.6507,
+      "num_input_tokens_seen": 52286020,
+      "step": 2473
+    },
+    {
+      "epoch": 0.29748091144111105,
+      "grad_norm": 1.7916149757883848,
+      "learning_rate": 3.2954275701108437e-06,
+      "loss": 0.6354,
+      "num_input_tokens_seen": 52306920,
+      "step": 2474
+    },
+    {
+      "epoch": 0.29760115433175016,
+      "grad_norm": 3.04305122713625,
+      "learning_rate": 3.294833986026275e-06,
+      "loss": 0.6865,
+      "num_input_tokens_seen": 52329880,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2977213972223892,
+      "grad_norm": 1.911358301830498,
+      "learning_rate": 3.2942402055164197e-06,
+      "loss": 0.8544,
+      "num_input_tokens_seen": 52348235,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2978416401130283,
+      "grad_norm": 3.4287036945148817,
+      "learning_rate": 3.2936462286713546e-06,
+      "loss": 0.7068,
+      "num_input_tokens_seen": 52366305,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2979618830036674,
+      "grad_norm": 1.857681706163987,
+      "learning_rate": 3.2930520555811846e-06,
+      "loss": 0.7702,
+      "num_input_tokens_seen": 52385650,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2980821258943065,
+      "grad_norm": 1.9187220620943757,
+      "learning_rate": 3.292457686336046e-06,
+      "loss": 0.7943,
+      "num_input_tokens_seen": 52404690,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2982023687849456,
+      "grad_norm": 0.83889210324733,
+      "learning_rate": 3.291863121026105e-06,
+      "loss": 0.647,
+      "num_input_tokens_seen": 52468190,
+      "step": 2480
+    },
+    {
+      "epoch": 0.29832261167558466,
+      "grad_norm": 2.08143999539811,
+      "learning_rate": 3.291268359741555e-06,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 52491995,
+      "step": 2481
+    },
+    {
+      "epoch": 0.29844285456622377,
+      "grad_norm": 2.0939528029122765,
+      "learning_rate": 3.2906734025726213e-06,
+      "loss": 0.7832,
+      "num_input_tokens_seen": 52510980,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2985630974568629,
+      "grad_norm": 2.482796718990891,
+      "learning_rate": 3.290078249609559e-06,
+      "loss": 0.8749,
+      "num_input_tokens_seen": 52530120,
+      "step": 2483
+    },
+    {
+      "epoch": 0.29868334034750194,
+      "grad_norm": 2.287437119430872,
+      "learning_rate": 3.2894829009426514e-06,
+      "loss": 0.8733,
+      "num_input_tokens_seen": 52547675,
+      "step": 2484
+    },
+    {
+      "epoch": 0.29880358323814105,
+      "grad_norm": 2.0172152854300367,
+      "learning_rate": 3.288887356662213e-06,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 52568730,
+      "step": 2485
+    },
+    {
+      "epoch": 0.29892382612878016,
+      "grad_norm": 0.9688866212572336,
+      "learning_rate": 3.288291616858588e-06,
+      "loss": 0.6127,
+      "num_input_tokens_seen": 52623840,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2990440690194192,
+      "grad_norm": 1.751736459862644,
+      "learning_rate": 3.287695681622149e-06,
+      "loss": 0.7675,
+      "num_input_tokens_seen": 52642910,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2991643119100583,
+      "grad_norm": 11.101667752029197,
+      "learning_rate": 3.2870995510432982e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 52661110,
+      "step": 2488
+    },
+    {
+      "epoch": 0.29928455480069743,
+      "grad_norm": 1.9221717110801153,
+      "learning_rate": 3.2865032252124697e-06,
+      "loss": 0.7636,
+      "num_input_tokens_seen": 52681345,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2994047976913365,
+      "grad_norm": 1.604462359663118,
+      "learning_rate": 3.2859067042201243e-06,
+      "loss": 0.7684,
+      "num_input_tokens_seen": 52703105,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2995250405819756,
+      "grad_norm": 2.015002945199685,
+      "learning_rate": 3.2853099881567544e-06,
+      "loss": 0.768,
+      "num_input_tokens_seen": 52721225,
+      "step": 2491
+    },
+    {
+      "epoch": 0.29964528347261465,
+      "grad_norm": 2.4415352218107262,
+      "learning_rate": 3.284713077112881e-06,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 52740375,
+      "step": 2492
+    },
+    {
+      "epoch": 0.29976552636325376,
+      "grad_norm": 3.0967529356376633,
+      "learning_rate": 3.284115971179056e-06,
+      "loss": 0.8609,
+      "num_input_tokens_seen": 52754125,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2998857692538929,
+      "grad_norm": 1.8933973354510039,
+      "learning_rate": 3.283518670445859e-06,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 52771755,
+      "step": 2494
+    },
+    {
+      "epoch": 0.30000601214453193,
+      "grad_norm": 0.7803843287154192,
+      "learning_rate": 3.2829211750038995e-06,
+      "loss": 0.5716,
+      "num_input_tokens_seen": 52840105,
+      "step": 2495
+    },
+    {
+      "epoch": 0.30012625503517104,
+      "grad_norm": 1.8411967158197522,
+      "learning_rate": 3.2823234849438183e-06,
+      "loss": 0.8758,
+      "num_input_tokens_seen": 52857860,
+      "step": 2496
+    },
+    {
+      "epoch": 0.30024649792581015,
+      "grad_norm": 3.748644010109616,
+      "learning_rate": 3.281725600356284e-06,
+      "loss": 0.7408,
+      "num_input_tokens_seen": 52877955,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3003667408164492,
+      "grad_norm": 2.0917554133314273,
+      "learning_rate": 3.281127521331995e-06,
+      "loss": 0.6641,
+      "num_input_tokens_seen": 52898855,
+      "step": 2498
+    },
+    {
+      "epoch": 0.3004869837070883,
+      "grad_norm": 0.8967966368909067,
+      "learning_rate": 3.2805292479616798e-06,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 52957440,
+      "step": 2499
+    },
+    {
+      "epoch": 0.30060722659772743,
+      "grad_norm": 3.4630790154255675,
+      "learning_rate": 3.2799307803360955e-06,
+      "loss": 0.9137,
+      "num_input_tokens_seen": 52973090,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3007274694883665,
+      "grad_norm": 1.4162701957226582,
+      "learning_rate": 3.27933211854603e-06,
+      "loss": 0.811,
+      "num_input_tokens_seen": 52991865,
+      "step": 2501
+    },
+    {
+      "epoch": 0.3008477123790056,
+      "grad_norm": 1.7011882807860528,
+      "learning_rate": 3.278733262682299e-06,
+      "loss": 0.8623,
+      "num_input_tokens_seen": 53009440,
+      "step": 2502
+    },
+    {
+      "epoch": 0.3009679552696447,
+      "grad_norm": 2.3751784450082543,
+      "learning_rate": 3.278134212835749e-06,
+      "loss": 0.8151,
+      "num_input_tokens_seen": 53028515,
+      "step": 2503
+    },
+    {
+      "epoch": 0.30108819816028376,
+      "grad_norm": 2.1668155791787616,
+      "learning_rate": 3.2775349690972547e-06,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 53042385,
+      "step": 2504
+    },
+    {
+      "epoch": 0.30120844105092287,
+      "grad_norm": 0.9486798203615799,
+      "learning_rate": 3.276935531557722e-06,
+      "loss": 0.5637,
+      "num_input_tokens_seen": 53107325,
+      "step": 2505
+    },
+    {
+      "epoch": 0.301328683941562,
+      "grad_norm": 2.4167899835553324,
+      "learning_rate": 3.2763359003080833e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 53124000,
+      "step": 2506
+    },
+    {
+      "epoch": 0.30144892683220104,
+      "grad_norm": 0.9779399931882427,
+      "learning_rate": 3.2757360754393047e-06,
+      "loss": 0.6735,
+      "num_input_tokens_seen": 53187790,
+      "step": 2507
+    },
+    {
+      "epoch": 0.30156916972284015,
+      "grad_norm": 2.378923066630368,
+      "learning_rate": 3.2751360570423767e-06,
+      "loss": 0.637,
+      "num_input_tokens_seen": 53205895,
+      "step": 2508
+    },
+    {
+      "epoch": 0.3016894126134792,
+      "grad_norm": 2.431824990597903,
+      "learning_rate": 3.2745358452083236e-06,
+      "loss": 0.7553,
+      "num_input_tokens_seen": 53228515,
+      "step": 2509
+    },
+    {
+      "epoch": 0.3018096555041183,
+      "grad_norm": 1.337845004814104,
+      "learning_rate": 3.2739354400281955e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 53249455,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3019298983947574,
+      "grad_norm": 0.9845911362696367,
+      "learning_rate": 3.2733348415930744e-06,
+      "loss": 0.6883,
+      "num_input_tokens_seen": 53311045,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3020501412853965,
+      "grad_norm": 2.1019031194051676,
+      "learning_rate": 3.27273404999407e-06,
+      "loss": 0.814,
+      "num_input_tokens_seen": 53332985,
+      "step": 2512
+    },
+    {
+      "epoch": 0.3021703841760356,
+      "grad_norm": 0.8197240349656524,
+      "learning_rate": 3.272133065322322e-06,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 53390975,
+      "step": 2513
+    },
+    {
+      "epoch": 0.3022906270666747,
+      "grad_norm": 1.7445627268185064,
+      "learning_rate": 3.271531887669e-06,
+      "loss": 0.7833,
+      "num_input_tokens_seen": 53410755,
+      "step": 2514
+    },
+    {
+      "epoch": 0.30241086995731375,
+      "grad_norm": 2.4318481507289484,
+      "learning_rate": 3.2709305171253015e-06,
+      "loss": 0.6357,
+      "num_input_tokens_seen": 53430595,
+      "step": 2515
+    },
+    {
+      "epoch": 0.30253111284795287,
+      "grad_norm": 1.8127750855080975,
+      "learning_rate": 3.270328953782453e-06,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 53450115,
+      "step": 2516
+    },
+    {
+      "epoch": 0.302651355738592,
+      "grad_norm": 3.0239995682327114,
+      "learning_rate": 3.2697271977317137e-06,
+      "loss": 0.7878,
+      "num_input_tokens_seen": 53462600,
+      "step": 2517
+    },
+    {
+      "epoch": 0.30277159862923103,
+      "grad_norm": 1.7550830901960661,
+      "learning_rate": 3.269125249064367e-06,
+      "loss": 0.7742,
+      "num_input_tokens_seen": 53482015,
+      "step": 2518
+    },
+    {
+      "epoch": 0.30289184151987014,
+      "grad_norm": 6.035178215518512,
+      "learning_rate": 3.26852310787173e-06,
+      "loss": 0.8282,
+      "num_input_tokens_seen": 53501925,
+      "step": 2519
+    },
+    {
+      "epoch": 0.30301208441050925,
+      "grad_norm": 1.98512020886652,
+      "learning_rate": 3.267920774245145e-06,
+      "loss": 0.7412,
+      "num_input_tokens_seen": 53521050,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3031323273011483,
+      "grad_norm": 3.2800418889949605,
+      "learning_rate": 3.267318248275988e-06,
+      "loss": 0.841,
+      "num_input_tokens_seen": 53539885,
+      "step": 2521
+    },
+    {
+      "epoch": 0.3032525701917874,
+      "grad_norm": 2.500616261269978,
+      "learning_rate": 3.266715530055659e-06,
+      "loss": 0.6561,
+      "num_input_tokens_seen": 53557755,
+      "step": 2522
+    },
+    {
+      "epoch": 0.30337281308242653,
+      "grad_norm": 1.7281600443888308,
+      "learning_rate": 3.2661126196755927e-06,
+      "loss": 0.8001,
+      "num_input_tokens_seen": 53576585,
+      "step": 2523
+    },
+    {
+      "epoch": 0.3034930559730656,
+      "grad_norm": 1.1425042884390375,
+      "learning_rate": 3.265509517227248e-06,
+      "loss": 0.6048,
+      "num_input_tokens_seen": 53633120,
+      "step": 2524
+    },
+    {
+      "epoch": 0.3036132988637047,
+      "grad_norm": 1.9615512724426913,
+      "learning_rate": 3.2649062228021154e-06,
+      "loss": 0.7983,
+      "num_input_tokens_seen": 53650690,
+      "step": 2525
+    },
+    {
+      "epoch": 0.30373354175434375,
+      "grad_norm": 2.1809817491609684,
+      "learning_rate": 3.2643027364917145e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 53670530,
+      "step": 2526
+    },
+    {
+      "epoch": 0.30385378464498286,
+      "grad_norm": 1.8760831990023663,
+      "learning_rate": 3.263699058387594e-06,
+      "loss": 0.8585,
+      "num_input_tokens_seen": 53687685,
+      "step": 2527
+    },
+    {
+      "epoch": 0.30397402753562197,
+      "grad_norm": 2.1256814571681844,
+      "learning_rate": 3.2630951885813315e-06,
+      "loss": 0.9042,
+      "num_input_tokens_seen": 53704800,
+      "step": 2528
+    },
+    {
+      "epoch": 0.304094270426261,
+      "grad_norm": 1.9710145323797088,
+      "learning_rate": 3.262491127164533e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 53723335,
+      "step": 2529
+    },
+    {
+      "epoch": 0.30421451331690014,
+      "grad_norm": 2.47045534128996,
+      "learning_rate": 3.2618868742288337e-06,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 53739980,
+      "step": 2530
+    },
+    {
+      "epoch": 0.30433475620753925,
+      "grad_norm": 3.483552643808775,
+      "learning_rate": 3.261282429865899e-06,
+      "loss": 0.7268,
+      "num_input_tokens_seen": 53757705,
+      "step": 2531
+    },
+    {
+      "epoch": 0.3044549990981783,
+      "grad_norm": 1.8065887059284331,
+      "learning_rate": 3.2606777941674225e-06,
+      "loss": 0.723,
+      "num_input_tokens_seen": 53776080,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3045752419888174,
+      "grad_norm": 2.155245326436034,
+      "learning_rate": 3.2600729672251276e-06,
+      "loss": 0.8364,
+      "num_input_tokens_seen": 53793515,
+      "step": 2533
+    },
+    {
+      "epoch": 0.3046954848794565,
+      "grad_norm": 2.7798033321988713,
+      "learning_rate": 3.259467949130765e-06,
+      "loss": 0.6544,
+      "num_input_tokens_seen": 53814645,
+      "step": 2534
+    },
+    {
+      "epoch": 0.3048157277700956,
+      "grad_norm": 3.329007330186855,
+      "learning_rate": 3.2588627399761164e-06,
+      "loss": 0.8325,
+      "num_input_tokens_seen": 53830360,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3049359706607347,
+      "grad_norm": 1.6784768787855624,
+      "learning_rate": 3.2582573398529903e-06,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 53847435,
+      "step": 2536
+    },
+    {
+      "epoch": 0.3050562135513738,
+      "grad_norm": 2.726315746049457,
+      "learning_rate": 3.2576517488532265e-06,
+      "loss": 0.7432,
+      "num_input_tokens_seen": 53863505,
+      "step": 2537
+    },
+    {
+      "epoch": 0.30517645644201286,
+      "grad_norm": 1.6952580132360224,
+      "learning_rate": 3.257045967068692e-06,
+      "loss": 0.8602,
+      "num_input_tokens_seen": 53882480,
+      "step": 2538
+    },
+    {
+      "epoch": 0.30529669933265197,
+      "grad_norm": 1.8713597780020776,
+      "learning_rate": 3.2564399945912848e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 53901990,
+      "step": 2539
+    },
+    {
+      "epoch": 0.305416942223291,
+      "grad_norm": 2.372692098960834,
+      "learning_rate": 3.2558338315129287e-06,
+      "loss": 0.8159,
+      "num_input_tokens_seen": 53919855,
+      "step": 2540
+    },
+    {
+      "epoch": 0.30553718511393013,
+      "grad_norm": 1.9393419318240728,
+      "learning_rate": 3.2552274779255785e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 53940505,
+      "step": 2541
+    },
+    {
+      "epoch": 0.30565742800456924,
+      "grad_norm": 2.2690890857295574,
+      "learning_rate": 3.2546209339212184e-06,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 53959245,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3057776708952083,
+      "grad_norm": 1.8853416978386144,
+      "learning_rate": 3.25401419959186e-06,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 53979575,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3058979137858474,
+      "grad_norm": 1.9475239188249012,
+      "learning_rate": 3.253407275029545e-06,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 53998200,
+      "step": 2544
+    },
+    {
+      "epoch": 0.3060181566764865,
+      "grad_norm": 2.1327806900429143,
+      "learning_rate": 3.2528001603263425e-06,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 54019990,
+      "step": 2545
+    },
+    {
+      "epoch": 0.3061383995671256,
+      "grad_norm": 1.863831689282036,
+      "learning_rate": 3.2521928555743514e-06,
+      "loss": 0.8104,
+      "num_input_tokens_seen": 54037055,
+      "step": 2546
+    },
+    {
+      "epoch": 0.3062586424577647,
+      "grad_norm": 2.3708063763844365,
+      "learning_rate": 3.251585360865701e-06,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 54054775,
+      "step": 2547
+    },
+    {
+      "epoch": 0.3063788853484038,
+      "grad_norm": 2.6478345952131472,
+      "learning_rate": 3.250977676292545e-06,
+      "loss": 0.7493,
+      "num_input_tokens_seen": 54072735,
+      "step": 2548
+    },
+    {
+      "epoch": 0.30649912823904285,
+      "grad_norm": 2.013453104655136,
+      "learning_rate": 3.2503698019470712e-06,
+      "loss": 0.7878,
+      "num_input_tokens_seen": 54088225,
+      "step": 2549
+    },
+    {
+      "epoch": 0.30661937112968196,
+      "grad_norm": 2.2244452127710765,
+      "learning_rate": 3.249761737921492e-06,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 54104475,
+      "step": 2550
+    },
+    {
+      "epoch": 0.30673961402032107,
+      "grad_norm": 2.0166115706203906,
+      "learning_rate": 3.249153484308051e-06,
+      "loss": 0.7423,
+      "num_input_tokens_seen": 54122810,
+      "step": 2551
+    },
+    {
+      "epoch": 0.3068598569109601,
+      "grad_norm": 3.05913576240648,
+      "learning_rate": 3.2485450411990194e-06,
+      "loss": 0.7744,
+      "num_input_tokens_seen": 54141885,
+      "step": 2552
+    },
+    {
+      "epoch": 0.30698009980159924,
+      "grad_norm": 6.694322438531677,
+      "learning_rate": 3.2479364086866983e-06,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 54161860,
+      "step": 2553
+    },
+    {
+      "epoch": 0.30710034269223835,
+      "grad_norm": 1.9947849475630965,
+      "learning_rate": 3.247327586863416e-06,
+      "loss": 0.8066,
+      "num_input_tokens_seen": 54182460,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3072205855828774,
+      "grad_norm": 3.2180216038254907,
+      "learning_rate": 3.2467185758215304e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 54201920,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3073408284735165,
+      "grad_norm": 2.6023250757146714,
+      "learning_rate": 3.246109375653428e-06,
+      "loss": 0.8481,
+      "num_input_tokens_seen": 54218405,
+      "step": 2556
+    },
+    {
+      "epoch": 0.30746107136415557,
+      "grad_norm": 1.9423469434797664,
+      "learning_rate": 3.2454999864515243e-06,
+      "loss": 0.7801,
+      "num_input_tokens_seen": 54237500,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3075813142547947,
+      "grad_norm": 2.021568294685621,
+      "learning_rate": 3.244890408308263e-06,
+      "loss": 0.6856,
+      "num_input_tokens_seen": 54257925,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3077015571454338,
+      "grad_norm": 3.0515948112502467,
+      "learning_rate": 3.2442806413161165e-06,
+      "loss": 0.6104,
+      "num_input_tokens_seen": 54277290,
+      "step": 2559
+    },
+    {
+      "epoch": 0.30782180003607285,
+      "grad_norm": 2.1490474422507195,
+      "learning_rate": 3.2436706855675856e-06,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 54294410,
+      "step": 2560
+    },
+    {
+      "epoch": 0.30794204292671196,
+      "grad_norm": 3.010655875980436,
+      "learning_rate": 3.2430605411552012e-06,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 54314245,
+      "step": 2561
+    },
+    {
+      "epoch": 0.30806228581735107,
+      "grad_norm": 0.8946030648815743,
+      "learning_rate": 3.2424502081715205e-06,
+      "loss": 0.7157,
+      "num_input_tokens_seen": 54377080,
+      "step": 2562
+    },
+    {
+      "epoch": 0.3081825287079901,
+      "grad_norm": 2.2138366673970027,
+      "learning_rate": 3.241839686709132e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 54397735,
+      "step": 2563
+    },
+    {
+      "epoch": 0.30830277159862923,
+      "grad_norm": 3.0243852354493845,
+      "learning_rate": 3.2412289768606495e-06,
+      "loss": 0.8195,
+      "num_input_tokens_seen": 54414025,
+      "step": 2564
+    },
+    {
+      "epoch": 0.30842301448926834,
+      "grad_norm": 1.6179158188721627,
+      "learning_rate": 3.240618078718718e-06,
+      "loss": 0.8168,
+      "num_input_tokens_seen": 54435205,
+      "step": 2565
+    },
+    {
+      "epoch": 0.3085432573799074,
+      "grad_norm": 2.0768916323215336,
+      "learning_rate": 3.240006992376011e-06,
+      "loss": 0.7389,
+      "num_input_tokens_seen": 54454550,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3086635002705465,
+      "grad_norm": 2.473560702463057,
+      "learning_rate": 3.2393957179252284e-06,
+      "loss": 0.7532,
+      "num_input_tokens_seen": 54470805,
+      "step": 2567
+    },
+    {
+      "epoch": 0.3087837431611856,
+      "grad_norm": 2.591371038379765,
+      "learning_rate": 3.2387842554591016e-06,
+      "loss": 0.805,
+      "num_input_tokens_seen": 54491340,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3089039860518247,
+      "grad_norm": 2.248663049410497,
+      "learning_rate": 3.238172605070388e-06,
+      "loss": 0.8656,
+      "num_input_tokens_seen": 54506475,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3090242289424638,
+      "grad_norm": 2.407108476749192,
+      "learning_rate": 3.2375607668518745e-06,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 54519230,
+      "step": 2570
+    },
+    {
+      "epoch": 0.30914447183310284,
+      "grad_norm": 2.1303557265605098,
+      "learning_rate": 3.236948740896377e-06,
+      "loss": 0.8958,
+      "num_input_tokens_seen": 54533750,
+      "step": 2571
+    },
+    {
+      "epoch": 0.30926471472374195,
+      "grad_norm": 1.3948547590926945,
+      "learning_rate": 3.2363365272967384e-06,
+      "loss": 0.836,
+      "num_input_tokens_seen": 54556040,
+      "step": 2572
+    },
+    {
+      "epoch": 0.30938495761438106,
+      "grad_norm": 2.564899293169109,
+      "learning_rate": 3.235724126145832e-06,
+      "loss": 0.8058,
+      "num_input_tokens_seen": 54571795,
+      "step": 2573
+    },
+    {
+      "epoch": 0.3095052005050201,
+      "grad_norm": 1.7378805506607162,
+      "learning_rate": 3.235111537536558e-06,
+      "loss": 0.7723,
+      "num_input_tokens_seen": 54592330,
+      "step": 2574
+    },
+    {
+      "epoch": 0.30962544339565923,
+      "grad_norm": 2.289403195162928,
+      "learning_rate": 3.2344987615618456e-06,
+      "loss": 0.8289,
+      "num_input_tokens_seen": 54611885,
+      "step": 2575
+    },
+    {
+      "epoch": 0.30974568628629834,
+      "grad_norm": 2.024184283736538,
+      "learning_rate": 3.2338857983146533e-06,
+      "loss": 0.7819,
+      "num_input_tokens_seen": 54633105,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3098659291769374,
+      "grad_norm": 2.2308105918247305,
+      "learning_rate": 3.233272647887966e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 54651715,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3099861720675765,
+      "grad_norm": 1.4902083919867717,
+      "learning_rate": 3.2326593103747985e-06,
+      "loss": 0.8881,
+      "num_input_tokens_seen": 54670450,
+      "step": 2578
+    },
+    {
+      "epoch": 0.3101064149582156,
+      "grad_norm": 2.287505695932403,
+      "learning_rate": 3.2320457858681936e-06,
+      "loss": 0.8442,
+      "num_input_tokens_seen": 54688560,
+      "step": 2579
+    },
+    {
+      "epoch": 0.31022665784885467,
+      "grad_norm": 2.482510842208662,
+      "learning_rate": 3.2314320744612228e-06,
+      "loss": 0.8551,
+      "num_input_tokens_seen": 54703580,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3103469007394938,
+      "grad_norm": 1.9490418576007904,
+      "learning_rate": 3.2308181762469854e-06,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 54721245,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3104671436301329,
+      "grad_norm": 8.407402748575718,
+      "learning_rate": 3.230204091318609e-06,
+      "loss": 0.7869,
+      "num_input_tokens_seen": 54741505,
+      "step": 2582
+    },
+    {
+      "epoch": 0.31058738652077195,
+      "grad_norm": 2.0418739515989945,
+      "learning_rate": 3.2295898197692503e-06,
+      "loss": 0.8416,
+      "num_input_tokens_seen": 54760185,
+      "step": 2583
+    },
+    {
+      "epoch": 0.31070762941141106,
+      "grad_norm": 1.6905667288903958,
+      "learning_rate": 3.2289753616920935e-06,
+      "loss": 0.7874,
+      "num_input_tokens_seen": 54780925,
+      "step": 2584
+    },
+    {
+      "epoch": 0.31082787230205017,
+      "grad_norm": 3.1509423699755432,
+      "learning_rate": 3.228360717180352e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 54798170,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3109481151926892,
+      "grad_norm": 0.8804646717585519,
+      "learning_rate": 3.227745886327266e-06,
+      "loss": 0.6421,
+      "num_input_tokens_seen": 54856585,
+      "step": 2586
+    },
+    {
+      "epoch": 0.31106835808332833,
+      "grad_norm": 0.8218910219540508,
+      "learning_rate": 3.227130869226105e-06,
+      "loss": 0.5873,
+      "num_input_tokens_seen": 54913465,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3111886009739674,
+      "grad_norm": 2.663260104192462,
+      "learning_rate": 3.226515665970167e-06,
+      "loss": 0.8147,
+      "num_input_tokens_seen": 54930725,
+      "step": 2588
+    },
+    {
+      "epoch": 0.3113088438646065,
+      "grad_norm": 2.3708506181821973,
+      "learning_rate": 3.225900276652777e-06,
+      "loss": 0.8579,
+      "num_input_tokens_seen": 54947220,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3114290867552456,
+      "grad_norm": 1.7211735780761468,
+      "learning_rate": 3.2252847013672906e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 54969600,
+      "step": 2590
+    },
+    {
+      "epoch": 0.31154932964588467,
+      "grad_norm": 2.3995555681460847,
+      "learning_rate": 3.224668940207089e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 54988305,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3116695725365238,
+      "grad_norm": 2.493748193692999,
+      "learning_rate": 3.2240529932655828e-06,
+      "loss": 0.8646,
+      "num_input_tokens_seen": 55007290,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3117898154271629,
+      "grad_norm": 3.5062824869753424,
+      "learning_rate": 3.223436860636211e-06,
+      "loss": 0.8855,
+      "num_input_tokens_seen": 55022645,
+      "step": 2593
+    },
+    {
+      "epoch": 0.31191005831780194,
+      "grad_norm": 1.7783212237888943,
+      "learning_rate": 3.22282054241244e-06,
+      "loss": 0.7327,
+      "num_input_tokens_seen": 55045520,
+      "step": 2594
+    },
+    {
+      "epoch": 0.31203030120844105,
+      "grad_norm": 7.757792925429537,
+      "learning_rate": 3.222204038687765e-06,
+      "loss": 0.74,
+      "num_input_tokens_seen": 55058375,
+      "step": 2595
+    },
+    {
+      "epoch": 0.31215054409908016,
+      "grad_norm": 1.510014610148318,
+      "learning_rate": 3.221587349555709e-06,
+      "loss": 0.8743,
+      "num_input_tokens_seen": 55078355,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3122707869897192,
+      "grad_norm": 2.004749260195451,
+      "learning_rate": 3.2209704751098236e-06,
+      "loss": 0.685,
+      "num_input_tokens_seen": 55097105,
+      "step": 2597
+    },
+    {
+      "epoch": 0.31239102988035833,
+      "grad_norm": 2.870185452203878,
+      "learning_rate": 3.2203534154436875e-06,
+      "loss": 0.8292,
+      "num_input_tokens_seen": 55111180,
+      "step": 2598
+    },
+    {
+      "epoch": 0.31251127277099744,
+      "grad_norm": 2.723044984494789,
+      "learning_rate": 3.219736170650909e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 55131655,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3126315156616365,
+      "grad_norm": 3.0518605470923363,
+      "learning_rate": 3.2191187408251228e-06,
+      "loss": 0.8437,
+      "num_input_tokens_seen": 55148535,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3127517585522756,
+      "grad_norm": 2.2688737869722426,
+      "learning_rate": 3.218501126059993e-06,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 55163650,
+      "step": 2601
+    },
+    {
+      "epoch": 0.31287200144291466,
+      "grad_norm": 2.886450743997347,
+      "learning_rate": 3.2178833264492116e-06,
+      "loss": 0.8054,
+      "num_input_tokens_seen": 55182075,
+      "step": 2602
+    },
+    {
+      "epoch": 0.31299224433355377,
+      "grad_norm": 1.9127437455544876,
+      "learning_rate": 3.217265342086498e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 55202285,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3131124872241929,
+      "grad_norm": 2.3684801627086416,
+      "learning_rate": 3.216647173065599e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 55217470,
+      "step": 2604
+    },
+    {
+      "epoch": 0.31323273011483194,
+      "grad_norm": 3.034930382059138,
+      "learning_rate": 3.216028819480292e-06,
+      "loss": 0.7337,
+      "num_input_tokens_seen": 55238530,
+      "step": 2605
+    },
+    {
+      "epoch": 0.31335297300547105,
+      "grad_norm": 2.298032041945607,
+      "learning_rate": 3.2154102814243793e-06,
+      "loss": 0.7669,
+      "num_input_tokens_seen": 55257390,
+      "step": 2606
+    },
+    {
+      "epoch": 0.31347321589611016,
+      "grad_norm": 4.813548964592026,
+      "learning_rate": 3.2147915589916937e-06,
+      "loss": 0.671,
+      "num_input_tokens_seen": 55278670,
+      "step": 2607
+    },
+    {
+      "epoch": 0.3135934587867492,
+      "grad_norm": 2.156926740136252,
+      "learning_rate": 3.2141726522760938e-06,
+      "loss": 0.8229,
+      "num_input_tokens_seen": 55296450,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3137137016773883,
+      "grad_norm": 0.7859195132797734,
+      "learning_rate": 3.2135535613714693e-06,
+      "loss": 0.5573,
+      "num_input_tokens_seen": 55359905,
+      "step": 2609
+    },
+    {
+      "epoch": 0.31383394456802743,
+      "grad_norm": 2.8678079467420643,
+      "learning_rate": 3.212934286371733e-06,
+      "loss": 0.9567,
+      "num_input_tokens_seen": 55376335,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3139541874586665,
+      "grad_norm": 2.9905998009294628,
+      "learning_rate": 3.2123148273708304e-06,
+      "loss": 0.8356,
+      "num_input_tokens_seen": 55396245,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3140744303493056,
+      "grad_norm": 2.2035238407951168,
+      "learning_rate": 3.211695184462733e-06,
+      "loss": 0.7571,
+      "num_input_tokens_seen": 55417140,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3141946732399447,
+      "grad_norm": 0.8666997879792196,
+      "learning_rate": 3.2110753577414388e-06,
+      "loss": 0.6396,
+      "num_input_tokens_seen": 55478440,
+      "step": 2613
+    },
+    {
+      "epoch": 0.31431491613058377,
+      "grad_norm": 2.3540436886758336,
+      "learning_rate": 3.2104553473009753e-06,
+      "loss": 0.7882,
+      "num_input_tokens_seen": 55496280,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3144351590212229,
+      "grad_norm": 2.1637888882836487,
+      "learning_rate": 3.209835153235399e-06,
+      "loss": 0.672,
+      "num_input_tokens_seen": 55517555,
+      "step": 2615
+    },
+    {
+      "epoch": 0.314555401911862,
+      "grad_norm": 2.1392589128444257,
+      "learning_rate": 3.2092147756387916e-06,
+      "loss": 0.6819,
+      "num_input_tokens_seen": 55537600,
+      "step": 2616
+    },
+    {
+      "epoch": 0.31467564480250104,
+      "grad_norm": 2.253962196281221,
+      "learning_rate": 3.208594214605264e-06,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 55555865,
+      "step": 2617
+    },
+    {
+      "epoch": 0.31479588769314015,
+      "grad_norm": 2.1776750893003722,
+      "learning_rate": 3.2079734702289553e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 55574480,
+      "step": 2618
+    },
+    {
+      "epoch": 0.3149161305837792,
+      "grad_norm": 0.8662034544883165,
+      "learning_rate": 3.207352542604031e-06,
+      "loss": 0.6405,
+      "num_input_tokens_seen": 55636535,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3150363734744183,
+      "grad_norm": 2.0336266554056017,
+      "learning_rate": 3.2067314318246864e-06,
+      "loss": 0.7775,
+      "num_input_tokens_seen": 55656970,
+      "step": 2620
+    },
+    {
+      "epoch": 0.31515661636505743,
+      "grad_norm": 2.3024180578021176,
+      "learning_rate": 3.206110137985143e-06,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 55676895,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3152768592556965,
+      "grad_norm": 2.9483211351038663,
+      "learning_rate": 3.2054886611796505e-06,
+      "loss": 0.9147,
+      "num_input_tokens_seen": 55695610,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3153971021463356,
+      "grad_norm": 0.9808980100879209,
+      "learning_rate": 3.204867001502487e-06,
+      "loss": 0.6881,
+      "num_input_tokens_seen": 55753985,
+      "step": 2623
+    },
+    {
+      "epoch": 0.3155173450369747,
+      "grad_norm": 1.9068825281597623,
+      "learning_rate": 3.2042451590479567e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 55774220,
+      "step": 2624
+    },
+    {
+      "epoch": 0.31563758792761376,
+      "grad_norm": 1.6909574445999993,
+      "learning_rate": 3.203623133910394e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 55792245,
+      "step": 2625
+    },
+    {
+      "epoch": 0.31575783081825287,
+      "grad_norm": 2.2947670621289595,
+      "learning_rate": 3.203000926184158e-06,
+      "loss": 0.7688,
+      "num_input_tokens_seen": 55810890,
+      "step": 2626
+    },
+    {
+      "epoch": 0.315878073708892,
+      "grad_norm": 1.8833738170149354,
+      "learning_rate": 3.202378535963639e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 55831525,
+      "step": 2627
+    },
+    {
+      "epoch": 0.31599831659953104,
+      "grad_norm": 1.5875752667451526,
+      "learning_rate": 3.2017559633432512e-06,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 55850875,
+      "step": 2628
+    },
+    {
+      "epoch": 0.31611855949017015,
+      "grad_norm": 1.848815010166794,
+      "learning_rate": 3.2011332084174398e-06,
+      "loss": 0.6571,
+      "num_input_tokens_seen": 55871465,
+      "step": 2629
+    },
+    {
+      "epoch": 0.31623880238080926,
+      "grad_norm": 1.6679081573850048,
+      "learning_rate": 3.2005102712806756e-06,
+      "loss": 0.8888,
+      "num_input_tokens_seen": 55890015,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3163590452714483,
+      "grad_norm": 2.630733947119322,
+      "learning_rate": 3.1998871520274575e-06,
+      "loss": 0.7225,
+      "num_input_tokens_seen": 55905070,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3164792881620874,
+      "grad_norm": 1.9738616773715132,
+      "learning_rate": 3.199263850752312e-06,
+      "loss": 0.8435,
+      "num_input_tokens_seen": 55925625,
+      "step": 2632
+    },
+    {
+      "epoch": 0.31659953105272653,
+      "grad_norm": 2.31469742893676,
+      "learning_rate": 3.198640367549795e-06,
+      "loss": 0.8622,
+      "num_input_tokens_seen": 55944240,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3167197739433656,
+      "grad_norm": 1.9467912583415112,
+      "learning_rate": 3.198016702514487e-06,
+      "loss": 0.8582,
+      "num_input_tokens_seen": 55964240,
+      "step": 2634
+    },
+    {
+      "epoch": 0.3168400168340047,
+      "grad_norm": 1.983358707807838,
+      "learning_rate": 3.1973928557409977e-06,
+      "loss": 0.8415,
+      "num_input_tokens_seen": 55982000,
+      "step": 2635
+    },
+    {
+      "epoch": 0.31696025972464376,
+      "grad_norm": 2.3048669051762034,
+      "learning_rate": 3.1967688273239636e-06,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 56001525,
+      "step": 2636
+    },
+    {
+      "epoch": 0.31708050261528287,
+      "grad_norm": 2.55983154032076,
+      "learning_rate": 3.1961446173580503e-06,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 56018185,
+      "step": 2637
+    },
+    {
+      "epoch": 0.317200745505922,
+      "grad_norm": 1.961665145226088,
+      "learning_rate": 3.1955202259379502e-06,
+      "loss": 0.7676,
+      "num_input_tokens_seen": 56039635,
+      "step": 2638
+    },
+    {
+      "epoch": 0.31732098839656103,
+      "grad_norm": 1.9511712643348056,
+      "learning_rate": 3.194895653158381e-06,
+      "loss": 0.8116,
+      "num_input_tokens_seen": 56058295,
+      "step": 2639
+    },
+    {
+      "epoch": 0.31744123128720014,
+      "grad_norm": 0.7836212932774443,
+      "learning_rate": 3.194270899114093e-06,
+      "loss": 0.5941,
+      "num_input_tokens_seen": 56123810,
+      "step": 2640
+    },
+    {
+      "epoch": 0.31756147417783925,
+      "grad_norm": 1.7294662182718141,
+      "learning_rate": 3.1936459638998575e-06,
+      "loss": 0.8195,
+      "num_input_tokens_seen": 56141145,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3176817170684783,
+      "grad_norm": 1.9321852015725185,
+      "learning_rate": 3.193020847610479e-06,
+      "loss": 0.8276,
+      "num_input_tokens_seen": 56161185,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3178019599591174,
+      "grad_norm": 2.6795087348816193,
+      "learning_rate": 3.1923955503407875e-06,
+      "loss": 0.7154,
+      "num_input_tokens_seen": 56178855,
+      "step": 2643
+    },
+    {
+      "epoch": 0.31792220284975653,
+      "grad_norm": 1.9949795294196624,
+      "learning_rate": 3.191770072185638e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 56195570,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3180424457403956,
+      "grad_norm": 2.807456405073622,
+      "learning_rate": 3.1911444132399165e-06,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 56211860,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3181626886310347,
+      "grad_norm": 2.153447572373795,
+      "learning_rate": 3.190518573598534e-06,
+      "loss": 0.8698,
+      "num_input_tokens_seen": 56228185,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3182829315216738,
+      "grad_norm": 1.5792194095603191,
+      "learning_rate": 3.1898925533564308e-06,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 56249375,
+      "step": 2647
+    },
+    {
+      "epoch": 0.31840317441231286,
+      "grad_norm": 8.325502809464872,
+      "learning_rate": 3.1892663526085733e-06,
+      "loss": 0.6427,
+      "num_input_tokens_seen": 56267470,
+      "step": 2648
+    },
+    {
+      "epoch": 0.31852341730295197,
+      "grad_norm": 0.7764030664179364,
+      "learning_rate": 3.188639971449956e-06,
+      "loss": 0.5965,
+      "num_input_tokens_seen": 56333240,
+      "step": 2649
+    },
+    {
+      "epoch": 0.318643660193591,
+      "grad_norm": 2.1659826537997584,
+      "learning_rate": 3.1880134099755995e-06,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 56352595,
+      "step": 2650
+    },
+    {
+      "epoch": 0.31876390308423014,
+      "grad_norm": 2.0736932254822813,
+      "learning_rate": 3.1873866682805535e-06,
+      "loss": 0.6956,
+      "num_input_tokens_seen": 56373010,
+      "step": 2651
+    },
+    {
+      "epoch": 0.31888414597486925,
+      "grad_norm": 1.8221722209252493,
+      "learning_rate": 3.186759746459894e-06,
+      "loss": 0.8844,
+      "num_input_tokens_seen": 56391840,
+      "step": 2652
+    },
+    {
+      "epoch": 0.3190043888655083,
+      "grad_norm": 1.8086018648329212,
+      "learning_rate": 3.186132644608725e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 56410300,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3191246317561474,
+      "grad_norm": 6.54584810830888,
+      "learning_rate": 3.1855053628221763e-06,
+      "loss": 0.712,
+      "num_input_tokens_seen": 56429275,
+      "step": 2654
+    },
+    {
+      "epoch": 0.3192448746467865,
+      "grad_norm": 2.3628032075870515,
+      "learning_rate": 3.184877901195407e-06,
+      "loss": 0.8966,
+      "num_input_tokens_seen": 56445690,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3193651175374256,
+      "grad_norm": 0.8533714042087319,
+      "learning_rate": 3.184250259823602e-06,
+      "loss": 0.6629,
+      "num_input_tokens_seen": 56507940,
+      "step": 2656
+    },
+    {
+      "epoch": 0.3194853604280647,
+      "grad_norm": 2.403070835149598,
+      "learning_rate": 3.1836224388019744e-06,
+      "loss": 0.8079,
+      "num_input_tokens_seen": 56522950,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3196056033187038,
+      "grad_norm": 2.069428387188427,
+      "learning_rate": 3.1829944382257633e-06,
+      "loss": 0.7631,
+      "num_input_tokens_seen": 56540800,
+      "step": 2658
+    },
+    {
+      "epoch": 0.31972584620934286,
+      "grad_norm": 2.30651719670016,
+      "learning_rate": 3.1823662581902373e-06,
+      "loss": 0.823,
+      "num_input_tokens_seen": 56558205,
+      "step": 2659
+    },
+    {
+      "epoch": 0.31984608909998197,
+      "grad_norm": 2.6735213260650372,
+      "learning_rate": 3.1817378987906896e-06,
+      "loss": 0.7462,
+      "num_input_tokens_seen": 56577430,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3199663319906211,
+      "grad_norm": 4.34219838385865,
+      "learning_rate": 3.181109360122442e-06,
+      "loss": 0.7931,
+      "num_input_tokens_seen": 56594740,
+      "step": 2661
+    },
+    {
+      "epoch": 0.32008657488126013,
+      "grad_norm": 2.2130761536745718,
+      "learning_rate": 3.180480642280844e-06,
+      "loss": 0.7815,
+      "num_input_tokens_seen": 56611595,
+      "step": 2662
+    },
+    {
+      "epoch": 0.32020681777189924,
+      "grad_norm": 1.567823631727516,
+      "learning_rate": 3.1798517453612714e-06,
+      "loss": 0.7243,
+      "num_input_tokens_seen": 56631120,
+      "step": 2663
+    },
+    {
+      "epoch": 0.32032706066253835,
+      "grad_norm": 1.9712893537711884,
+      "learning_rate": 3.1792226694591265e-06,
+      "loss": 0.749,
+      "num_input_tokens_seen": 56652225,
+      "step": 2664
+    },
+    {
+      "epoch": 0.3204473035531774,
+      "grad_norm": 1.7701556814905914,
+      "learning_rate": 3.178593414669841e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 56670530,
+      "step": 2665
+    },
+    {
+      "epoch": 0.3205675464438165,
+      "grad_norm": 2.2098971223214807,
+      "learning_rate": 3.1779639810888707e-06,
+      "loss": 0.702,
+      "num_input_tokens_seen": 56689845,
+      "step": 2666
+    },
+    {
+      "epoch": 0.3206877893344556,
+      "grad_norm": 1.7246063743039204,
+      "learning_rate": 3.1773343688117017e-06,
+      "loss": 0.7581,
+      "num_input_tokens_seen": 56710475,
+      "step": 2667
+    },
+    {
+      "epoch": 0.3208080322250947,
+      "grad_norm": 2.178756099863218,
+      "learning_rate": 3.1767045779338445e-06,
+      "loss": 0.8376,
+      "num_input_tokens_seen": 56727855,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3209282751157338,
+      "grad_norm": 2.3564056461653387,
+      "learning_rate": 3.176074608550839e-06,
+      "loss": 0.9053,
+      "num_input_tokens_seen": 56743395,
+      "step": 2669
+    },
+    {
+      "epoch": 0.32104851800637285,
+      "grad_norm": 2.470435478527772,
+      "learning_rate": 3.17544446075825e-06,
+      "loss": 0.8199,
+      "num_input_tokens_seen": 56762280,
+      "step": 2670
+    },
+    {
+      "epoch": 0.32116876089701196,
+      "grad_norm": 1.5967077732301016,
+      "learning_rate": 3.174814134651671e-06,
+      "loss": 0.7087,
+      "num_input_tokens_seen": 56784550,
+      "step": 2671
+    },
+    {
+      "epoch": 0.3212890037876511,
+      "grad_norm": 1.6741571286119414,
+      "learning_rate": 3.1741836303267215e-06,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 56803805,
+      "step": 2672
+    },
+    {
+      "epoch": 0.32140924667829013,
+      "grad_norm": 1.7881467202150503,
+      "learning_rate": 3.1735529478790496e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 56821515,
+      "step": 2673
+    },
+    {
+      "epoch": 0.32152948956892924,
+      "grad_norm": 2.0331375900292885,
+      "learning_rate": 3.1729220874043277e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 56843495,
+      "step": 2674
+    },
+    {
+      "epoch": 0.32164973245956835,
+      "grad_norm": 0.7922160094215054,
+      "learning_rate": 3.172291048998259e-06,
+      "loss": 0.5831,
+      "num_input_tokens_seen": 56903575,
+      "step": 2675
+    },
+    {
+      "epoch": 0.3217699753502074,
+      "grad_norm": 2.0327577553204743,
+      "learning_rate": 3.1716598327565694e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 56922935,
+      "step": 2676
+    },
+    {
+      "epoch": 0.3218902182408465,
+      "grad_norm": 1.4745353666774943,
+      "learning_rate": 3.171028438775015e-06,
+      "loss": 0.8375,
+      "num_input_tokens_seen": 56941850,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3220104611314856,
+      "grad_norm": 6.3525311680619705,
+      "learning_rate": 3.170396867149377e-06,
+      "loss": 0.8444,
+      "num_input_tokens_seen": 56959575,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3221307040221247,
+      "grad_norm": 2.071384362681274,
+      "learning_rate": 3.1697651179754657e-06,
+      "loss": 0.8522,
+      "num_input_tokens_seen": 56977955,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3222509469127638,
+      "grad_norm": 1.8745238138653095,
+      "learning_rate": 3.169133191349115e-06,
+      "loss": 0.7322,
+      "num_input_tokens_seen": 57000245,
+      "step": 2680
+    },
+    {
+      "epoch": 0.32237118980340285,
+      "grad_norm": 2.139584523458616,
+      "learning_rate": 3.1685010873661898e-06,
+      "loss": 0.8389,
+      "num_input_tokens_seen": 57019140,
+      "step": 2681
+    },
+    {
+      "epoch": 0.32249143269404196,
+      "grad_norm": 2.426915584321933,
+      "learning_rate": 3.167868806122578e-06,
+      "loss": 0.7983,
+      "num_input_tokens_seen": 57037910,
+      "step": 2682
+    },
+    {
+      "epoch": 0.32261167558468107,
+      "grad_norm": 1.8521185395363011,
+      "learning_rate": 3.1672363477141968e-06,
+      "loss": 0.6615,
+      "num_input_tokens_seen": 57056925,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3227319184753201,
+      "grad_norm": 2.2596497437861354,
+      "learning_rate": 3.16660371223699e-06,
+      "loss": 0.8503,
+      "num_input_tokens_seen": 57077305,
+      "step": 2684
+    },
+    {
+      "epoch": 0.32285216136595923,
+      "grad_norm": 2.0402950920581726,
+      "learning_rate": 3.1659708997869278e-06,
+      "loss": 0.8573,
+      "num_input_tokens_seen": 57094940,
+      "step": 2685
+    },
+    {
+      "epoch": 0.32297240425659834,
+      "grad_norm": 1.652098663387129,
+      "learning_rate": 3.1653379104600067e-06,
+      "loss": 0.7417,
+      "num_input_tokens_seen": 57114805,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3230926471472374,
+      "grad_norm": 1.9652968726869164,
+      "learning_rate": 3.1647047443522516e-06,
+      "loss": 0.6963,
+      "num_input_tokens_seen": 57135330,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3232128900378765,
+      "grad_norm": 1.639736119044479,
+      "learning_rate": 3.164071401559713e-06,
+      "loss": 0.806,
+      "num_input_tokens_seen": 57152450,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3233331329285156,
+      "grad_norm": 1.819350762997877,
+      "learning_rate": 3.1634378821784678e-06,
+      "loss": 0.7048,
+      "num_input_tokens_seen": 57172385,
+      "step": 2689
+    },
+    {
+      "epoch": 0.3234533758191547,
+      "grad_norm": 6.549150051219702,
+      "learning_rate": 3.1628041863046208e-06,
+      "loss": 0.7366,
+      "num_input_tokens_seen": 57189520,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3235736187097938,
+      "grad_norm": 2.0620647112057653,
+      "learning_rate": 3.162170314034304e-06,
+      "loss": 0.908,
+      "num_input_tokens_seen": 57206655,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3236938616004329,
+      "grad_norm": 1.6422682829267043,
+      "learning_rate": 3.1615362654636738e-06,
+      "loss": 0.7972,
+      "num_input_tokens_seen": 57227115,
+      "step": 2692
+    },
+    {
+      "epoch": 0.32381410449107195,
+      "grad_norm": 2.9180452439611364,
+      "learning_rate": 3.1609020406889163e-06,
+      "loss": 0.8648,
+      "num_input_tokens_seen": 57244270,
+      "step": 2693
+    },
+    {
+      "epoch": 0.32393434738171106,
+      "grad_norm": 2.147408519500242,
+      "learning_rate": 3.1602676398062416e-06,
+      "loss": 0.8423,
+      "num_input_tokens_seen": 57262900,
+      "step": 2694
+    },
+    {
+      "epoch": 0.3240545902723502,
+      "grad_norm": 2.8768352107211665,
+      "learning_rate": 3.1596330629118886e-06,
+      "loss": 0.6135,
+      "num_input_tokens_seen": 57282590,
+      "step": 2695
+    },
+    {
+      "epoch": 0.32417483316298923,
+      "grad_norm": 2.1293200529911225,
+      "learning_rate": 3.158998310102122e-06,
+      "loss": 0.7291,
+      "num_input_tokens_seen": 57300940,
+      "step": 2696
+    },
+    {
+      "epoch": 0.32429507605362834,
+      "grad_norm": 2.025354213499799,
+      "learning_rate": 3.1583633814732337e-06,
+      "loss": 0.8374,
+      "num_input_tokens_seen": 57320180,
+      "step": 2697
+    },
+    {
+      "epoch": 0.3244153189442674,
+      "grad_norm": 2.6121707740462736,
+      "learning_rate": 3.157728277121541e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 57338075,
+      "step": 2698
+    },
+    {
+      "epoch": 0.3245355618349065,
+      "grad_norm": 2.6456435470681865,
+      "learning_rate": 3.1570929971433897e-06,
+      "loss": 0.7804,
+      "num_input_tokens_seen": 57353580,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3246558047255456,
+      "grad_norm": 1.9632892797007728,
+      "learning_rate": 3.1564575416351504e-06,
+      "loss": 0.8239,
+      "num_input_tokens_seen": 57372000,
+      "step": 2700
+    },
+    {
+      "epoch": 0.32477604761618467,
+      "grad_norm": 2.5675351813695864,
+      "learning_rate": 3.1558219106932215e-06,
+      "loss": 0.7435,
+      "num_input_tokens_seen": 57391135,
+      "step": 2701
+    },
+    {
+      "epoch": 0.3248962905068238,
+      "grad_norm": 1.6371075018688064,
+      "learning_rate": 3.155186104414027e-06,
+      "loss": 0.8532,
+      "num_input_tokens_seen": 57410490,
+      "step": 2702
+    },
+    {
+      "epoch": 0.3250165333974629,
+      "grad_norm": 1.7372905081356436,
+      "learning_rate": 3.15455012289402e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 57429855,
+      "step": 2703
+    },
+    {
+      "epoch": 0.32513677628810195,
+      "grad_norm": 1.7288619320609986,
+      "learning_rate": 3.153913966229677e-06,
+      "loss": 0.8342,
+      "num_input_tokens_seen": 57448695,
+      "step": 2704
+    },
+    {
+      "epoch": 0.32525701917874106,
+      "grad_norm": 0.8349752911309698,
+      "learning_rate": 3.1532776345175027e-06,
+      "loss": 0.5158,
+      "num_input_tokens_seen": 57513560,
+      "step": 2705
+    },
+    {
+      "epoch": 0.32537726206938017,
+      "grad_norm": 2.0010251145225255,
+      "learning_rate": 3.152641127854028e-06,
+      "loss": 0.7826,
+      "num_input_tokens_seen": 57531710,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3254975049600192,
+      "grad_norm": 2.3279306915581137,
+      "learning_rate": 3.1520044463358116e-06,
+      "loss": 0.8108,
+      "num_input_tokens_seen": 57548160,
+      "step": 2707
+    },
+    {
+      "epoch": 0.32561774785065833,
+      "grad_norm": 1.6226381097249127,
+      "learning_rate": 3.1513675900594354e-06,
+      "loss": 0.7943,
+      "num_input_tokens_seen": 57566305,
+      "step": 2708
+    },
+    {
+      "epoch": 0.32573799074129745,
+      "grad_norm": 2.298219233863723,
+      "learning_rate": 3.1507305591215117e-06,
+      "loss": 0.8636,
+      "num_input_tokens_seen": 57583935,
+      "step": 2709
+    },
+    {
+      "epoch": 0.3258582336319365,
+      "grad_norm": 0.727161206180856,
+      "learning_rate": 3.150093353618677e-06,
+      "loss": 0.5815,
+      "num_input_tokens_seen": 57648385,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3259784765225756,
+      "grad_norm": 3.163989019330951,
+      "learning_rate": 3.149455973647596e-06,
+      "loss": 0.8708,
+      "num_input_tokens_seen": 57666165,
+      "step": 2711
+    },
+    {
+      "epoch": 0.32609871941321467,
+      "grad_norm": 2.176791539098969,
+      "learning_rate": 3.1488184193049563e-06,
+      "loss": 0.7658,
+      "num_input_tokens_seen": 57685420,
+      "step": 2712
+    },
+    {
+      "epoch": 0.3262189623038538,
+      "grad_norm": 1.7633302677195908,
+      "learning_rate": 3.1481806906874767e-06,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 57706450,
+      "step": 2713
+    },
+    {
+      "epoch": 0.3263392051944929,
+      "grad_norm": 2.1733410738226047,
+      "learning_rate": 3.147542787891899e-06,
+      "loss": 0.8722,
+      "num_input_tokens_seen": 57725515,
+      "step": 2714
+    },
+    {
+      "epoch": 0.32645944808513194,
+      "grad_norm": 2.247417633752184,
+      "learning_rate": 3.1469047110149926e-06,
+      "loss": 0.7485,
+      "num_input_tokens_seen": 57743975,
+      "step": 2715
+    },
+    {
+      "epoch": 0.32657969097577105,
+      "grad_norm": 1.8267752316455788,
+      "learning_rate": 3.1462664601535537e-06,
+      "loss": 0.8442,
+      "num_input_tokens_seen": 57763405,
+      "step": 2716
+    },
+    {
+      "epoch": 0.32669993386641016,
+      "grad_norm": 1.8744550556417996,
+      "learning_rate": 3.145628035404404e-06,
+      "loss": 0.7952,
+      "num_input_tokens_seen": 57782325,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3268201767570492,
+      "grad_norm": 0.8716495105232366,
+      "learning_rate": 3.1449894368643922e-06,
+      "loss": 0.6015,
+      "num_input_tokens_seen": 57844360,
+      "step": 2718
+    },
+    {
+      "epoch": 0.32694041964768833,
+      "grad_norm": 1.5618477147636627,
+      "learning_rate": 3.1443506646303934e-06,
+      "loss": 0.7182,
+      "num_input_tokens_seen": 57865380,
+      "step": 2719
+    },
+    {
+      "epoch": 0.32706066253832744,
+      "grad_norm": 2.7422734548256553,
+      "learning_rate": 3.1437117187993086e-06,
+      "loss": 0.6703,
+      "num_input_tokens_seen": 57887420,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3271809054289665,
+      "grad_norm": 1.6234747529148854,
+      "learning_rate": 3.143072599468065e-06,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 57906965,
+      "step": 2721
+    },
+    {
+      "epoch": 0.3273011483196056,
+      "grad_norm": 1.6064520349385707,
+      "learning_rate": 3.1424333067336174e-06,
+      "loss": 0.7516,
+      "num_input_tokens_seen": 57929450,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3274213912102447,
+      "grad_norm": 1.8748318849504562,
+      "learning_rate": 3.141793840692945e-06,
+      "loss": 0.7659,
+      "num_input_tokens_seen": 57949920,
+      "step": 2723
+    },
+    {
+      "epoch": 0.32754163410088377,
+      "grad_norm": 2.1334997271929996,
+      "learning_rate": 3.1411542014430553e-06,
+      "loss": 0.6101,
+      "num_input_tokens_seen": 57970720,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3276618769915229,
+      "grad_norm": 1.7573760131612186,
+      "learning_rate": 3.1405143890809804e-06,
+      "loss": 0.8196,
+      "num_input_tokens_seen": 57989735,
+      "step": 2725
+    },
+    {
+      "epoch": 0.327782119882162,
+      "grad_norm": 4.459579605018711,
+      "learning_rate": 3.1398744037037796e-06,
+      "loss": 0.7013,
+      "num_input_tokens_seen": 58008790,
+      "step": 2726
+    },
+    {
+      "epoch": 0.32790236277280105,
+      "grad_norm": 1.9377888080148526,
+      "learning_rate": 3.139234245408538e-06,
+      "loss": 0.832,
+      "num_input_tokens_seen": 58027390,
+      "step": 2727
+    },
+    {
+      "epoch": 0.32802260566344016,
+      "grad_norm": 1.4230882857102058,
+      "learning_rate": 3.1385939142923666e-06,
+      "loss": 0.7622,
+      "num_input_tokens_seen": 58049500,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3281428485540792,
+      "grad_norm": 2.328767722205342,
+      "learning_rate": 3.137953410452405e-06,
+      "loss": 0.7796,
+      "num_input_tokens_seen": 58069490,
+      "step": 2729
+    },
+    {
+      "epoch": 0.3282630914447183,
+      "grad_norm": 2.0193660226844967,
+      "learning_rate": 3.137312733985814e-06,
+      "loss": 0.7459,
+      "num_input_tokens_seen": 58091810,
+      "step": 2730
+    },
+    {
+      "epoch": 0.32838333433535744,
+      "grad_norm": 1.909184868042113,
+      "learning_rate": 3.136671884989787e-06,
+      "loss": 0.7432,
+      "num_input_tokens_seen": 58111440,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3285035772259965,
+      "grad_norm": 3.250984534581054,
+      "learning_rate": 3.1360308635615383e-06,
+      "loss": 0.8685,
+      "num_input_tokens_seen": 58129700,
+      "step": 2732
+    },
+    {
+      "epoch": 0.3286238201166356,
+      "grad_norm": 1.8013503046713124,
+      "learning_rate": 3.135389669798311e-06,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 58147480,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3287440630072747,
+      "grad_norm": 1.8713527684415947,
+      "learning_rate": 3.134748303797373e-06,
+      "loss": 0.7975,
+      "num_input_tokens_seen": 58164570,
+      "step": 2734
+    },
+    {
+      "epoch": 0.32886430589791377,
+      "grad_norm": 2.6957095076688375,
+      "learning_rate": 3.1341067656560203e-06,
+      "loss": 0.804,
+      "num_input_tokens_seen": 58182135,
+      "step": 2735
+    },
+    {
+      "epoch": 0.3289845487885529,
+      "grad_norm": 2.741633372466873,
+      "learning_rate": 3.133465055471572e-06,
+      "loss": 0.8631,
+      "num_input_tokens_seen": 58201640,
+      "step": 2736
+    },
+    {
+      "epoch": 0.329104791679192,
+      "grad_norm": 3.47478827457579,
+      "learning_rate": 3.1328231733413767e-06,
+      "loss": 0.6489,
+      "num_input_tokens_seen": 58218000,
+      "step": 2737
+    },
+    {
+      "epoch": 0.32922503456983104,
+      "grad_norm": 2.4625185237992993,
+      "learning_rate": 3.1321811193628067e-06,
+      "loss": 0.9008,
+      "num_input_tokens_seen": 58235865,
+      "step": 2738
+    },
+    {
+      "epoch": 0.32934527746047015,
+      "grad_norm": 2.501407061504325,
+      "learning_rate": 3.131538893633261e-06,
+      "loss": 0.7063,
+      "num_input_tokens_seen": 58255145,
+      "step": 2739
+    },
+    {
+      "epoch": 0.32946552035110926,
+      "grad_norm": 2.1514585083941826,
+      "learning_rate": 3.1308964962501648e-06,
+      "loss": 0.7775,
+      "num_input_tokens_seen": 58274690,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3295857632417483,
+      "grad_norm": 2.5706237545260247,
+      "learning_rate": 3.1302539273109693e-06,
+      "loss": 0.8631,
+      "num_input_tokens_seen": 58291235,
+      "step": 2741
+    },
+    {
+      "epoch": 0.32970600613238743,
+      "grad_norm": 1.7458017510375092,
+      "learning_rate": 3.1296111869131513e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 58308380,
+      "step": 2742
+    },
+    {
+      "epoch": 0.32982624902302654,
+      "grad_norm": 2.756484539653937,
+      "learning_rate": 3.1289682751542153e-06,
+      "loss": 0.8444,
+      "num_input_tokens_seen": 58327660,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3299464919136656,
+      "grad_norm": 3.848639538076117,
+      "learning_rate": 3.1283251921316883e-06,
+      "loss": 0.7141,
+      "num_input_tokens_seen": 58345125,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3300667348043047,
+      "grad_norm": 2.6946446718393697,
+      "learning_rate": 3.127681937943128e-06,
+      "loss": 0.8133,
+      "num_input_tokens_seen": 58362935,
+      "step": 2745
+    },
+    {
+      "epoch": 0.33018697769494376,
+      "grad_norm": 2.949483909575551,
+      "learning_rate": 3.1270385126861134e-06,
+      "loss": 0.7524,
+      "num_input_tokens_seen": 58380640,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3303072205855829,
+      "grad_norm": 2.1452245056967523,
+      "learning_rate": 3.1263949164582533e-06,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 58400010,
+      "step": 2747
+    },
+    {
+      "epoch": 0.330427463476222,
+      "grad_norm": 2.174426347375206,
+      "learning_rate": 3.1257511493571797e-06,
+      "loss": 0.7787,
+      "num_input_tokens_seen": 58418235,
+      "step": 2748
+    },
+    {
+      "epoch": 0.33054770636686104,
+      "grad_norm": 2.5200604964587408,
+      "learning_rate": 3.125107211480552e-06,
+      "loss": 0.7806,
+      "num_input_tokens_seen": 58437890,
+      "step": 2749
+    },
+    {
+      "epoch": 0.33066794925750015,
+      "grad_norm": 1.856137017745213,
+      "learning_rate": 3.1244631029260536e-06,
+      "loss": 0.7902,
+      "num_input_tokens_seen": 58456945,
+      "step": 2750
+    },
+    {
+      "epoch": 0.33078819214813926,
+      "grad_norm": 0.7748134796518019,
+      "learning_rate": 3.1238188237913984e-06,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 58521205,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3309084350387783,
+      "grad_norm": 2.554187944704898,
+      "learning_rate": 3.12317437417432e-06,
+      "loss": 0.7616,
+      "num_input_tokens_seen": 58540430,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3310286779294174,
+      "grad_norm": 2.7429809662048465,
+      "learning_rate": 3.122529754172582e-06,
+      "loss": 0.8333,
+      "num_input_tokens_seen": 58557035,
+      "step": 2753
+    },
+    {
+      "epoch": 0.33114892082005654,
+      "grad_norm": 2.0071136049488625,
+      "learning_rate": 3.1218849638839736e-06,
+      "loss": 0.7247,
+      "num_input_tokens_seen": 58576015,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3312691637106956,
+      "grad_norm": 2.0085487201826986,
+      "learning_rate": 3.121240003406308e-06,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 58594585,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3313894066013347,
+      "grad_norm": 2.7553226068053096,
+      "learning_rate": 3.120594872837425e-06,
+      "loss": 0.7155,
+      "num_input_tokens_seen": 58612975,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3315096494919738,
+      "grad_norm": 0.8672724726438519,
+      "learning_rate": 3.1199495722751906e-06,
+      "loss": 0.6446,
+      "num_input_tokens_seen": 58672225,
+      "step": 2757
+    },
+    {
+      "epoch": 0.33162989238261287,
+      "grad_norm": 1.61481718679592,
+      "learning_rate": 3.1193041018174972e-06,
+      "loss": 0.8393,
+      "num_input_tokens_seen": 58692660,
+      "step": 2758
+    },
+    {
+      "epoch": 0.331750135273252,
+      "grad_norm": 2.149748696279514,
+      "learning_rate": 3.118658461562261e-06,
+      "loss": 0.9451,
+      "num_input_tokens_seen": 58708480,
+      "step": 2759
+    },
+    {
+      "epoch": 0.33187037816389103,
+      "grad_norm": 2.038008284711372,
+      "learning_rate": 3.1180126516074254e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 58729805,
+      "step": 2760
+    },
+    {
+      "epoch": 0.33199062105453014,
+      "grad_norm": 3.352699363379197,
+      "learning_rate": 3.1173666720509603e-06,
+      "loss": 0.83,
+      "num_input_tokens_seen": 58746460,
+      "step": 2761
+    },
+    {
+      "epoch": 0.33211086394516925,
+      "grad_norm": 1.884559777331994,
+      "learning_rate": 3.1167205229908586e-06,
+      "loss": 0.6822,
+      "num_input_tokens_seen": 58767055,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3322311068358083,
+      "grad_norm": 2.572815908871451,
+      "learning_rate": 3.116074204525142e-06,
+      "loss": 0.6225,
+      "num_input_tokens_seen": 58784950,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3323513497264474,
+      "grad_norm": 1.6029764448841326,
+      "learning_rate": 3.1154277167518553e-06,
+      "loss": 0.826,
+      "num_input_tokens_seen": 58806285,
+      "step": 2764
+    },
+    {
+      "epoch": 0.33247159261708653,
+      "grad_norm": 0.8608217438447363,
+      "learning_rate": 3.114781059769072e-06,
+      "loss": 0.6219,
+      "num_input_tokens_seen": 58857330,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3325918355077256,
+      "grad_norm": 3.2520588667326584,
+      "learning_rate": 3.1141342336748874e-06,
+      "loss": 0.6841,
+      "num_input_tokens_seen": 58876610,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3327120783983647,
+      "grad_norm": 1.5406510525943908,
+      "learning_rate": 3.1134872385674253e-06,
+      "loss": 0.812,
+      "num_input_tokens_seen": 58900485,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3328323212890038,
+      "grad_norm": 2.4035155038083045,
+      "learning_rate": 3.112840074544835e-06,
+      "loss": 0.8587,
+      "num_input_tokens_seen": 58919585,
+      "step": 2768
+    },
+    {
+      "epoch": 0.33295256417964286,
+      "grad_norm": 2.1754588966571236,
+      "learning_rate": 3.11219274170529e-06,
+      "loss": 0.6265,
+      "num_input_tokens_seen": 58941115,
+      "step": 2769
+    },
+    {
+      "epoch": 0.333072807070282,
+      "grad_norm": 2.167767765999262,
+      "learning_rate": 3.1115452401469903e-06,
+      "loss": 0.8087,
+      "num_input_tokens_seen": 58961235,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3331930499609211,
+      "grad_norm": 1.931759854733021,
+      "learning_rate": 3.1108975699681613e-06,
+      "loss": 0.858,
+      "num_input_tokens_seen": 58978350,
+      "step": 2771
+    },
+    {
+      "epoch": 0.33331329285156014,
+      "grad_norm": 2.019200170330616,
+      "learning_rate": 3.1102497312670542e-06,
+      "loss": 0.7111,
+      "num_input_tokens_seen": 58996075,
+      "step": 2772
+    },
+    {
+      "epoch": 0.33343353574219925,
+      "grad_norm": 4.258261403095521,
+      "learning_rate": 3.109601724141946e-06,
+      "loss": 0.8,
+      "num_input_tokens_seen": 59014790,
+      "step": 2773
+    },
+    {
+      "epoch": 0.33355377863283836,
+      "grad_norm": 1.6806437356975537,
+      "learning_rate": 3.108953548691138e-06,
+      "loss": 0.681,
+      "num_input_tokens_seen": 59034595,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3336740215234774,
+      "grad_norm": 2.8488263989856635,
+      "learning_rate": 3.108305205012959e-06,
+      "loss": 0.7168,
+      "num_input_tokens_seen": 59055010,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3337942644141165,
+      "grad_norm": 2.1918573878030116,
+      "learning_rate": 3.107656693205761e-06,
+      "loss": 0.8728,
+      "num_input_tokens_seen": 59074170,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3339145073047556,
+      "grad_norm": 2.5064266077010506,
+      "learning_rate": 3.107008013367924e-06,
+      "loss": 0.6937,
+      "num_input_tokens_seen": 59092685,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3340347501953947,
+      "grad_norm": 3.7241341908238352,
+      "learning_rate": 3.1063591655978507e-06,
+      "loss": 0.8622,
+      "num_input_tokens_seen": 59108355,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3341549930860338,
+      "grad_norm": 1.8034720128038548,
+      "learning_rate": 3.105710149993972e-06,
+      "loss": 0.7858,
+      "num_input_tokens_seen": 59127405,
+      "step": 2779
+    },
+    {
+      "epoch": 0.33427523597667286,
+      "grad_norm": 1.7007959847279115,
+      "learning_rate": 3.1050609666547427e-06,
+      "loss": 0.8508,
+      "num_input_tokens_seen": 59146685,
+      "step": 2780
+    },
+    {
+      "epoch": 0.33439547886731197,
+      "grad_norm": 3.635785976873577,
+      "learning_rate": 3.104411615678644e-06,
+      "loss": 0.7639,
+      "num_input_tokens_seen": 59165255,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3345157217579511,
+      "grad_norm": 3.132579974044228,
+      "learning_rate": 3.1037620971641803e-06,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 59184765,
+      "step": 2782
+    },
+    {
+      "epoch": 0.33463596464859013,
+      "grad_norm": 3.3596132386151862,
+      "learning_rate": 3.1031124112098844e-06,
+      "loss": 0.6493,
+      "num_input_tokens_seen": 59202695,
+      "step": 2783
+    },
+    {
+      "epoch": 0.33475620753922924,
+      "grad_norm": 2.413872215076397,
+      "learning_rate": 3.1024625579143127e-06,
+      "loss": 0.7169,
+      "num_input_tokens_seen": 59219935,
+      "step": 2784
+    },
+    {
+      "epoch": 0.33487645042986836,
+      "grad_norm": 1.734038045042645,
+      "learning_rate": 3.101812537376048e-06,
+      "loss": 0.7197,
+      "num_input_tokens_seen": 59238675,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3349966933205074,
+      "grad_norm": 2.0553162598388326,
+      "learning_rate": 3.101162349693697e-06,
+      "loss": 0.8379,
+      "num_input_tokens_seen": 59256690,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3351169362111465,
+      "grad_norm": 1.729847794193637,
+      "learning_rate": 3.100511994965893e-06,
+      "loss": 0.695,
+      "num_input_tokens_seen": 59276365,
+      "step": 2787
+    },
+    {
+      "epoch": 0.33523717910178563,
+      "grad_norm": 1.9528519302845946,
+      "learning_rate": 3.0998614732912947e-06,
+      "loss": 0.8521,
+      "num_input_tokens_seen": 59295460,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3353574219924247,
+      "grad_norm": 1.8970150449988303,
+      "learning_rate": 3.0992107847685855e-06,
+      "loss": 0.673,
+      "num_input_tokens_seen": 59312895,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3354776648830638,
+      "grad_norm": 1.6691314417713952,
+      "learning_rate": 3.0985599294964736e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 59332170,
+      "step": 2790
+    },
+    {
+      "epoch": 0.33559790777370285,
+      "grad_norm": 4.688649555518731,
+      "learning_rate": 3.097908907573695e-06,
+      "loss": 0.7003,
+      "num_input_tokens_seen": 59349870,
+      "step": 2791
+    },
+    {
+      "epoch": 0.33571815066434196,
+      "grad_norm": 2.1372373782661604,
+      "learning_rate": 3.0972577190990067e-06,
+      "loss": 0.8875,
+      "num_input_tokens_seen": 59368070,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3358383935549811,
+      "grad_norm": 2.3743460348214316,
+      "learning_rate": 3.096606364171196e-06,
+      "loss": 0.7991,
+      "num_input_tokens_seen": 59387580,
+      "step": 2793
+    },
+    {
+      "epoch": 0.33595863644562013,
+      "grad_norm": 1.9269947028243861,
+      "learning_rate": 3.0959548428890703e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 59406170,
+      "step": 2794
+    },
+    {
+      "epoch": 0.33607887933625924,
+      "grad_norm": 1.8237253708863697,
+      "learning_rate": 3.095303155351468e-06,
+      "loss": 0.8312,
+      "num_input_tokens_seen": 59426095,
+      "step": 2795
+    },
+    {
+      "epoch": 0.33619912222689835,
+      "grad_norm": 3.2811700261462478,
+      "learning_rate": 3.0946513016572464e-06,
+      "loss": 0.7871,
+      "num_input_tokens_seen": 59444720,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3363193651175374,
+      "grad_norm": 3.8493121665902885,
+      "learning_rate": 3.0939992819052938e-06,
+      "loss": 0.7625,
+      "num_input_tokens_seen": 59461950,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3364396080081765,
+      "grad_norm": 2.335749678065399,
+      "learning_rate": 3.0933470961945193e-06,
+      "loss": 0.811,
+      "num_input_tokens_seen": 59479965,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3365598508988156,
+      "grad_norm": 2.30020559902466,
+      "learning_rate": 3.09269474462386e-06,
+      "loss": 0.6896,
+      "num_input_tokens_seen": 59499255,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3366800937894547,
+      "grad_norm": 2.438991638638325,
+      "learning_rate": 3.092042227292276e-06,
+      "loss": 0.8197,
+      "num_input_tokens_seen": 59515810,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3368003366800938,
+      "grad_norm": 2.302191166023087,
+      "learning_rate": 3.0913895442987557e-06,
+      "loss": 0.8764,
+      "num_input_tokens_seen": 59536495,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3369205795707329,
+      "grad_norm": 1.6902246842435145,
+      "learning_rate": 3.090736695742308e-06,
+      "loss": 0.8435,
+      "num_input_tokens_seen": 59557345,
+      "step": 2802
+    },
+    {
+      "epoch": 0.33704082246137196,
+      "grad_norm": 2.709622204578694,
+      "learning_rate": 3.0900836817219713e-06,
+      "loss": 0.5072,
+      "num_input_tokens_seen": 59573495,
+      "step": 2803
+    },
+    {
+      "epoch": 0.33716106535201107,
+      "grad_norm": 1.949156738012,
+      "learning_rate": 3.089430502336807e-06,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 59593185,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3372813082426502,
+      "grad_norm": 2.8325630355072677,
+      "learning_rate": 3.088777157685902e-06,
+      "loss": 0.8919,
+      "num_input_tokens_seen": 59608495,
+      "step": 2805
+    },
+    {
+      "epoch": 0.33740155113328923,
+      "grad_norm": 2.026935667165047,
+      "learning_rate": 3.088123647868367e-06,
+      "loss": 0.8535,
+      "num_input_tokens_seen": 59624765,
+      "step": 2806
+    },
+    {
+      "epoch": 0.33752179402392835,
+      "grad_norm": 2.3353906795228587,
+      "learning_rate": 3.0874699729833405e-06,
+      "loss": 0.81,
+      "num_input_tokens_seen": 59645855,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3376420369145674,
+      "grad_norm": 1.6741238948668111,
+      "learning_rate": 3.086816133129983e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 59665835,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3377622798052065,
+      "grad_norm": 1.9207040290975084,
+      "learning_rate": 3.0861621284074826e-06,
+      "loss": 0.7652,
+      "num_input_tokens_seen": 59686080,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3378825226958456,
+      "grad_norm": 1.5637863864427821,
+      "learning_rate": 3.085507958915051e-06,
+      "loss": 0.7299,
+      "num_input_tokens_seen": 59704230,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3380027655864847,
+      "grad_norm": 1.798883797434924,
+      "learning_rate": 3.0848536247519253e-06,
+      "loss": 0.707,
+      "num_input_tokens_seen": 59725535,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3381230084771238,
+      "grad_norm": 3.018216391490923,
+      "learning_rate": 3.0841991260173663e-06,
+      "loss": 0.8514,
+      "num_input_tokens_seen": 59745160,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3382432513677629,
+      "grad_norm": 1.7386044483791776,
+      "learning_rate": 3.0835444628106634e-06,
+      "loss": 0.7916,
+      "num_input_tokens_seen": 59763860,
+      "step": 2813
+    },
+    {
+      "epoch": 0.33836349425840195,
+      "grad_norm": 1.9071974860174172,
+      "learning_rate": 3.082889635231126e-06,
+      "loss": 0.8262,
+      "num_input_tokens_seen": 59782240,
+      "step": 2814
+    },
+    {
+      "epoch": 0.33848373714904106,
+      "grad_norm": 4.490045600224103,
+      "learning_rate": 3.0822346433780925e-06,
+      "loss": 0.7647,
+      "num_input_tokens_seen": 59802685,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3386039800396802,
+      "grad_norm": 2.495872181672995,
+      "learning_rate": 3.0815794873509237e-06,
+      "loss": 0.8673,
+      "num_input_tokens_seen": 59820690,
+      "step": 2816
+    },
+    {
+      "epoch": 0.33872422293031923,
+      "grad_norm": 1.970151639515993,
+      "learning_rate": 3.0809241672490066e-06,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 59838580,
+      "step": 2817
+    },
+    {
+      "epoch": 0.33884446582095834,
+      "grad_norm": 1.5991942417581513,
+      "learning_rate": 3.080268683171753e-06,
+      "loss": 0.8408,
+      "num_input_tokens_seen": 59858590,
+      "step": 2818
+    },
+    {
+      "epoch": 0.33896470871159745,
+      "grad_norm": 2.7415043629581843,
+      "learning_rate": 3.0796130352185985e-06,
+      "loss": 0.8872,
+      "num_input_tokens_seen": 59875165,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3390849516022365,
+      "grad_norm": 1.8129485181822191,
+      "learning_rate": 3.0789572234890057e-06,
+      "loss": 0.6669,
+      "num_input_tokens_seen": 59896525,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3392051944928756,
+      "grad_norm": 2.4069161910217174,
+      "learning_rate": 3.0783012480824596e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 59915390,
+      "step": 2821
+    },
+    {
+      "epoch": 0.33932543738351467,
+      "grad_norm": 5.3240901826034195,
+      "learning_rate": 3.077645109098471e-06,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 59931380,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3394456802741538,
+      "grad_norm": 1.722209468626472,
+      "learning_rate": 3.076988806636577e-06,
+      "loss": 0.709,
+      "num_input_tokens_seen": 59948860,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3395659231647929,
+      "grad_norm": 2.113534641354055,
+      "learning_rate": 3.076332340796337e-06,
+      "loss": 0.878,
+      "num_input_tokens_seen": 59968190,
+      "step": 2824
+    },
+    {
+      "epoch": 0.33968616605543195,
+      "grad_norm": 1.7360650334619434,
+      "learning_rate": 3.075675711677337e-06,
+      "loss": 0.7998,
+      "num_input_tokens_seen": 59988005,
+      "step": 2825
+    },
+    {
+      "epoch": 0.33980640894607106,
+      "grad_norm": 2.161625221142269,
+      "learning_rate": 3.0750189193791865e-06,
+      "loss": 0.7698,
+      "num_input_tokens_seen": 60007310,
+      "step": 2826
+    },
+    {
+      "epoch": 0.33992665183671017,
+      "grad_norm": 3.5400258198806966,
+      "learning_rate": 3.0743619640015207e-06,
+      "loss": 0.7054,
+      "num_input_tokens_seen": 60027280,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3400468947273492,
+      "grad_norm": 1.8901926877729285,
+      "learning_rate": 3.073704845643999e-06,
+      "loss": 0.9171,
+      "num_input_tokens_seen": 60044125,
+      "step": 2828
+    },
+    {
+      "epoch": 0.34016713761798834,
+      "grad_norm": 3.352970335615011,
+      "learning_rate": 3.0730475644063063e-06,
+      "loss": 0.7793,
+      "num_input_tokens_seen": 60058945,
+      "step": 2829
+    },
+    {
+      "epoch": 0.34028738050862745,
+      "grad_norm": 1.6493481265206238,
+      "learning_rate": 3.072390120388151e-06,
+      "loss": 0.6534,
+      "num_input_tokens_seen": 60076990,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3404076233992665,
+      "grad_norm": 2.474475380315649,
+      "learning_rate": 3.071732513689267e-06,
+      "loss": 0.709,
+      "num_input_tokens_seen": 60095245,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3405278662899056,
+      "grad_norm": 6.846865583761615,
+      "learning_rate": 3.0710747444094125e-06,
+      "loss": 0.6727,
+      "num_input_tokens_seen": 60112995,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3406481091805447,
+      "grad_norm": 2.1503353467810804,
+      "learning_rate": 3.070416812648372e-06,
+      "loss": 0.6458,
+      "num_input_tokens_seen": 60136165,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3407683520711838,
+      "grad_norm": 2.031735930049402,
+      "learning_rate": 3.069758718505951e-06,
+      "loss": 0.6598,
+      "num_input_tokens_seen": 60157625,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3408885949618229,
+      "grad_norm": 1.7753580427927367,
+      "learning_rate": 3.0691004620819836e-06,
+      "loss": 0.804,
+      "num_input_tokens_seen": 60177475,
+      "step": 2835
+    },
+    {
+      "epoch": 0.341008837852462,
+      "grad_norm": 0.8385604745430449,
+      "learning_rate": 3.0684420434763254e-06,
+      "loss": 0.6361,
+      "num_input_tokens_seen": 60243380,
+      "step": 2836
+    },
+    {
+      "epoch": 0.34112908074310105,
+      "grad_norm": 1.8362738978232802,
+      "learning_rate": 3.06778346278886e-06,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 60261935,
+      "step": 2837
+    },
+    {
+      "epoch": 0.34124932363374016,
+      "grad_norm": 2.2381543922175866,
+      "learning_rate": 3.0671247201194906e-06,
+      "loss": 0.7804,
+      "num_input_tokens_seen": 60283790,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3413695665243792,
+      "grad_norm": 1.9379672220776607,
+      "learning_rate": 3.066465815568151e-06,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 60304340,
+      "step": 2839
+    },
+    {
+      "epoch": 0.34148980941501833,
+      "grad_norm": 1.7784290958254294,
+      "learning_rate": 3.0658067492347947e-06,
+      "loss": 0.6779,
+      "num_input_tokens_seen": 60326700,
+      "step": 2840
+    },
+    {
+      "epoch": 0.34161005230565744,
+      "grad_norm": 2.1823654584586563,
+      "learning_rate": 3.0651475212194023e-06,
+      "loss": 0.6668,
+      "num_input_tokens_seen": 60345675,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3417302951962965,
+      "grad_norm": 1.5510014926499225,
+      "learning_rate": 3.064488131621977e-06,
+      "loss": 0.7465,
+      "num_input_tokens_seen": 60368720,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3418505380869356,
+      "grad_norm": 1.779202623474596,
+      "learning_rate": 3.063828580542549e-06,
+      "loss": 0.7341,
+      "num_input_tokens_seen": 60389635,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3419707809775747,
+      "grad_norm": 1.7754439046263575,
+      "learning_rate": 3.0631688680811706e-06,
+      "loss": 0.7209,
+      "num_input_tokens_seen": 60408980,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3420910238682138,
+      "grad_norm": 1.9932715738635667,
+      "learning_rate": 3.062508994337921e-06,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 60428305,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3422112667588529,
+      "grad_norm": 2.874751239830218,
+      "learning_rate": 3.0618489594129013e-06,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 60446165,
+      "step": 2846
+    },
+    {
+      "epoch": 0.342331509649492,
+      "grad_norm": 2.0439749162215923,
+      "learning_rate": 3.061188763406239e-06,
+      "loss": 0.705,
+      "num_input_tokens_seen": 60462030,
+      "step": 2847
+    },
+    {
+      "epoch": 0.34245175254013105,
+      "grad_norm": 2.5900072155005884,
+      "learning_rate": 3.060528406418085e-06,
+      "loss": 0.8229,
+      "num_input_tokens_seen": 60481600,
+      "step": 2848
+    },
+    {
+      "epoch": 0.34257199543077016,
+      "grad_norm": 1.7684958294198165,
+      "learning_rate": 3.0598678885486145e-06,
+      "loss": 0.6206,
+      "num_input_tokens_seen": 60503860,
+      "step": 2849
+    },
+    {
+      "epoch": 0.34269223832140927,
+      "grad_norm": 1.9303883221766656,
+      "learning_rate": 3.0592072098980282e-06,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 60523240,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3428124812120483,
+      "grad_norm": 2.129521971104975,
+      "learning_rate": 3.0585463705665514e-06,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 60543335,
+      "step": 2851
+    },
+    {
+      "epoch": 0.34293272410268744,
+      "grad_norm": 17.322176662358288,
+      "learning_rate": 3.0578853706544304e-06,
+      "loss": 0.7083,
+      "num_input_tokens_seen": 60560445,
+      "step": 2852
+    },
+    {
+      "epoch": 0.34305296699332655,
+      "grad_norm": 2.0827704907175724,
+      "learning_rate": 3.0572242102619404e-06,
+      "loss": 0.6452,
+      "num_input_tokens_seen": 60577320,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3431732098839656,
+      "grad_norm": 1.8626898159940488,
+      "learning_rate": 3.0565628894893776e-06,
+      "loss": 0.7999,
+      "num_input_tokens_seen": 60597675,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3432934527746047,
+      "grad_norm": 2.1442553149012427,
+      "learning_rate": 3.055901408437066e-06,
+      "loss": 0.7426,
+      "num_input_tokens_seen": 60615920,
+      "step": 2855
+    },
+    {
+      "epoch": 0.34341369566524377,
+      "grad_norm": 1.7409326367148166,
+      "learning_rate": 3.055239767205349e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 60637390,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3435339385558829,
+      "grad_norm": 1.9046985936687577,
+      "learning_rate": 3.054577965894599e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 60653255,
+      "step": 2857
+    },
+    {
+      "epoch": 0.343654181446522,
+      "grad_norm": 1.5901725571863656,
+      "learning_rate": 3.0539160046052094e-06,
+      "loss": 0.6985,
+      "num_input_tokens_seen": 60672675,
+      "step": 2858
+    },
+    {
+      "epoch": 0.34377442433716104,
+      "grad_norm": 2.408357923079108,
+      "learning_rate": 3.0532538834376003e-06,
+      "loss": 0.6956,
+      "num_input_tokens_seen": 60691955,
+      "step": 2859
+    },
+    {
+      "epoch": 0.34389466722780015,
+      "grad_norm": 2.143233525566655,
+      "learning_rate": 3.0525916024922143e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 60710860,
+      "step": 2860
+    },
+    {
+      "epoch": 0.34401491011843927,
+      "grad_norm": 2.621784356142421,
+      "learning_rate": 3.0519291618695193e-06,
+      "loss": 0.8386,
+      "num_input_tokens_seen": 60727980,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3441351530090783,
+      "grad_norm": 1.822456067354058,
+      "learning_rate": 3.051266561670007e-06,
+      "loss": 0.757,
+      "num_input_tokens_seen": 60746765,
+      "step": 2862
+    },
+    {
+      "epoch": 0.34425539589971743,
+      "grad_norm": 1.8990869297627924,
+      "learning_rate": 3.0506038019941933e-06,
+      "loss": 0.8942,
+      "num_input_tokens_seen": 60766495,
+      "step": 2863
+    },
+    {
+      "epoch": 0.34437563879035654,
+      "grad_norm": 2.2645115204256214,
+      "learning_rate": 3.049940882942617e-06,
+      "loss": 0.676,
+      "num_input_tokens_seen": 60785000,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3444958816809956,
+      "grad_norm": 1.9680746893813048,
+      "learning_rate": 3.0492778046158448e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 60806140,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3446161245716347,
+      "grad_norm": 3.3283965873512136,
+      "learning_rate": 3.0486145671144633e-06,
+      "loss": 0.7623,
+      "num_input_tokens_seen": 60825650,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3447363674622738,
+      "grad_norm": 1.9987305200009258,
+      "learning_rate": 3.047951170539086e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 60844995,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3448566103529129,
+      "grad_norm": 1.844060554489792,
+      "learning_rate": 3.047287614990349e-06,
+      "loss": 0.8364,
+      "num_input_tokens_seen": 60862635,
+      "step": 2868
+    },
+    {
+      "epoch": 0.344976853243552,
+      "grad_norm": 2.2583330194911873,
+      "learning_rate": 3.046623900568914e-06,
+      "loss": 0.6115,
+      "num_input_tokens_seen": 60884920,
+      "step": 2869
+    },
+    {
+      "epoch": 0.34509709613419104,
+      "grad_norm": 2.5237396072914366,
+      "learning_rate": 3.045960027375465e-06,
+      "loss": 0.6877,
+      "num_input_tokens_seen": 60902475,
+      "step": 2870
+    },
+    {
+      "epoch": 0.34521733902483015,
+      "grad_norm": 12.745458521410642,
+      "learning_rate": 3.045295995510712e-06,
+      "loss": 0.815,
+      "num_input_tokens_seen": 60919165,
+      "step": 2871
+    },
+    {
+      "epoch": 0.34533758191546926,
+      "grad_norm": 1.7615115863204773,
+      "learning_rate": 3.0446318050753865e-06,
+      "loss": 0.7282,
+      "num_input_tokens_seen": 60939365,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3454578248061083,
+      "grad_norm": 2.0581348142581812,
+      "learning_rate": 3.0439674561702474e-06,
+      "loss": 0.7802,
+      "num_input_tokens_seen": 60958585,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3455780676967474,
+      "grad_norm": 2.144542562852442,
+      "learning_rate": 3.0433029488960756e-06,
+      "loss": 0.8803,
+      "num_input_tokens_seen": 60976910,
+      "step": 2874
+    },
+    {
+      "epoch": 0.34569831058738654,
+      "grad_norm": 2.140651205472323,
+      "learning_rate": 3.0426382833536756e-06,
+      "loss": 0.6014,
+      "num_input_tokens_seen": 60999985,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3458185534780256,
+      "grad_norm": 2.6059404639916135,
+      "learning_rate": 3.041973459643877e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 61019160,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3459387963686647,
+      "grad_norm": 2.3503444344744016,
+      "learning_rate": 3.041308477867534e-06,
+      "loss": 0.6683,
+      "num_input_tokens_seen": 61040130,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3460590392593038,
+      "grad_norm": 2.2131216137972465,
+      "learning_rate": 3.0406433381255214e-06,
+      "loss": 0.8356,
+      "num_input_tokens_seen": 61057885,
+      "step": 2878
+    },
+    {
+      "epoch": 0.34617928214994287,
+      "grad_norm": 2.47492379675046,
+      "learning_rate": 3.0399780405187425e-06,
+      "loss": 0.8165,
+      "num_input_tokens_seen": 61076600,
+      "step": 2879
+    },
+    {
+      "epoch": 0.346299525040582,
+      "grad_norm": 1.989623315912117,
+      "learning_rate": 3.0393125851481216e-06,
+      "loss": 0.7786,
+      "num_input_tokens_seen": 61096195,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3464197679312211,
+      "grad_norm": 3.8850219727060677,
+      "learning_rate": 3.038646972114608e-06,
+      "loss": 0.8624,
+      "num_input_tokens_seen": 61112240,
+      "step": 2881
+    },
+    {
+      "epoch": 0.34654001082186014,
+      "grad_norm": 1.8425940482598202,
+      "learning_rate": 3.037981201519174e-06,
+      "loss": 0.6665,
+      "num_input_tokens_seen": 61132560,
+      "step": 2882
+    },
+    {
+      "epoch": 0.34666025371249926,
+      "grad_norm": 2.0319774137362763,
+      "learning_rate": 3.0373152734628175e-06,
+      "loss": 0.7051,
+      "num_input_tokens_seen": 61150560,
+      "step": 2883
+    },
+    {
+      "epoch": 0.34678049660313837,
+      "grad_norm": 1.866237152941598,
+      "learning_rate": 3.0366491880465593e-06,
+      "loss": 0.7542,
+      "num_input_tokens_seen": 61168300,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3469007394937774,
+      "grad_norm": 1.8519316468060965,
+      "learning_rate": 3.035982945371443e-06,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 61189715,
+      "step": 2885
+    },
+    {
+      "epoch": 0.34702098238441653,
+      "grad_norm": 8.631024270860742,
+      "learning_rate": 3.035316545538537e-06,
+      "loss": 0.8454,
+      "num_input_tokens_seen": 61208230,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3471412252750556,
+      "grad_norm": 2.5055331254599857,
+      "learning_rate": 3.034649988648935e-06,
+      "loss": 0.7953,
+      "num_input_tokens_seen": 61227715,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3472614681656947,
+      "grad_norm": 1.8982922427974478,
+      "learning_rate": 3.033983274803752e-06,
+      "loss": 0.8086,
+      "num_input_tokens_seen": 61247225,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3473817110563338,
+      "grad_norm": 2.183143383289986,
+      "learning_rate": 3.0333164041041283e-06,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 61263260,
+      "step": 2889
+    },
+    {
+      "epoch": 0.34750195394697286,
+      "grad_norm": 2.341490343468796,
+      "learning_rate": 3.0326493766512277e-06,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 61282400,
+      "step": 2890
+    },
+    {
+      "epoch": 0.347622196837612,
+      "grad_norm": 2.8541983930137103,
+      "learning_rate": 3.0319821925462377e-06,
+      "loss": 0.7553,
+      "num_input_tokens_seen": 61305215,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3477424397282511,
+      "grad_norm": 3.1364684002491825,
+      "learning_rate": 3.0313148518903696e-06,
+      "loss": 0.9432,
+      "num_input_tokens_seen": 61324760,
+      "step": 2892
+    },
+    {
+      "epoch": 0.34786268261889014,
+      "grad_norm": 2.65460936854414,
+      "learning_rate": 3.0306473547848593e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 61341520,
+      "step": 2893
+    },
+    {
+      "epoch": 0.34798292550952925,
+      "grad_norm": 1.9559459850183425,
+      "learning_rate": 3.029979701330964e-06,
+      "loss": 0.7681,
+      "num_input_tokens_seen": 61360665,
+      "step": 2894
+    },
+    {
+      "epoch": 0.34810316840016836,
+      "grad_norm": 2.7312612243689376,
+      "learning_rate": 3.0293118916299668e-06,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 61378840,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3482234112908074,
+      "grad_norm": 1.8766845125557394,
+      "learning_rate": 3.0286439257831735e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 61398030,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3483436541814465,
+      "grad_norm": 2.121773912955864,
+      "learning_rate": 3.0279758038919156e-06,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 61415975,
+      "step": 2897
+    },
+    {
+      "epoch": 0.34846389707208564,
+      "grad_norm": 2.3054469818517993,
+      "learning_rate": 3.0273075260575455e-06,
+      "loss": 0.7788,
+      "num_input_tokens_seen": 61434595,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3485841399627247,
+      "grad_norm": 1.8807126879633407,
+      "learning_rate": 3.0266390923814404e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 61452375,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3487043828533638,
+      "grad_norm": 1.7987034315937522,
+      "learning_rate": 3.0259705029650008e-06,
+      "loss": 0.8192,
+      "num_input_tokens_seen": 61470025,
+      "step": 2900
+    },
+    {
+      "epoch": 0.34882462574400286,
+      "grad_norm": 1.7613147697472733,
+      "learning_rate": 3.025301757909652e-06,
+      "loss": 0.7245,
+      "num_input_tokens_seen": 61489940,
+      "step": 2901
+    },
+    {
+      "epoch": 0.34894486863464197,
+      "grad_norm": 2.25089082353769,
+      "learning_rate": 3.0246328573168414e-06,
+      "loss": 0.802,
+      "num_input_tokens_seen": 61510975,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3490651115252811,
+      "grad_norm": 2.0417213987288814,
+      "learning_rate": 3.0239638012880412e-06,
+      "loss": 0.781,
+      "num_input_tokens_seen": 61530590,
+      "step": 2903
+    },
+    {
+      "epoch": 0.34918535441592014,
+      "grad_norm": 2.7389571660604446,
+      "learning_rate": 3.023294589924746e-06,
+      "loss": 0.8024,
+      "num_input_tokens_seen": 61547245,
+      "step": 2904
+    },
+    {
+      "epoch": 0.34930559730655925,
+      "grad_norm": 2.2009726955070246,
+      "learning_rate": 3.022625223328476e-06,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 61568705,
+      "step": 2905
+    },
+    {
+      "epoch": 0.34942584019719836,
+      "grad_norm": 1.4464638904550626,
+      "learning_rate": 3.0219557016007727e-06,
+      "loss": 0.6865,
+      "num_input_tokens_seen": 61588555,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3495460830878374,
+      "grad_norm": 2.0068703189783434,
+      "learning_rate": 3.021286024843202e-06,
+      "loss": 0.6964,
+      "num_input_tokens_seen": 61606470,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3496663259784765,
+      "grad_norm": 1.1287761145393838,
+      "learning_rate": 3.0206161931573526e-06,
+      "loss": 0.6994,
+      "num_input_tokens_seen": 61658740,
+      "step": 2908
+    },
+    {
+      "epoch": 0.34978656886911563,
+      "grad_norm": 1.7189752630444677,
+      "learning_rate": 3.0199462066448388e-06,
+      "loss": 0.9246,
+      "num_input_tokens_seen": 61680655,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3499068117597547,
+      "grad_norm": 1.7926214310382942,
+      "learning_rate": 3.019276065407296e-06,
+      "loss": 0.6894,
+      "num_input_tokens_seen": 61699495,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3500270546503938,
+      "grad_norm": 2.2809386669017786,
+      "learning_rate": 3.018605769546385e-06,
+      "loss": 0.8036,
+      "num_input_tokens_seen": 61719770,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3501472975410329,
+      "grad_norm": 5.424755331499995,
+      "learning_rate": 3.0179353191637876e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 61738450,
+      "step": 2912
+    },
+    {
+      "epoch": 0.35026754043167196,
+      "grad_norm": 1.777364050476785,
+      "learning_rate": 3.0172647143612125e-06,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 61757820,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3503877833223111,
+      "grad_norm": 1.9990078594526504,
+      "learning_rate": 3.016593955240389e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 61776230,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3505080262129502,
+      "grad_norm": 0.821546981645845,
+      "learning_rate": 3.015923041903071e-06,
+      "loss": 0.6555,
+      "num_input_tokens_seen": 61842075,
+      "step": 2915
+    },
+    {
+      "epoch": 0.35062826910358924,
+      "grad_norm": 2.109389742120155,
+      "learning_rate": 3.0152519744510347e-06,
+      "loss": 0.8299,
+      "num_input_tokens_seen": 61861595,
+      "step": 2916
+    },
+    {
+      "epoch": 0.35074851199422835,
+      "grad_norm": 1.82926646141423,
+      "learning_rate": 3.014580752986082e-06,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 61880190,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3508687548848674,
+      "grad_norm": 2.1995226057289767,
+      "learning_rate": 3.0139093776100345e-06,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 61896500,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3509889977755065,
+      "grad_norm": 1.7525718147909617,
+      "learning_rate": 3.013237848424741e-06,
+      "loss": 0.7463,
+      "num_input_tokens_seen": 61915605,
+      "step": 2919
+    },
+    {
+      "epoch": 0.35110924066614563,
+      "grad_norm": 2.198443869401436,
+      "learning_rate": 3.012566165532072e-06,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 61934115,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3512294835567847,
+      "grad_norm": 2.455083748335562,
+      "learning_rate": 3.0118943290339207e-06,
+      "loss": 0.7628,
+      "num_input_tokens_seen": 61954045,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3513497264474238,
+      "grad_norm": 1.744022161688745,
+      "learning_rate": 3.011222339032204e-06,
+      "loss": 0.6749,
+      "num_input_tokens_seen": 61971915,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3514699693380629,
+      "grad_norm": 2.386340831268044,
+      "learning_rate": 3.0105501956288626e-06,
+      "loss": 0.6872,
+      "num_input_tokens_seen": 61992105,
+      "step": 2923
+    },
+    {
+      "epoch": 0.35159021222870196,
+      "grad_norm": 3.1679533381969858,
+      "learning_rate": 3.0098778989258594e-06,
+      "loss": 0.7361,
+      "num_input_tokens_seen": 62010435,
+      "step": 2924
+    },
+    {
+      "epoch": 0.35171045511934107,
+      "grad_norm": 2.300553385752973,
+      "learning_rate": 3.009205449025183e-06,
+      "loss": 0.8762,
+      "num_input_tokens_seen": 62026350,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3518306980099802,
+      "grad_norm": 2.0321620335432033,
+      "learning_rate": 3.0085328460288415e-06,
+      "loss": 0.6233,
+      "num_input_tokens_seen": 62042830,
+      "step": 2926
+    },
+    {
+      "epoch": 0.35195094090061924,
+      "grad_norm": 2.3221737163131704,
+      "learning_rate": 3.0078600900388694e-06,
+      "loss": 0.706,
+      "num_input_tokens_seen": 62062855,
+      "step": 2927
+    },
+    {
+      "epoch": 0.35207118379125835,
+      "grad_norm": 1.8354561557972358,
+      "learning_rate": 3.007187181157323e-06,
+      "loss": 0.7342,
+      "num_input_tokens_seen": 62082585,
+      "step": 2928
+    },
+    {
+      "epoch": 0.35219142668189746,
+      "grad_norm": 2.705029204385408,
+      "learning_rate": 3.006514119486282e-06,
+      "loss": 0.6695,
+      "num_input_tokens_seen": 62099135,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3523116695725365,
+      "grad_norm": 2.421581983842595,
+      "learning_rate": 3.005840905127849e-06,
+      "loss": 0.6886,
+      "num_input_tokens_seen": 62115760,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3524319124631756,
+      "grad_norm": 2.4685443049028115,
+      "learning_rate": 3.0051675381841516e-06,
+      "loss": 0.866,
+      "num_input_tokens_seen": 62132790,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3525521553538147,
+      "grad_norm": 1.5695941660375252,
+      "learning_rate": 3.0044940187573363e-06,
+      "loss": 0.7638,
+      "num_input_tokens_seen": 62153520,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3526723982444538,
+      "grad_norm": 2.016889098332623,
+      "learning_rate": 3.003820346949578e-06,
+      "loss": 0.6576,
+      "num_input_tokens_seen": 62171320,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3527926411350929,
+      "grad_norm": 2.681639029178212,
+      "learning_rate": 3.0031465228630708e-06,
+      "loss": 0.7949,
+      "num_input_tokens_seen": 62191925,
+      "step": 2934
+    },
+    {
+      "epoch": 0.35291288402573195,
+      "grad_norm": 2.6267459798754023,
+      "learning_rate": 3.0024725466000337e-06,
+      "loss": 0.8599,
+      "num_input_tokens_seen": 62211600,
+      "step": 2935
+    },
+    {
+      "epoch": 0.35303312691637107,
+      "grad_norm": 2.145509044175393,
+      "learning_rate": 3.0017984182627087e-06,
+      "loss": 0.7887,
+      "num_input_tokens_seen": 62230645,
+      "step": 2936
+    },
+    {
+      "epoch": 0.3531533698070102,
+      "grad_norm": 1.9341048340175657,
+      "learning_rate": 3.00112413795336e-06,
+      "loss": 0.8163,
+      "num_input_tokens_seen": 62250200,
+      "step": 2937
+    },
+    {
+      "epoch": 0.35327361269764923,
+      "grad_norm": 2.2827976822780376,
+      "learning_rate": 3.000449705774275e-06,
+      "loss": 0.801,
+      "num_input_tokens_seen": 62268160,
+      "step": 2938
+    },
+    {
+      "epoch": 0.35339385558828834,
+      "grad_norm": 2.105138400041192,
+      "learning_rate": 2.9997751218277663e-06,
+      "loss": 0.7094,
+      "num_input_tokens_seen": 62286035,
+      "step": 2939
+    },
+    {
+      "epoch": 0.35351409847892745,
+      "grad_norm": 2.34830310939989,
+      "learning_rate": 2.9991003862161655e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 62304695,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3536343413695665,
+      "grad_norm": 1.8490820719029954,
+      "learning_rate": 2.998425499041831e-06,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 62324930,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3537545842602056,
+      "grad_norm": 0.9134622313556636,
+      "learning_rate": 2.997750460407142e-06,
+      "loss": 0.6281,
+      "num_input_tokens_seen": 62386005,
+      "step": 2942
+    },
+    {
+      "epoch": 0.35387482715084473,
+      "grad_norm": 2.844586058302895,
+      "learning_rate": 2.9970752704145014e-06,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 62402940,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3539950700414838,
+      "grad_norm": 0.739508397987474,
+      "learning_rate": 2.9963999291663347e-06,
+      "loss": 0.5973,
+      "num_input_tokens_seen": 62468440,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3541153129321229,
+      "grad_norm": 2.776464374577834,
+      "learning_rate": 2.9957244367650915e-06,
+      "loss": 0.7467,
+      "num_input_tokens_seen": 62484405,
+      "step": 2945
+    },
+    {
+      "epoch": 0.354235555822762,
+      "grad_norm": 3.3124182071216297,
+      "learning_rate": 2.995048793313242e-06,
+      "loss": 0.8318,
+      "num_input_tokens_seen": 62501540,
+      "step": 2946
+    },
+    {
+      "epoch": 0.35435579871340106,
+      "grad_norm": 2.2550764404008317,
+      "learning_rate": 2.994372998913283e-06,
+      "loss": 0.7139,
+      "num_input_tokens_seen": 62519765,
+      "step": 2947
+    },
+    {
+      "epoch": 0.35447604160404017,
+      "grad_norm": 2.519482606541746,
+      "learning_rate": 2.9936970536677297e-06,
+      "loss": 0.6328,
+      "num_input_tokens_seen": 62539730,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3545962844946792,
+      "grad_norm": 2.371282385388547,
+      "learning_rate": 2.9930209576791244e-06,
+      "loss": 0.8133,
+      "num_input_tokens_seen": 62557925,
+      "step": 2949
+    },
+    {
+      "epoch": 0.35471652738531834,
+      "grad_norm": 2.3488908393388663,
+      "learning_rate": 2.9923447110500285e-06,
+      "loss": 0.6346,
+      "num_input_tokens_seen": 62576390,
+      "step": 2950
+    },
+    {
+      "epoch": 0.35483677027595745,
+      "grad_norm": 1.6458689680897258,
+      "learning_rate": 2.9916683138830295e-06,
+      "loss": 0.7464,
+      "num_input_tokens_seen": 62596775,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3549570131665965,
+      "grad_norm": 2.8413688045660117,
+      "learning_rate": 2.9909917662807353e-06,
+      "loss": 0.8089,
+      "num_input_tokens_seen": 62614295,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3550772560572356,
+      "grad_norm": 2.2302194121121053,
+      "learning_rate": 2.9903150683457783e-06,
+      "loss": 0.6954,
+      "num_input_tokens_seen": 62632560,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3551974989478747,
+      "grad_norm": 1.7761226409755375,
+      "learning_rate": 2.9896382201808126e-06,
+      "loss": 0.6512,
+      "num_input_tokens_seen": 62649680,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3553177418385138,
+      "grad_norm": 2.1304844349914287,
+      "learning_rate": 2.988961221888516e-06,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 62666075,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3554379847291529,
+      "grad_norm": 2.232578825519733,
+      "learning_rate": 2.9882840735715884e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 62681880,
+      "step": 2956
+    },
+    {
+      "epoch": 0.355558227619792,
+      "grad_norm": 2.479938573774159,
+      "learning_rate": 2.9876067753327523e-06,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 62699330,
+      "step": 2957
+    },
+    {
+      "epoch": 0.35567847051043106,
+      "grad_norm": 2.1983717197979606,
+      "learning_rate": 2.986929327274754e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 62719630,
+      "step": 2958
+    },
+    {
+      "epoch": 0.35579871340107017,
+      "grad_norm": 1.770533305287497,
+      "learning_rate": 2.9862517295003617e-06,
+      "loss": 0.7774,
+      "num_input_tokens_seen": 62739765,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3559189562917093,
+      "grad_norm": 1.6030427383627186,
+      "learning_rate": 2.9855739821123654e-06,
+      "loss": 0.7241,
+      "num_input_tokens_seen": 62761065,
+      "step": 2960
+    },
+    {
+      "epoch": 0.35603919918234833,
+      "grad_norm": 1.9943763604621525,
+      "learning_rate": 2.9848960852135803e-06,
+      "loss": 0.8108,
+      "num_input_tokens_seen": 62780725,
+      "step": 2961
+    },
+    {
+      "epoch": 0.35615944207298744,
+      "grad_norm": 2.3115840141860677,
+      "learning_rate": 2.9842180389068417e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 62797755,
+      "step": 2962
+    },
+    {
+      "epoch": 0.35627968496362655,
+      "grad_norm": 0.8163308754617485,
+      "learning_rate": 2.98353984329501e-06,
+      "loss": 0.6227,
+      "num_input_tokens_seen": 62861820,
+      "step": 2963
+    },
+    {
+      "epoch": 0.3563999278542656,
+      "grad_norm": 1.7175358141494248,
+      "learning_rate": 2.982861498480965e-06,
+      "loss": 0.7102,
+      "num_input_tokens_seen": 62883920,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3565201707449047,
+      "grad_norm": 2.0082443680284277,
+      "learning_rate": 2.9821830045676126e-06,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 62903340,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3566404136355438,
+      "grad_norm": 1.8173087076135495,
+      "learning_rate": 2.9815043616578793e-06,
+      "loss": 0.7123,
+      "num_input_tokens_seen": 62923855,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3567606565261829,
+      "grad_norm": 2.056619677034643,
+      "learning_rate": 2.9808255698547145e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 62946375,
+      "step": 2967
+    },
+    {
+      "epoch": 0.356880899416822,
+      "grad_norm": 2.0887074427988335,
+      "learning_rate": 2.980146629261091e-06,
+      "loss": 0.7839,
+      "num_input_tokens_seen": 62965980,
+      "step": 2968
+    },
+    {
+      "epoch": 0.35700114230746105,
+      "grad_norm": 2.0145808948915787,
+      "learning_rate": 2.979467539980003e-06,
+      "loss": 0.8084,
+      "num_input_tokens_seen": 62982490,
+      "step": 2969
+    },
+    {
+      "epoch": 0.35712138519810016,
+      "grad_norm": 2.0953020651916066,
+      "learning_rate": 2.9787883021144675e-06,
+      "loss": 0.7652,
+      "num_input_tokens_seen": 62999325,
+      "step": 2970
+    },
+    {
+      "epoch": 0.35724162808873927,
+      "grad_norm": 2.2707959005445884,
+      "learning_rate": 2.9781089157675255e-06,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 63017505,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3573618709793783,
+      "grad_norm": 1.6867566023035991,
+      "learning_rate": 2.9774293810422384e-06,
+      "loss": 0.8743,
+      "num_input_tokens_seen": 63037900,
+      "step": 2972
+    },
+    {
+      "epoch": 0.35748211387001744,
+      "grad_norm": 2.776112663456812,
+      "learning_rate": 2.9767496980416913e-06,
+      "loss": 0.8936,
+      "num_input_tokens_seen": 63056915,
+      "step": 2973
+    },
+    {
+      "epoch": 0.35760235676065655,
+      "grad_norm": 2.630724038701508,
+      "learning_rate": 2.9760698668689914e-06,
+      "loss": 0.8093,
+      "num_input_tokens_seen": 63072860,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3577225996512956,
+      "grad_norm": 2.0763864326175367,
+      "learning_rate": 2.975389887627269e-06,
+      "loss": 0.7103,
+      "num_input_tokens_seen": 63095180,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3578428425419347,
+      "grad_norm": 2.6509259135292416,
+      "learning_rate": 2.9747097604196764e-06,
+      "loss": 0.8914,
+      "num_input_tokens_seen": 63111545,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3579630854325738,
+      "grad_norm": 0.6986299447828513,
+      "learning_rate": 2.9740294853493875e-06,
+      "loss": 0.5956,
+      "num_input_tokens_seen": 63182825,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3580833283232129,
+      "grad_norm": 2.1230706515879,
+      "learning_rate": 2.9733490625196004e-06,
+      "loss": 0.6761,
+      "num_input_tokens_seen": 63202405,
+      "step": 2978
+    },
+    {
+      "epoch": 0.358203571213852,
+      "grad_norm": 5.338732050837974,
+      "learning_rate": 2.9726684920335344e-06,
+      "loss": 0.7591,
+      "num_input_tokens_seen": 63219990,
+      "step": 2979
+    },
+    {
+      "epoch": 0.35832381410449105,
+      "grad_norm": 2.1166918865470423,
+      "learning_rate": 2.971987773994432e-06,
+      "loss": 0.8144,
+      "num_input_tokens_seen": 63235895,
+      "step": 2980
+    },
+    {
+      "epoch": 0.35844405699513016,
+      "grad_norm": 2.05126906180908,
+      "learning_rate": 2.9713069085055566e-06,
+      "loss": 0.8266,
+      "num_input_tokens_seen": 63253925,
+      "step": 2981
+    },
+    {
+      "epoch": 0.35856429988576927,
+      "grad_norm": 2.7956062423960777,
+      "learning_rate": 2.9706258956701958e-06,
+      "loss": 0.7866,
+      "num_input_tokens_seen": 63273635,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3586845427764083,
+      "grad_norm": 2.318693226356334,
+      "learning_rate": 2.969944735591658e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 63292165,
+      "step": 2983
+    },
+    {
+      "epoch": 0.35880478566704743,
+      "grad_norm": 3.218758764542374,
+      "learning_rate": 2.9692634283732747e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 63310235,
+      "step": 2984
+    },
+    {
+      "epoch": 0.35892502855768654,
+      "grad_norm": 2.33091356567361,
+      "learning_rate": 2.9685819741184007e-06,
+      "loss": 0.7861,
+      "num_input_tokens_seen": 63328395,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3590452714483256,
+      "grad_norm": 2.2145966297723865,
+      "learning_rate": 2.967900372930411e-06,
+      "loss": 0.6835,
+      "num_input_tokens_seen": 63346625,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3591655143389647,
+      "grad_norm": 4.425013318949799,
+      "learning_rate": 2.9672186249127046e-06,
+      "loss": 0.7923,
+      "num_input_tokens_seen": 63365810,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3592857572296038,
+      "grad_norm": 1.879413631296592,
+      "learning_rate": 2.9665367301687014e-06,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 63383775,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3594060001202429,
+      "grad_norm": 1.7996845906629475,
+      "learning_rate": 2.965854688801845e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 63405555,
+      "step": 2989
+    },
+    {
+      "epoch": 0.359526243010882,
+      "grad_norm": 1.9114645875708873,
+      "learning_rate": 2.9651725009156e-06,
+      "loss": 0.7557,
+      "num_input_tokens_seen": 63423020,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3596464859015211,
+      "grad_norm": 1.7552493677084657,
+      "learning_rate": 2.964490166613454e-06,
+      "loss": 0.7399,
+      "num_input_tokens_seen": 63442665,
+      "step": 2991
+    },
+    {
+      "epoch": 0.35976672879216015,
+      "grad_norm": 0.9002352766720059,
+      "learning_rate": 2.9638076859989167e-06,
+      "loss": 0.6019,
+      "num_input_tokens_seen": 63498250,
+      "step": 2992
+    },
+    {
+      "epoch": 0.35988697168279926,
+      "grad_norm": 1.7374379380919434,
+      "learning_rate": 2.9631250591755196e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 63520685,
+      "step": 2993
+    },
+    {
+      "epoch": 0.36000721457343837,
+      "grad_norm": 1.7819665656384225,
+      "learning_rate": 2.9624422862468174e-06,
+      "loss": 0.5795,
+      "num_input_tokens_seen": 63543235,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3601274574640774,
+      "grad_norm": 1.566950416111105,
+      "learning_rate": 2.9617593673163853e-06,
+      "loss": 0.6988,
+      "num_input_tokens_seen": 63561775,
+      "step": 2995
+    },
+    {
+      "epoch": 0.36024770035471654,
+      "grad_norm": 2.6489034606829023,
+      "learning_rate": 2.9610763024878216e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 63577000,
+      "step": 2996
+    },
+    {
+      "epoch": 0.3603679432453556,
+      "grad_norm": 1.8572348243700725,
+      "learning_rate": 2.960393091864747e-06,
+      "loss": 0.9045,
+      "num_input_tokens_seen": 63595100,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3604881861359947,
+      "grad_norm": 1.6965993778822916,
+      "learning_rate": 2.959709735550804e-06,
+      "loss": 0.7416,
+      "num_input_tokens_seen": 63614415,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3606084290266338,
+      "grad_norm": 2.272856191762267,
+      "learning_rate": 2.9590262336496575e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 63633865,
+      "step": 2999
+    },
+    {
+      "epoch": 0.36072867191727287,
+      "grad_norm": 8.09684440520056,
+      "learning_rate": 2.958342586264993e-06,
+      "loss": 0.8395,
+      "num_input_tokens_seen": 63651720,
+      "step": 3000
+    },
+    {
+      "epoch": 0.360848914807912,
+      "grad_norm": 2.1448108093479075,
+      "learning_rate": 2.957658793500521e-06,
+      "loss": 0.7411,
+      "num_input_tokens_seen": 63669520,
+      "step": 3001
+    },
+    {
+      "epoch": 0.3609691576985511,
+      "grad_norm": 2.7568503637191433,
+      "learning_rate": 2.9569748554599713e-06,
+      "loss": 0.7197,
+      "num_input_tokens_seen": 63684850,
+      "step": 3002
+    },
+    {
+      "epoch": 0.36108940058919015,
+      "grad_norm": 2.59302159519918,
+      "learning_rate": 2.956290772247097e-06,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 63703245,
+      "step": 3003
+    },
+    {
+      "epoch": 0.36120964347982926,
+      "grad_norm": 1.6845370843586114,
+      "learning_rate": 2.9556065439656724e-06,
+      "loss": 0.7384,
+      "num_input_tokens_seen": 63722015,
+      "step": 3004
+    },
+    {
+      "epoch": 0.36132988637046837,
+      "grad_norm": 1.8873139956097023,
+      "learning_rate": 2.9549221707194952e-06,
+      "loss": 0.8198,
+      "num_input_tokens_seen": 63740585,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3614501292611074,
+      "grad_norm": 2.627524458220418,
+      "learning_rate": 2.9542376526123835e-06,
+      "loss": 0.7312,
+      "num_input_tokens_seen": 63759355,
+      "step": 3006
+    },
+    {
+      "epoch": 0.36157037215174653,
+      "grad_norm": 2.0020328864982626,
+      "learning_rate": 2.9535529897481796e-06,
+      "loss": 0.8393,
+      "num_input_tokens_seen": 63776620,
+      "step": 3007
+    },
+    {
+      "epoch": 0.36169061504238564,
+      "grad_norm": 2.5911731016441015,
+      "learning_rate": 2.9528681822307446e-06,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 63793190,
+      "step": 3008
+    },
+    {
+      "epoch": 0.3618108579330247,
+      "grad_norm": 2.653503032899075,
+      "learning_rate": 2.9521832301639642e-06,
+      "loss": 0.8189,
+      "num_input_tokens_seen": 63812485,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3619311008236638,
+      "grad_norm": 2.654552430917696,
+      "learning_rate": 2.9514981336517448e-06,
+      "loss": 0.7346,
+      "num_input_tokens_seen": 63831975,
+      "step": 3010
+    },
+    {
+      "epoch": 0.36205134371430286,
+      "grad_norm": 2.0408650024544466,
+      "learning_rate": 2.950812892798015e-06,
+      "loss": 0.8074,
+      "num_input_tokens_seen": 63852590,
+      "step": 3011
+    },
+    {
+      "epoch": 0.362171586604942,
+      "grad_norm": 4.129123537617905,
+      "learning_rate": 2.950127507706725e-06,
+      "loss": 0.8669,
+      "num_input_tokens_seen": 63872930,
+      "step": 3012
+    },
+    {
+      "epoch": 0.3622918294955811,
+      "grad_norm": 1.5023418424140664,
+      "learning_rate": 2.949441978481848e-06,
+      "loss": 0.8802,
+      "num_input_tokens_seen": 63893550,
+      "step": 3013
+    },
+    {
+      "epoch": 0.36241207238622014,
+      "grad_norm": 1.9226042728650827,
+      "learning_rate": 2.9487563052273774e-06,
+      "loss": 0.7923,
+      "num_input_tokens_seen": 63910030,
+      "step": 3014
+    },
+    {
+      "epoch": 0.36253231527685925,
+      "grad_norm": 1.8449124732240965,
+      "learning_rate": 2.94807048804733e-06,
+      "loss": 0.8552,
+      "num_input_tokens_seen": 63929370,
+      "step": 3015
+    },
+    {
+      "epoch": 0.36265255816749836,
+      "grad_norm": 1.9996652290596735,
+      "learning_rate": 2.9473845270457434e-06,
+      "loss": 0.8923,
+      "num_input_tokens_seen": 63945905,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3627728010581374,
+      "grad_norm": 2.536462746047324,
+      "learning_rate": 2.946698422326677e-06,
+      "loss": 0.697,
+      "num_input_tokens_seen": 63963085,
+      "step": 3017
+    },
+    {
+      "epoch": 0.36289304394877653,
+      "grad_norm": 2.1359093062228625,
+      "learning_rate": 2.946012173994213e-06,
+      "loss": 0.7911,
+      "num_input_tokens_seen": 63982590,
+      "step": 3018
+    },
+    {
+      "epoch": 0.36301328683941564,
+      "grad_norm": 1.5035475918395433,
+      "learning_rate": 2.945325782152454e-06,
+      "loss": 0.6826,
+      "num_input_tokens_seen": 64005345,
+      "step": 3019
+    },
+    {
+      "epoch": 0.3631335297300547,
+      "grad_norm": 2.2323894326599247,
+      "learning_rate": 2.9446392469055257e-06,
+      "loss": 0.79,
+      "num_input_tokens_seen": 64023100,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3632537726206938,
+      "grad_norm": 1.737221133606062,
+      "learning_rate": 2.9439525683575745e-06,
+      "loss": 0.7973,
+      "num_input_tokens_seen": 64041740,
+      "step": 3021
+    },
+    {
+      "epoch": 0.3633740155113329,
+      "grad_norm": 2.08503699609348,
+      "learning_rate": 2.943265746612769e-06,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 64061030,
+      "step": 3022
+    },
+    {
+      "epoch": 0.36349425840197197,
+      "grad_norm": 1.7919308736027637,
+      "learning_rate": 2.9425787817753007e-06,
+      "loss": 0.7648,
+      "num_input_tokens_seen": 64079410,
+      "step": 3023
+    },
+    {
+      "epoch": 0.3636145012926111,
+      "grad_norm": 1.7915654004491517,
+      "learning_rate": 2.94189167394938e-06,
+      "loss": 0.7058,
+      "num_input_tokens_seen": 64101565,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3637347441832502,
+      "grad_norm": 2.0363889965813677,
+      "learning_rate": 2.941204423239241e-06,
+      "loss": 0.8063,
+      "num_input_tokens_seen": 64120160,
+      "step": 3025
+    },
+    {
+      "epoch": 0.36385498707388925,
+      "grad_norm": 2.0367616428703386,
+      "learning_rate": 2.9405170297491395e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 64139875,
+      "step": 3026
+    },
+    {
+      "epoch": 0.36397522996452836,
+      "grad_norm": 2.0612594899023007,
+      "learning_rate": 2.939829493583353e-06,
+      "loss": 0.7947,
+      "num_input_tokens_seen": 64156240,
+      "step": 3027
+    },
+    {
+      "epoch": 0.3640954728551674,
+      "grad_norm": 2.7556091863319243,
+      "learning_rate": 2.9391418148461785e-06,
+      "loss": 0.8385,
+      "num_input_tokens_seen": 64173375,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3642157157458065,
+      "grad_norm": 2.1433199359107844,
+      "learning_rate": 2.938453993641938e-06,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 64191470,
+      "step": 3029
+    },
+    {
+      "epoch": 0.36433595863644563,
+      "grad_norm": 4.405542301298672,
+      "learning_rate": 2.937766030074973e-06,
+      "loss": 0.7076,
+      "num_input_tokens_seen": 64208445,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3644562015270847,
+      "grad_norm": 2.4287224704479686,
+      "learning_rate": 2.937077924249647e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 64230755,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3645764444177238,
+      "grad_norm": 3.2121224738148015,
+      "learning_rate": 2.9363896762703443e-06,
+      "loss": 0.7531,
+      "num_input_tokens_seen": 64247540,
+      "step": 3032
+    },
+    {
+      "epoch": 0.3646966873083629,
+      "grad_norm": 1.913137721761046,
+      "learning_rate": 2.9357012862414725e-06,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 64266620,
+      "step": 3033
+    },
+    {
+      "epoch": 0.36481693019900197,
+      "grad_norm": 1.9202922316932172,
+      "learning_rate": 2.935012754267459e-06,
+      "loss": 0.7069,
+      "num_input_tokens_seen": 64288550,
+      "step": 3034
+    },
+    {
+      "epoch": 0.3649371730896411,
+      "grad_norm": 2.5904623241121456,
+      "learning_rate": 2.934324080452755e-06,
+      "loss": 0.7549,
+      "num_input_tokens_seen": 64306060,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3650574159802802,
+      "grad_norm": 1.4694981981703936,
+      "learning_rate": 2.93363526490183e-06,
+      "loss": 0.77,
+      "num_input_tokens_seen": 64325850,
+      "step": 3036
+    },
+    {
+      "epoch": 0.36517765887091924,
+      "grad_norm": 1.963954593361555,
+      "learning_rate": 2.9329463077191783e-06,
+      "loss": 0.6972,
+      "num_input_tokens_seen": 64348945,
+      "step": 3037
+    },
+    {
+      "epoch": 0.36529790176155835,
+      "grad_norm": 2.438383738817773,
+      "learning_rate": 2.9322572090093135e-06,
+      "loss": 0.6448,
+      "num_input_tokens_seen": 64367370,
+      "step": 3038
+    },
+    {
+      "epoch": 0.36541814465219746,
+      "grad_norm": 2.894055218806962,
+      "learning_rate": 2.9315679688767713e-06,
+      "loss": 0.7557,
+      "num_input_tokens_seen": 64385100,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3655383875428365,
+      "grad_norm": 1.6316060157478773,
+      "learning_rate": 2.9308785874261085e-06,
+      "loss": 0.6618,
+      "num_input_tokens_seen": 64405010,
+      "step": 3040
+    },
+    {
+      "epoch": 0.36565863043347563,
+      "grad_norm": 1.876012026933462,
+      "learning_rate": 2.9301890647619045e-06,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 64424025,
+      "step": 3041
+    },
+    {
+      "epoch": 0.36577887332411474,
+      "grad_norm": 2.7787159407877757,
+      "learning_rate": 2.929499400988759e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 64444905,
+      "step": 3042
+    },
+    {
+      "epoch": 0.3658991162147538,
+      "grad_norm": 2.1899876800885205,
+      "learning_rate": 2.9288095962112927e-06,
+      "loss": 0.65,
+      "num_input_tokens_seen": 64465330,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3660193591053929,
+      "grad_norm": 1.7544531307199034,
+      "learning_rate": 2.92811965053415e-06,
+      "loss": 0.8411,
+      "num_input_tokens_seen": 64482220,
+      "step": 3044
+    },
+    {
+      "epoch": 0.36613960199603196,
+      "grad_norm": 2.0742417733507046,
+      "learning_rate": 2.9274295640619946e-06,
+      "loss": 0.7826,
+      "num_input_tokens_seen": 64499070,
+      "step": 3045
+    },
+    {
+      "epoch": 0.36625984488667107,
+      "grad_norm": 1.7883302662293092,
+      "learning_rate": 2.9267393368995103e-06,
+      "loss": 0.7836,
+      "num_input_tokens_seen": 64518020,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3663800877773102,
+      "grad_norm": 2.3428617523356845,
+      "learning_rate": 2.926048969151407e-06,
+      "loss": 0.7489,
+      "num_input_tokens_seen": 64535025,
+      "step": 3047
+    },
+    {
+      "epoch": 0.36650033066794924,
+      "grad_norm": 2.0296602825414576,
+      "learning_rate": 2.92535846092241e-06,
+      "loss": 0.6777,
+      "num_input_tokens_seen": 64553760,
+      "step": 3048
+    },
+    {
+      "epoch": 0.36662057355858835,
+      "grad_norm": 1.7049836248754247,
+      "learning_rate": 2.9246678123172704e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 64573570,
+      "step": 3049
+    },
+    {
+      "epoch": 0.36674081644922746,
+      "grad_norm": 2.515448628718151,
+      "learning_rate": 2.923977023440759e-06,
+      "loss": 0.7385,
+      "num_input_tokens_seen": 64591595,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3668610593398665,
+      "grad_norm": 1.627641199869569,
+      "learning_rate": 2.9232860943976686e-06,
+      "loss": 0.679,
+      "num_input_tokens_seen": 64612050,
+      "step": 3051
+    },
+    {
+      "epoch": 0.3669813022305056,
+      "grad_norm": 1.6979178840805242,
+      "learning_rate": 2.9225950252928115e-06,
+      "loss": 0.834,
+      "num_input_tokens_seen": 64632620,
+      "step": 3052
+    },
+    {
+      "epoch": 0.36710154512114473,
+      "grad_norm": 2.498423249056174,
+      "learning_rate": 2.9219038162310225e-06,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 64650540,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3672217880117838,
+      "grad_norm": 2.489071678827203,
+      "learning_rate": 2.921212467317157e-06,
+      "loss": 0.818,
+      "num_input_tokens_seen": 64669705,
+      "step": 3054
+    },
+    {
+      "epoch": 0.3673420309024229,
+      "grad_norm": 2.8663521346021597,
+      "learning_rate": 2.920520978656093e-06,
+      "loss": 0.8045,
+      "num_input_tokens_seen": 64686390,
+      "step": 3055
+    },
+    {
+      "epoch": 0.367462273793062,
+      "grad_norm": 2.279010603084803,
+      "learning_rate": 2.9198293503527286e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 64707715,
+      "step": 3056
+    },
+    {
+      "epoch": 0.36758251668370107,
+      "grad_norm": 0.7723522210277356,
+      "learning_rate": 2.919137582511983e-06,
+      "loss": 0.6206,
+      "num_input_tokens_seen": 64763875,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3677027595743402,
+      "grad_norm": 2.201065780820753,
+      "learning_rate": 2.9184456752387964e-06,
+      "loss": 0.6371,
+      "num_input_tokens_seen": 64780520,
+      "step": 3058
+    },
+    {
+      "epoch": 0.36782300246497923,
+      "grad_norm": 1.93935933391005,
+      "learning_rate": 2.917753628638132e-06,
+      "loss": 0.6978,
+      "num_input_tokens_seen": 64800545,
+      "step": 3059
+    },
+    {
+      "epoch": 0.36794324535561834,
+      "grad_norm": 2.9581608736066745,
+      "learning_rate": 2.9170614428149716e-06,
+      "loss": 0.694,
+      "num_input_tokens_seen": 64818600,
+      "step": 3060
+    },
+    {
+      "epoch": 0.36806348824625745,
+      "grad_norm": 3.8907633849036585,
+      "learning_rate": 2.9163691178743195e-06,
+      "loss": 0.8667,
+      "num_input_tokens_seen": 64836970,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3681837311368965,
+      "grad_norm": 2.3375353506131606,
+      "learning_rate": 2.9156766539212006e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 64854335,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3683039740275356,
+      "grad_norm": 2.1947106055976895,
+      "learning_rate": 2.9149840510606614e-06,
+      "loss": 0.7144,
+      "num_input_tokens_seen": 64872710,
+      "step": 3063
+    },
+    {
+      "epoch": 0.36842421691817473,
+      "grad_norm": 1.0359801152085388,
+      "learning_rate": 2.914291309397769e-06,
+      "loss": 0.686,
+      "num_input_tokens_seen": 64929900,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3685444598088138,
+      "grad_norm": 2.264296314002788,
+      "learning_rate": 2.9135984290376117e-06,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 64948485,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3686647026994529,
+      "grad_norm": 1.7814301011671463,
+      "learning_rate": 2.9129054100853e-06,
+      "loss": 0.8244,
+      "num_input_tokens_seen": 64967045,
+      "step": 3066
+    },
+    {
+      "epoch": 0.368784945590092,
+      "grad_norm": 7.660560302742403,
+      "learning_rate": 2.912212252645963e-06,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 64989350,
+      "step": 3067
+    },
+    {
+      "epoch": 0.36890518848073106,
+      "grad_norm": 2.7484492038557082,
+      "learning_rate": 2.9115189568247523e-06,
+      "loss": 0.7556,
+      "num_input_tokens_seen": 65006630,
+      "step": 3068
+    },
+    {
+      "epoch": 0.36902543137137017,
+      "grad_norm": 2.517116487646983,
+      "learning_rate": 2.910825522726841e-06,
+      "loss": 0.9067,
+      "num_input_tokens_seen": 65023875,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3691456742620093,
+      "grad_norm": 2.1179414415213693,
+      "learning_rate": 2.9101319504574215e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 65040035,
+      "step": 3070
+    },
+    {
+      "epoch": 0.36926591715264834,
+      "grad_norm": 2.2375768482085525,
+      "learning_rate": 2.909438240121709e-06,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 65060030,
+      "step": 3071
+    },
+    {
+      "epoch": 0.36938616004328745,
+      "grad_norm": 1.8559233221752391,
+      "learning_rate": 2.9087443918249385e-06,
+      "loss": 0.6966,
+      "num_input_tokens_seen": 65080770,
+      "step": 3072
+    },
+    {
+      "epoch": 0.36950640293392656,
+      "grad_norm": 2.140031799912235,
+      "learning_rate": 2.908050405672367e-06,
+      "loss": 0.7873,
+      "num_input_tokens_seen": 65100035,
+      "step": 3073
+    },
+    {
+      "epoch": 0.3696266458245656,
+      "grad_norm": 1.9494531830024715,
+      "learning_rate": 2.9073562817692703e-06,
+      "loss": 0.7851,
+      "num_input_tokens_seen": 65118440,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3697468887152047,
+      "grad_norm": 0.7905469202754124,
+      "learning_rate": 2.906662020220947e-06,
+      "loss": 0.6066,
+      "num_input_tokens_seen": 65180650,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3698671316058438,
+      "grad_norm": 3.0442508619622384,
+      "learning_rate": 2.905967621132716e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 65197980,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3699873744964829,
+      "grad_norm": 2.283642788158564,
+      "learning_rate": 2.905273084609918e-06,
+      "loss": 0.7504,
+      "num_input_tokens_seen": 65219045,
+      "step": 3077
+    },
+    {
+      "epoch": 0.370107617387122,
+      "grad_norm": 0.9049793891796653,
+      "learning_rate": 2.904578410757912e-06,
+      "loss": 0.6469,
+      "num_input_tokens_seen": 65278870,
+      "step": 3078
+    },
+    {
+      "epoch": 0.37022786027776106,
+      "grad_norm": 2.0538122310865927,
+      "learning_rate": 2.9038835996820807e-06,
+      "loss": 0.6639,
+      "num_input_tokens_seen": 65296200,
+      "step": 3079
+    },
+    {
+      "epoch": 0.37034810316840017,
+      "grad_norm": 5.630774452643147,
+      "learning_rate": 2.9031886514878258e-06,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 65314475,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3704683460590393,
+      "grad_norm": 2.2202352077880287,
+      "learning_rate": 2.902493566280571e-06,
+      "loss": 0.866,
+      "num_input_tokens_seen": 65332300,
+      "step": 3081
+    },
+    {
+      "epoch": 0.37058858894967833,
+      "grad_norm": 1.8727087801933924,
+      "learning_rate": 2.9017983441657595e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 65349350,
+      "step": 3082
+    },
+    {
+      "epoch": 0.37070883184031744,
+      "grad_norm": 3.2382469697965375,
+      "learning_rate": 2.9011029852488564e-06,
+      "loss": 0.7556,
+      "num_input_tokens_seen": 65366305,
+      "step": 3083
+    },
+    {
+      "epoch": 0.37082907473095655,
+      "grad_norm": 1.01107184616564,
+      "learning_rate": 2.9004074896353465e-06,
+      "loss": 0.6761,
+      "num_input_tokens_seen": 65420025,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3709493176215956,
+      "grad_norm": 1.9543161183791637,
+      "learning_rate": 2.8997118574307362e-06,
+      "loss": 0.8105,
+      "num_input_tokens_seen": 65436700,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3710695605122347,
+      "grad_norm": 2.155761927349376,
+      "learning_rate": 2.899016088740553e-06,
+      "loss": 0.737,
+      "num_input_tokens_seen": 65454530,
+      "step": 3086
+    },
+    {
+      "epoch": 0.37118980340287383,
+      "grad_norm": 2.4060505105930634,
+      "learning_rate": 2.898320183670344e-06,
+      "loss": 0.7902,
+      "num_input_tokens_seen": 65471665,
+      "step": 3087
+    },
+    {
+      "epoch": 0.3713100462935129,
+      "grad_norm": 7.531004831037133,
+      "learning_rate": 2.8976241423256767e-06,
+      "loss": 0.8779,
+      "num_input_tokens_seen": 65491480,
+      "step": 3088
+    },
+    {
+      "epoch": 0.371430289184152,
+      "grad_norm": 2.3132560982893535,
+      "learning_rate": 2.896927964812142e-06,
+      "loss": 0.6782,
+      "num_input_tokens_seen": 65511765,
+      "step": 3089
+    },
+    {
+      "epoch": 0.37155053207479105,
+      "grad_norm": 2.598919673939607,
+      "learning_rate": 2.8962316512353465e-06,
+      "loss": 0.7524,
+      "num_input_tokens_seen": 65529030,
+      "step": 3090
+    },
+    {
+      "epoch": 0.37167077496543016,
+      "grad_norm": 1.8580821851049534,
+      "learning_rate": 2.8955352017009233e-06,
+      "loss": 0.7461,
+      "num_input_tokens_seen": 65547995,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3717910178560693,
+      "grad_norm": 2.1379644631233687,
+      "learning_rate": 2.8948386163145212e-06,
+      "loss": 0.7674,
+      "num_input_tokens_seen": 65566925,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3719112607467083,
+      "grad_norm": 1.885953940845471,
+      "learning_rate": 2.8941418951818135e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 65586205,
+      "step": 3093
+    },
+    {
+      "epoch": 0.37203150363734744,
+      "grad_norm": 2.275389973348851,
+      "learning_rate": 2.89344503840849e-06,
+      "loss": 0.7089,
+      "num_input_tokens_seen": 65603440,
+      "step": 3094
+    },
+    {
+      "epoch": 0.37215174652798655,
+      "grad_norm": 2.4022548280684246,
+      "learning_rate": 2.8927480461002653e-06,
+      "loss": 0.7087,
+      "num_input_tokens_seen": 65623130,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3722719894186256,
+      "grad_norm": 11.143873019262802,
+      "learning_rate": 2.892050918362872e-06,
+      "loss": 0.8512,
+      "num_input_tokens_seen": 65637905,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3723922323092647,
+      "grad_norm": 0.9478235789567104,
+      "learning_rate": 2.891353655302063e-06,
+      "loss": 0.6145,
+      "num_input_tokens_seen": 65691680,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3725124751999038,
+      "grad_norm": 2.0488815779033938,
+      "learning_rate": 2.8906562570236137e-06,
+      "loss": 0.847,
+      "num_input_tokens_seen": 65709310,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3726327180905429,
+      "grad_norm": 1.5379489531203288,
+      "learning_rate": 2.8899587236333186e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 65727970,
+      "step": 3099
+    },
+    {
+      "epoch": 0.372752960981182,
+      "grad_norm": 1.9187932948780124,
+      "learning_rate": 2.8892610552369917e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 65749905,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3728732038718211,
+      "grad_norm": 1.956083552989559,
+      "learning_rate": 2.8885632519404704e-06,
+      "loss": 0.8148,
+      "num_input_tokens_seen": 65769895,
+      "step": 3101
+    },
+    {
+      "epoch": 0.37299344676246016,
+      "grad_norm": 1.9128049549934065,
+      "learning_rate": 2.8878653138496102e-06,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 65790110,
+      "step": 3102
+    },
+    {
+      "epoch": 0.37311368965309927,
+      "grad_norm": 2.5534878452091028,
+      "learning_rate": 2.8871672410702878e-06,
+      "loss": 0.7457,
+      "num_input_tokens_seen": 65807190,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3732339325437384,
+      "grad_norm": 1.7605907227540802,
+      "learning_rate": 2.8864690337084008e-06,
+      "loss": 0.813,
+      "num_input_tokens_seen": 65826185,
+      "step": 3104
+    },
+    {
+      "epoch": 0.37335417543437743,
+      "grad_norm": 3.741086890953342,
+      "learning_rate": 2.885770691869866e-06,
+      "loss": 0.7714,
+      "num_input_tokens_seen": 65846785,
+      "step": 3105
+    },
+    {
+      "epoch": 0.37347441832501654,
+      "grad_norm": 2.7954387323893894,
+      "learning_rate": 2.8850722156606207e-06,
+      "loss": 0.7436,
+      "num_input_tokens_seen": 65864895,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3735946612156556,
+      "grad_norm": 2.2239892431697506,
+      "learning_rate": 2.8843736051866252e-06,
+      "loss": 0.6685,
+      "num_input_tokens_seen": 65883540,
+      "step": 3107
+    },
+    {
+      "epoch": 0.3737149041062947,
+      "grad_norm": 1.5507169942103427,
+      "learning_rate": 2.8836748605538557e-06,
+      "loss": 0.6847,
+      "num_input_tokens_seen": 65904900,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3738351469969338,
+      "grad_norm": 4.082018640371153,
+      "learning_rate": 2.882975981868313e-06,
+      "loss": 0.6283,
+      "num_input_tokens_seen": 65925005,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3739553898875729,
+      "grad_norm": 2.344429963585106,
+      "learning_rate": 2.8822769692360165e-06,
+      "loss": 0.6876,
+      "num_input_tokens_seen": 65946085,
+      "step": 3110
+    },
+    {
+      "epoch": 0.374075632778212,
+      "grad_norm": 2.3160869694092936,
+      "learning_rate": 2.881577822763005e-06,
+      "loss": 0.7651,
+      "num_input_tokens_seen": 65963755,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3741958756688511,
+      "grad_norm": 2.1012165506469582,
+      "learning_rate": 2.880878542555338e-06,
+      "loss": 0.8759,
+      "num_input_tokens_seen": 65981240,
+      "step": 3112
+    },
+    {
+      "epoch": 0.37431611855949015,
+      "grad_norm": 2.249785145647857,
+      "learning_rate": 2.8801791287190976e-06,
+      "loss": 0.8004,
+      "num_input_tokens_seen": 65998955,
+      "step": 3113
+    },
+    {
+      "epoch": 0.37443636145012926,
+      "grad_norm": 3.0832348730518295,
+      "learning_rate": 2.8794795813603817e-06,
+      "loss": 0.8509,
+      "num_input_tokens_seen": 66014140,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3745566043407684,
+      "grad_norm": 2.1879430614656044,
+      "learning_rate": 2.878779900585314e-06,
+      "loss": 0.8146,
+      "num_input_tokens_seen": 66031700,
+      "step": 3115
+    },
+    {
+      "epoch": 0.37467684723140743,
+      "grad_norm": 2.144839827741358,
+      "learning_rate": 2.8780800865000328e-06,
+      "loss": 0.7543,
+      "num_input_tokens_seen": 66052730,
+      "step": 3116
+    },
+    {
+      "epoch": 0.37479709012204654,
+      "grad_norm": 1.033878072715553,
+      "learning_rate": 2.877380139210702e-06,
+      "loss": 0.6525,
+      "num_input_tokens_seen": 66111120,
+      "step": 3117
+    },
+    {
+      "epoch": 0.37491733301268565,
+      "grad_norm": 6.552799514834817,
+      "learning_rate": 2.876680058823501e-06,
+      "loss": 0.7464,
+      "num_input_tokens_seen": 66131240,
+      "step": 3118
+    },
+    {
+      "epoch": 0.3750375759033247,
+      "grad_norm": 2.3640126969618147,
+      "learning_rate": 2.8759798454446323e-06,
+      "loss": 0.6613,
+      "num_input_tokens_seen": 66154125,
+      "step": 3119
+    },
+    {
+      "epoch": 0.3751578187939638,
+      "grad_norm": 4.519550186111322,
+      "learning_rate": 2.8752794991803173e-06,
+      "loss": 0.8101,
+      "num_input_tokens_seen": 66171530,
+      "step": 3120
+    },
+    {
+      "epoch": 0.37527806168460287,
+      "grad_norm": 7.341990834089306,
+      "learning_rate": 2.874579020136798e-06,
+      "loss": 0.7484,
+      "num_input_tokens_seen": 66187005,
+      "step": 3121
+    },
+    {
+      "epoch": 0.375398304575242,
+      "grad_norm": 2.2947574508359625,
+      "learning_rate": 2.873878408420337e-06,
+      "loss": 0.8453,
+      "num_input_tokens_seen": 66206800,
+      "step": 3122
+    },
+    {
+      "epoch": 0.3755185474658811,
+      "grad_norm": 1.7964611695745767,
+      "learning_rate": 2.873177664137216e-06,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 66227450,
+      "step": 3123
+    },
+    {
+      "epoch": 0.37563879035652015,
+      "grad_norm": 2.0733710424574454,
+      "learning_rate": 2.872476787393738e-06,
+      "loss": 0.6866,
+      "num_input_tokens_seen": 66251290,
+      "step": 3124
+    },
+    {
+      "epoch": 0.37575903324715926,
+      "grad_norm": 2.783782881930464,
+      "learning_rate": 2.871775778296225e-06,
+      "loss": 0.8738,
+      "num_input_tokens_seen": 66268100,
+      "step": 3125
+    },
+    {
+      "epoch": 0.37587927613779837,
+      "grad_norm": 2.526670701982994,
+      "learning_rate": 2.8710746369510196e-06,
+      "loss": 0.7877,
+      "num_input_tokens_seen": 66285805,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3759995190284374,
+      "grad_norm": 2.4288263163627817,
+      "learning_rate": 2.8703733634644846e-06,
+      "loss": 0.8293,
+      "num_input_tokens_seen": 66300280,
+      "step": 3127
+    },
+    {
+      "epoch": 0.37611976191907653,
+      "grad_norm": 1.6937571026810703,
+      "learning_rate": 2.8696719579430014e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 66319155,
+      "step": 3128
+    },
+    {
+      "epoch": 0.37624000480971564,
+      "grad_norm": 3.0614512671743634,
+      "learning_rate": 2.8689704204929747e-06,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 66338055,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3763602477003547,
+      "grad_norm": 1.942790693679138,
+      "learning_rate": 2.8682687512208253e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 66356785,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3764804905909938,
+      "grad_norm": 1.9450081062139966,
+      "learning_rate": 2.8675669502329972e-06,
+      "loss": 0.8071,
+      "num_input_tokens_seen": 66378035,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3766007334816329,
+      "grad_norm": 7.847783767388033,
+      "learning_rate": 2.8668650176359524e-06,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 66395575,
+      "step": 3132
+    },
+    {
+      "epoch": 0.376720976372272,
+      "grad_norm": 2.4890157923517484,
+      "learning_rate": 2.866162953536174e-06,
+      "loss": 0.7881,
+      "num_input_tokens_seen": 66416265,
+      "step": 3133
+    },
+    {
+      "epoch": 0.3768412192629111,
+      "grad_norm": 1.7937575236750953,
+      "learning_rate": 2.8654607580401634e-06,
+      "loss": 0.746,
+      "num_input_tokens_seen": 66435720,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3769614621535502,
+      "grad_norm": 0.9339269846878173,
+      "learning_rate": 2.8647584312544446e-06,
+      "loss": 0.6817,
+      "num_input_tokens_seen": 66500645,
+      "step": 3135
+    },
+    {
+      "epoch": 0.37708170504418925,
+      "grad_norm": 1.8014387929617797,
+      "learning_rate": 2.864055973285559e-06,
+      "loss": 0.8514,
+      "num_input_tokens_seen": 66522365,
+      "step": 3136
+    },
+    {
+      "epoch": 0.37720194793482836,
+      "grad_norm": 1.852460490666136,
+      "learning_rate": 2.8633533842400698e-06,
+      "loss": 0.8661,
+      "num_input_tokens_seen": 66542285,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3773221908254674,
+      "grad_norm": 1.903543331292333,
+      "learning_rate": 2.8626506642245576e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 66560855,
+      "step": 3138
+    },
+    {
+      "epoch": 0.37744243371610653,
+      "grad_norm": 1.488639823456665,
+      "learning_rate": 2.8619478133456265e-06,
+      "loss": 0.6979,
+      "num_input_tokens_seen": 66583275,
+      "step": 3139
+    },
+    {
+      "epoch": 0.37756267660674564,
+      "grad_norm": 2.0544944680788535,
+      "learning_rate": 2.8612448317098974e-06,
+      "loss": 0.7192,
+      "num_input_tokens_seen": 66603330,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3776829194973847,
+      "grad_norm": 2.2425091155970116,
+      "learning_rate": 2.860541719424012e-06,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 66621410,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3778031623880238,
+      "grad_norm": 3.547883488470654,
+      "learning_rate": 2.8598384765946315e-06,
+      "loss": 0.792,
+      "num_input_tokens_seen": 66639785,
+      "step": 3142
+    },
+    {
+      "epoch": 0.3779234052786629,
+      "grad_norm": 2.2029576359245446,
+      "learning_rate": 2.859135103328438e-06,
+      "loss": 0.7191,
+      "num_input_tokens_seen": 66659235,
+      "step": 3143
+    },
+    {
+      "epoch": 0.37804364816930197,
+      "grad_norm": 3.861109233663844,
+      "learning_rate": 2.8584315997321325e-06,
+      "loss": 0.834,
+      "num_input_tokens_seen": 66677960,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3781638910599411,
+      "grad_norm": 3.412038625841604,
+      "learning_rate": 2.8577279659124356e-06,
+      "loss": 0.7739,
+      "num_input_tokens_seen": 66695355,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3782841339505802,
+      "grad_norm": 1.8578129467557964,
+      "learning_rate": 2.8570242019760885e-06,
+      "loss": 0.8228,
+      "num_input_tokens_seen": 66712635,
+      "step": 3146
+    },
+    {
+      "epoch": 0.37840437684121925,
+      "grad_norm": 5.292381263964881,
+      "learning_rate": 2.8563203080298516e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 66733130,
+      "step": 3147
+    },
+    {
+      "epoch": 0.37852461973185836,
+      "grad_norm": 2.503168178923822,
+      "learning_rate": 2.855616284180505e-06,
+      "loss": 0.8857,
+      "num_input_tokens_seen": 66749900,
+      "step": 3148
+    },
+    {
+      "epoch": 0.37864486262249747,
+      "grad_norm": 0.9636375568937535,
+      "learning_rate": 2.8549121305348477e-06,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 66809405,
+      "step": 3149
+    },
+    {
+      "epoch": 0.3787651055131365,
+      "grad_norm": 2.860839388834743,
+      "learning_rate": 2.8542078471997e-06,
+      "loss": 0.8223,
+      "num_input_tokens_seen": 66826740,
+      "step": 3150
+    },
+    {
+      "epoch": 0.37888534840377563,
+      "grad_norm": 1.842984744524347,
+      "learning_rate": 2.8535034342819013e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 66843870,
+      "step": 3151
+    },
+    {
+      "epoch": 0.37900559129441475,
+      "grad_norm": 1.5723033472035364,
+      "learning_rate": 2.85279889188831e-06,
+      "loss": 0.7161,
+      "num_input_tokens_seen": 66863965,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3791258341850538,
+      "grad_norm": 1.782499440148793,
+      "learning_rate": 2.852094220125805e-06,
+      "loss": 0.8012,
+      "num_input_tokens_seen": 66883195,
+      "step": 3153
+    },
+    {
+      "epoch": 0.3792460770756929,
+      "grad_norm": 2.823779504213679,
+      "learning_rate": 2.851389419101285e-06,
+      "loss": 0.7022,
+      "num_input_tokens_seen": 66901895,
+      "step": 3154
+    },
+    {
+      "epoch": 0.37936631996633197,
+      "grad_norm": 2.359293594206734,
+      "learning_rate": 2.8506844889216664e-06,
+      "loss": 0.7786,
+      "num_input_tokens_seen": 66921000,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3794865628569711,
+      "grad_norm": 0.9069679432226965,
+      "learning_rate": 2.849979429693887e-06,
+      "loss": 0.6675,
+      "num_input_tokens_seen": 66981705,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3796068057476102,
+      "grad_norm": 2.5502282171571755,
+      "learning_rate": 2.8492742415249042e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 66999070,
+      "step": 3157
+    },
+    {
+      "epoch": 0.37972704863824924,
+      "grad_norm": 1.9709765330500568,
+      "learning_rate": 2.848568924521694e-06,
+      "loss": 0.7593,
+      "num_input_tokens_seen": 67019570,
+      "step": 3158
+    },
+    {
+      "epoch": 0.37984729152888835,
+      "grad_norm": 2.2528051229234825,
+      "learning_rate": 2.8478634787912526e-06,
+      "loss": 0.7264,
+      "num_input_tokens_seen": 67037345,
+      "step": 3159
+    },
+    {
+      "epoch": 0.37996753441952746,
+      "grad_norm": 2.940462319070035,
+      "learning_rate": 2.8471579044405954e-06,
+      "loss": 0.7635,
+      "num_input_tokens_seen": 67056795,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3800877773101665,
+      "grad_norm": 1.686815102329516,
+      "learning_rate": 2.846452201576758e-06,
+      "loss": 0.7463,
+      "num_input_tokens_seen": 67075890,
+      "step": 3161
+    },
+    {
+      "epoch": 0.38020802020080563,
+      "grad_norm": 1.0648944070946078,
+      "learning_rate": 2.845746370306795e-06,
+      "loss": 0.6504,
+      "num_input_tokens_seen": 67140800,
+      "step": 3162
+    },
+    {
+      "epoch": 0.38032826309144474,
+      "grad_norm": 2.1074651090939116,
+      "learning_rate": 2.84504041073778e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 67158935,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3804485059820838,
+      "grad_norm": 2.198766616879029,
+      "learning_rate": 2.844334322976806e-06,
+      "loss": 0.798,
+      "num_input_tokens_seen": 67178870,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3805687488727229,
+      "grad_norm": 9.09663277233536,
+      "learning_rate": 2.843628107130987e-06,
+      "loss": 0.8278,
+      "num_input_tokens_seen": 67197130,
+      "step": 3165
+    },
+    {
+      "epoch": 0.380688991763362,
+      "grad_norm": 0.7830222457122498,
+      "learning_rate": 2.8429217633074545e-06,
+      "loss": 0.5527,
+      "num_input_tokens_seen": 67259660,
+      "step": 3166
+    },
+    {
+      "epoch": 0.38080923465400107,
+      "grad_norm": 2.037505354391273,
+      "learning_rate": 2.842215291613361e-06,
+      "loss": 0.8273,
+      "num_input_tokens_seen": 67277760,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3809294775446402,
+      "grad_norm": 0.8465271274401072,
+      "learning_rate": 2.841508692155877e-06,
+      "loss": 0.6425,
+      "num_input_tokens_seen": 67340905,
+      "step": 3168
+    },
+    {
+      "epoch": 0.38104972043527924,
+      "grad_norm": 1.7681966864986254,
+      "learning_rate": 2.840801965042194e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 67360085,
+      "step": 3169
+    },
+    {
+      "epoch": 0.38116996332591835,
+      "grad_norm": 2.2557065318987655,
+      "learning_rate": 2.840095110379521e-06,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 67379325,
+      "step": 3170
+    },
+    {
+      "epoch": 0.38129020621655746,
+      "grad_norm": 0.7234750766036715,
+      "learning_rate": 2.8393881282750884e-06,
+      "loss": 0.5503,
+      "num_input_tokens_seen": 67441875,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3814104491071965,
+      "grad_norm": 2.9612366795459892,
+      "learning_rate": 2.8386810188361435e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 67458915,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3815306919978356,
+      "grad_norm": 2.4768780469564975,
+      "learning_rate": 2.837973782169955e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 67477010,
+      "step": 3173
+    },
+    {
+      "epoch": 0.38165093488847474,
+      "grad_norm": 0.8930919426773212,
+      "learning_rate": 2.8372664183838096e-06,
+      "loss": 0.6238,
+      "num_input_tokens_seen": 67539750,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3817711777791138,
+      "grad_norm": 2.2688095836356656,
+      "learning_rate": 2.836558927585015e-06,
+      "loss": 0.6863,
+      "num_input_tokens_seen": 67556440,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3818914206697529,
+      "grad_norm": 2.9819648635912137,
+      "learning_rate": 2.8358513098808957e-06,
+      "loss": 0.8098,
+      "num_input_tokens_seen": 67576475,
+      "step": 3176
+    },
+    {
+      "epoch": 0.382011663560392,
+      "grad_norm": 2.1356580426955025,
+      "learning_rate": 2.835143565378798e-06,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 67596660,
+      "step": 3177
+    },
+    {
+      "epoch": 0.38213190645103107,
+      "grad_norm": 2.2040490822486816,
+      "learning_rate": 2.8344356941860847e-06,
+      "loss": 0.7749,
+      "num_input_tokens_seen": 67616010,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3822521493416702,
+      "grad_norm": 2.6126826308733335,
+      "learning_rate": 2.8337276964101403e-06,
+      "loss": 0.6625,
+      "num_input_tokens_seen": 67636170,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3823723922323093,
+      "grad_norm": 5.052353377952411,
+      "learning_rate": 2.833019572158367e-06,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 67654325,
+      "step": 3180
+    },
+    {
+      "epoch": 0.38249263512294834,
+      "grad_norm": 2.997275011567839,
+      "learning_rate": 2.8323113215381872e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 67672390,
+      "step": 3181
+    },
+    {
+      "epoch": 0.38261287801358745,
+      "grad_norm": 2.166972875921478,
+      "learning_rate": 2.8316029446570416e-06,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 67690190,
+      "step": 3182
+    },
+    {
+      "epoch": 0.38273312090422656,
+      "grad_norm": 2.6722048896581603,
+      "learning_rate": 2.8308944416223904e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 67706560,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3828533637948656,
+      "grad_norm": 2.8594806509272312,
+      "learning_rate": 2.8301858125417134e-06,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 67726120,
+      "step": 3184
+    },
+    {
+      "epoch": 0.38297360668550473,
+      "grad_norm": 3.683428415865929,
+      "learning_rate": 2.8294770575225087e-06,
+      "loss": 0.7413,
+      "num_input_tokens_seen": 67745970,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3830938495761438,
+      "grad_norm": 5.559087873767548,
+      "learning_rate": 2.828768176672293e-06,
+      "loss": 0.8346,
+      "num_input_tokens_seen": 67764805,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3832140924667829,
+      "grad_norm": 2.447176294654098,
+      "learning_rate": 2.8280591700986048e-06,
+      "loss": 0.7202,
+      "num_input_tokens_seen": 67786390,
+      "step": 3187
+    },
+    {
+      "epoch": 0.383334335357422,
+      "grad_norm": 2.4761889783888043,
+      "learning_rate": 2.8273500379089986e-06,
+      "loss": 0.7514,
+      "num_input_tokens_seen": 67805550,
+      "step": 3188
+    },
+    {
+      "epoch": 0.38345457824806106,
+      "grad_norm": 3.866308095564575,
+      "learning_rate": 2.8266407802110496e-06,
+      "loss": 0.7896,
+      "num_input_tokens_seen": 67823525,
+      "step": 3189
+    },
+    {
+      "epoch": 0.3835748211387002,
+      "grad_norm": 2.356838230972668,
+      "learning_rate": 2.8259313971123506e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 67844365,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3836950640293393,
+      "grad_norm": 2.388726334124802,
+      "learning_rate": 2.825221888720517e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 67864775,
+      "step": 3191
+    },
+    {
+      "epoch": 0.38381530691997834,
+      "grad_norm": 1.812982963796719,
+      "learning_rate": 2.824512255143178e-06,
+      "loss": 0.8074,
+      "num_input_tokens_seen": 67883730,
+      "step": 3192
+    },
+    {
+      "epoch": 0.38393554981061745,
+      "grad_norm": 1.9273416151562353,
+      "learning_rate": 2.8238024964879855e-06,
+      "loss": 0.7896,
+      "num_input_tokens_seen": 67904345,
+      "step": 3193
+    },
+    {
+      "epoch": 0.38405579270125656,
+      "grad_norm": 2.3238009444830365,
+      "learning_rate": 2.823092612862609e-06,
+      "loss": 0.7657,
+      "num_input_tokens_seen": 67922560,
+      "step": 3194
+    },
+    {
+      "epoch": 0.3841760355918956,
+      "grad_norm": 2.4195216232002754,
+      "learning_rate": 2.822382604374738e-06,
+      "loss": 0.7852,
+      "num_input_tokens_seen": 67941205,
+      "step": 3195
+    },
+    {
+      "epoch": 0.3842962784825347,
+      "grad_norm": 2.7879406733664625,
+      "learning_rate": 2.8216724711320793e-06,
+      "loss": 0.6478,
+      "num_input_tokens_seen": 67960050,
+      "step": 3196
+    },
+    {
+      "epoch": 0.38441652137317384,
+      "grad_norm": 1.9042961595927328,
+      "learning_rate": 2.820962213242361e-06,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 67979100,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3845367642638129,
+      "grad_norm": 2.4164326987298734,
+      "learning_rate": 2.8202518308133273e-06,
+      "loss": 0.8441,
+      "num_input_tokens_seen": 67996095,
+      "step": 3198
+    },
+    {
+      "epoch": 0.384657007154452,
+      "grad_norm": 2.008362776419249,
+      "learning_rate": 2.8195413239527426e-06,
+      "loss": 0.7297,
+      "num_input_tokens_seen": 68015555,
+      "step": 3199
+    },
+    {
+      "epoch": 0.38477725004509106,
+      "grad_norm": 3.177418803460515,
+      "learning_rate": 2.8188306927683906e-06,
+      "loss": 0.8037,
+      "num_input_tokens_seen": 68034745,
+      "step": 3200
+    },
+    {
+      "epoch": 0.38489749293573017,
+      "grad_norm": 3.1801369650805147,
+      "learning_rate": 2.818119937368074e-06,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 68053100,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3850177358263693,
+      "grad_norm": 2.501343470982094,
+      "learning_rate": 2.817409057859613e-06,
+      "loss": 0.6515,
+      "num_input_tokens_seen": 68071810,
+      "step": 3202
+    },
+    {
+      "epoch": 0.38513797871700833,
+      "grad_norm": 2.7637374189971657,
+      "learning_rate": 2.8166980543508482e-06,
+      "loss": 0.7845,
+      "num_input_tokens_seen": 68087420,
+      "step": 3203
+    },
+    {
+      "epoch": 0.38525822160764744,
+      "grad_norm": 2.117494427375198,
+      "learning_rate": 2.8159869269496375e-06,
+      "loss": 0.7976,
+      "num_input_tokens_seen": 68105640,
+      "step": 3204
+    },
+    {
+      "epoch": 0.38537846449828655,
+      "grad_norm": 1.801854297618171,
+      "learning_rate": 2.8152756757638593e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 68123860,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3854987073889256,
+      "grad_norm": 2.1129966074856363,
+      "learning_rate": 2.8145643009014093e-06,
+      "loss": 0.8353,
+      "num_input_tokens_seen": 68142075,
+      "step": 3206
+    },
+    {
+      "epoch": 0.3856189502795647,
+      "grad_norm": 1.8512615219835713,
+      "learning_rate": 2.8138528024702023e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 68159690,
+      "step": 3207
+    },
+    {
+      "epoch": 0.38573919317020383,
+      "grad_norm": 2.7392680086253827,
+      "learning_rate": 2.8131411805781717e-06,
+      "loss": 0.7191,
+      "num_input_tokens_seen": 68179535,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3858594360608429,
+      "grad_norm": 3.189225758197121,
+      "learning_rate": 2.8124294353332716e-06,
+      "loss": 0.6427,
+      "num_input_tokens_seen": 68197930,
+      "step": 3209
+    },
+    {
+      "epoch": 0.385979678951482,
+      "grad_norm": 2.8845284035144494,
+      "learning_rate": 2.811717566843471e-06,
+      "loss": 0.771,
+      "num_input_tokens_seen": 68217310,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3860999218421211,
+      "grad_norm": 2.4025780283711846,
+      "learning_rate": 2.811005575216762e-06,
+      "loss": 0.6919,
+      "num_input_tokens_seen": 68235745,
+      "step": 3211
+    },
+    {
+      "epoch": 0.38622016473276016,
+      "grad_norm": 1.8744117195874066,
+      "learning_rate": 2.810293460561151e-06,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 68257100,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3863404076233993,
+      "grad_norm": 2.067944246693746,
+      "learning_rate": 2.8095812229846674e-06,
+      "loss": 0.6743,
+      "num_input_tokens_seen": 68276780,
+      "step": 3213
+    },
+    {
+      "epoch": 0.3864606505140384,
+      "grad_norm": 2.3616425111489168,
+      "learning_rate": 2.808868862595355e-06,
+      "loss": 0.691,
+      "num_input_tokens_seen": 68296745,
+      "step": 3214
+    },
+    {
+      "epoch": 0.38658089340467744,
+      "grad_norm": 2.228010454743298,
+      "learning_rate": 2.8081563795012795e-06,
+      "loss": 0.7955,
+      "num_input_tokens_seen": 68316090,
+      "step": 3215
+    },
+    {
+      "epoch": 0.38670113629531655,
+      "grad_norm": 2.4642472329205027,
+      "learning_rate": 2.807443773810524e-06,
+      "loss": 0.7312,
+      "num_input_tokens_seen": 68337070,
+      "step": 3216
+    },
+    {
+      "epoch": 0.3868213791859556,
+      "grad_norm": 19.586500644173306,
+      "learning_rate": 2.80673104563119e-06,
+      "loss": 0.8905,
+      "num_input_tokens_seen": 68357415,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3869416220765947,
+      "grad_norm": 1.935875430167441,
+      "learning_rate": 2.8060181950713976e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 68373925,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3870618649672338,
+      "grad_norm": 2.826078153465582,
+      "learning_rate": 2.805305222239286e-06,
+      "loss": 0.8065,
+      "num_input_tokens_seen": 68390900,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3871821078578729,
+      "grad_norm": 2.056429348415702,
+      "learning_rate": 2.8045921272430126e-06,
+      "loss": 0.7305,
+      "num_input_tokens_seen": 68410300,
+      "step": 3220
+    },
+    {
+      "epoch": 0.387302350748512,
+      "grad_norm": 2.299329305554979,
+      "learning_rate": 2.803878910190753e-06,
+      "loss": 0.7619,
+      "num_input_tokens_seen": 68426940,
+      "step": 3221
+    },
+    {
+      "epoch": 0.3874225936391511,
+      "grad_norm": 3.4718564099401212,
+      "learning_rate": 2.8031655711907017e-06,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 68440365,
+      "step": 3222
+    },
+    {
+      "epoch": 0.38754283652979016,
+      "grad_norm": 2.1749807856426346,
+      "learning_rate": 2.8024521103510723e-06,
+      "loss": 0.8059,
+      "num_input_tokens_seen": 68456855,
+      "step": 3223
+    },
+    {
+      "epoch": 0.38766307942042927,
+      "grad_norm": 1.8249165888820273,
+      "learning_rate": 2.8017385277800952e-06,
+      "loss": 0.7469,
+      "num_input_tokens_seen": 68474930,
+      "step": 3224
+    },
+    {
+      "epoch": 0.3877833223110684,
+      "grad_norm": 2.006745255579459,
+      "learning_rate": 2.8010248235860213e-06,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 68494765,
+      "step": 3225
+    },
+    {
+      "epoch": 0.38790356520170743,
+      "grad_norm": 0.8561899931852645,
+      "learning_rate": 2.800310997877119e-06,
+      "loss": 0.665,
+      "num_input_tokens_seen": 68555650,
+      "step": 3226
+    },
+    {
+      "epoch": 0.38802380809234654,
+      "grad_norm": 2.149592460969267,
+      "learning_rate": 2.799597050761674e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 68571575,
+      "step": 3227
+    },
+    {
+      "epoch": 0.38814405098298566,
+      "grad_norm": 1.8775909686476138,
+      "learning_rate": 2.7988829823479924e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 68589685,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3882642938736247,
+      "grad_norm": 2.2605340733905304,
+      "learning_rate": 2.7981687927443976e-06,
+      "loss": 0.6408,
+      "num_input_tokens_seen": 68606205,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3883845367642638,
+      "grad_norm": 1.9181067704478187,
+      "learning_rate": 2.797454482059231e-06,
+      "loss": 0.852,
+      "num_input_tokens_seen": 68626080,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3885047796549029,
+      "grad_norm": 1.647765499404315,
+      "learning_rate": 2.796740050400854e-06,
+      "loss": 0.8394,
+      "num_input_tokens_seen": 68645100,
+      "step": 3231
+    },
+    {
+      "epoch": 0.388625022545542,
+      "grad_norm": 0.8532408791801639,
+      "learning_rate": 2.7960254978776448e-06,
+      "loss": 0.6178,
+      "num_input_tokens_seen": 68706910,
+      "step": 3232
+    },
+    {
+      "epoch": 0.3887452654361811,
+      "grad_norm": 2.435094191181028,
+      "learning_rate": 2.7953108245980006e-06,
+      "loss": 0.8108,
+      "num_input_tokens_seen": 68725145,
+      "step": 3233
+    },
+    {
+      "epoch": 0.38886550832682015,
+      "grad_norm": 1.5641339285657376,
+      "learning_rate": 2.7945960306703365e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 68747850,
+      "step": 3234
+    },
+    {
+      "epoch": 0.38898575121745926,
+      "grad_norm": 1.8871605136948788,
+      "learning_rate": 2.793881116203087e-06,
+      "loss": 0.657,
+      "num_input_tokens_seen": 68767835,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3891059941080984,
+      "grad_norm": 2.7825940514451557,
+      "learning_rate": 2.793166081304702e-06,
+      "loss": 0.8235,
+      "num_input_tokens_seen": 68788050,
+      "step": 3236
+    },
+    {
+      "epoch": 0.38922623699873743,
+      "grad_norm": 2.211229418210362,
+      "learning_rate": 2.7924509260836543e-06,
+      "loss": 0.8199,
+      "num_input_tokens_seen": 68806895,
+      "step": 3237
+    },
+    {
+      "epoch": 0.38934647988937654,
+      "grad_norm": 1.9152738637345454,
+      "learning_rate": 2.7917356506484302e-06,
+      "loss": 0.6831,
+      "num_input_tokens_seen": 68825735,
+      "step": 3238
+    },
+    {
+      "epoch": 0.38946672278001565,
+      "grad_norm": 2.357229773764005,
+      "learning_rate": 2.791020255107538e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 68842825,
+      "step": 3239
+    },
+    {
+      "epoch": 0.3895869656706547,
+      "grad_norm": 1.5063397346567262,
+      "learning_rate": 2.790304739569502e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 68862445,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3897072085612938,
+      "grad_norm": 1.8553786764134097,
+      "learning_rate": 2.789589104142865e-06,
+      "loss": 0.8961,
+      "num_input_tokens_seen": 68879790,
+      "step": 3241
+    },
+    {
+      "epoch": 0.3898274514519329,
+      "grad_norm": 1.6941341482165353,
+      "learning_rate": 2.78887334893619e-06,
+      "loss": 0.7666,
+      "num_input_tokens_seen": 68897925,
+      "step": 3242
+    },
+    {
+      "epoch": 0.389947694342572,
+      "grad_norm": 0.8165173741996621,
+      "learning_rate": 2.788157474058054e-06,
+      "loss": 0.6475,
+      "num_input_tokens_seen": 68959920,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3900679372332111,
+      "grad_norm": 1.9749278910624048,
+      "learning_rate": 2.7874414796170555e-06,
+      "loss": 0.699,
+      "num_input_tokens_seen": 68981130,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3901881801238502,
+      "grad_norm": 4.2546129160662165,
+      "learning_rate": 2.7867253657218113e-06,
+      "loss": 0.833,
+      "num_input_tokens_seen": 68994740,
+      "step": 3245
+    },
+    {
+      "epoch": 0.39030842301448926,
+      "grad_norm": 1.8386320100310074,
+      "learning_rate": 2.7860091324809544e-06,
+      "loss": 0.7296,
+      "num_input_tokens_seen": 69015520,
+      "step": 3246
+    },
+    {
+      "epoch": 0.39042866590512837,
+      "grad_norm": 2.0772288424866585,
+      "learning_rate": 2.7852927800031377e-06,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 69035405,
+      "step": 3247
+    },
+    {
+      "epoch": 0.3905489087957674,
+      "grad_norm": 1.91006723307839,
+      "learning_rate": 2.7845763083970293e-06,
+      "loss": 0.8217,
+      "num_input_tokens_seen": 69055525,
+      "step": 3248
+    },
+    {
+      "epoch": 0.39066915168640653,
+      "grad_norm": 2.617120465654301,
+      "learning_rate": 2.78385971777132e-06,
+      "loss": 0.8192,
+      "num_input_tokens_seen": 69076335,
+      "step": 3249
+    },
+    {
+      "epoch": 0.39078939457704565,
+      "grad_norm": 1.7490050289284655,
+      "learning_rate": 2.7831430082347143e-06,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 69095260,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3909096374676847,
+      "grad_norm": 2.046149020869084,
+      "learning_rate": 2.7824261798959373e-06,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 69113160,
+      "step": 3251
+    },
+    {
+      "epoch": 0.3910298803583238,
+      "grad_norm": 1.8219220612114093,
+      "learning_rate": 2.78170923286373e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 69132480,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3911501232489629,
+      "grad_norm": 2.6743505168232025,
+      "learning_rate": 2.780992167246855e-06,
+      "loss": 0.8292,
+      "num_input_tokens_seen": 69149725,
+      "step": 3253
+    },
+    {
+      "epoch": 0.391270366139602,
+      "grad_norm": 1.0283085145022817,
+      "learning_rate": 2.780274983154088e-06,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 69208345,
+      "step": 3254
+    },
+    {
+      "epoch": 0.3913906090302411,
+      "grad_norm": 2.2577370138408432,
+      "learning_rate": 2.7795576806942268e-06,
+      "loss": 0.8156,
+      "num_input_tokens_seen": 69226870,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3915108519208802,
+      "grad_norm": 0.8175112549171334,
+      "learning_rate": 2.778840259976085e-06,
+      "loss": 0.5768,
+      "num_input_tokens_seen": 69281820,
+      "step": 3256
+    },
+    {
+      "epoch": 0.39163109481151925,
+      "grad_norm": 1.9688295861626535,
+      "learning_rate": 2.7781227211084955e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 69299770,
+      "step": 3257
+    },
+    {
+      "epoch": 0.39175133770215836,
+      "grad_norm": 1.9496064702054825,
+      "learning_rate": 2.7774050642003076e-06,
+      "loss": 0.8803,
+      "num_input_tokens_seen": 69320300,
+      "step": 3258
+    },
+    {
+      "epoch": 0.3918715805927975,
+      "grad_norm": 1.8700093173411128,
+      "learning_rate": 2.7766872893603896e-06,
+      "loss": 0.935,
+      "num_input_tokens_seen": 69339995,
+      "step": 3259
+    },
+    {
+      "epoch": 0.39199182348343653,
+      "grad_norm": 1.8084285975671681,
+      "learning_rate": 2.775969396697627e-06,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 69358220,
+      "step": 3260
+    },
+    {
+      "epoch": 0.39211206637407564,
+      "grad_norm": 1.9947594011564918,
+      "learning_rate": 2.7752513863209242e-06,
+      "loss": 0.8458,
+      "num_input_tokens_seen": 69376520,
+      "step": 3261
+    },
+    {
+      "epoch": 0.39223230926471475,
+      "grad_norm": 1.539650753480743,
+      "learning_rate": 2.7745332583392024e-06,
+      "loss": 0.8267,
+      "num_input_tokens_seen": 69393700,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3923525521553538,
+      "grad_norm": 2.2492313911851465,
+      "learning_rate": 2.7738150128614014e-06,
+      "loss": 0.7913,
+      "num_input_tokens_seen": 69410825,
+      "step": 3263
+    },
+    {
+      "epoch": 0.3924727950459929,
+      "grad_norm": 3.7170439930092716,
+      "learning_rate": 2.773096649996478e-06,
+      "loss": 0.8912,
+      "num_input_tokens_seen": 69427495,
+      "step": 3264
+    },
+    {
+      "epoch": 0.39259303793663197,
+      "grad_norm": 2.998718104645421,
+      "learning_rate": 2.772378169853408e-06,
+      "loss": 0.7961,
+      "num_input_tokens_seen": 69444785,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3927132808272711,
+      "grad_norm": 1.9705067567947931,
+      "learning_rate": 2.771659572541183e-06,
+      "loss": 0.7371,
+      "num_input_tokens_seen": 69462435,
+      "step": 3266
+    },
+    {
+      "epoch": 0.3928335237179102,
+      "grad_norm": 1.9960138044789726,
+      "learning_rate": 2.7709408581688143e-06,
+      "loss": 0.8694,
+      "num_input_tokens_seen": 69482140,
+      "step": 3267
+    },
+    {
+      "epoch": 0.39295376660854925,
+      "grad_norm": 1.579739037093937,
+      "learning_rate": 2.7702220268453307e-06,
+      "loss": 0.8754,
+      "num_input_tokens_seen": 69502220,
+      "step": 3268
+    },
+    {
+      "epoch": 0.39307400949918836,
+      "grad_norm": 1.9979618949720819,
+      "learning_rate": 2.7695030786797785e-06,
+      "loss": 0.8414,
+      "num_input_tokens_seen": 69517835,
+      "step": 3269
+    },
+    {
+      "epoch": 0.39319425238982747,
+      "grad_norm": 2.46552842637992,
+      "learning_rate": 2.76878401378122e-06,
+      "loss": 0.7347,
+      "num_input_tokens_seen": 69535640,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3933144952804665,
+      "grad_norm": 0.8294253344023248,
+      "learning_rate": 2.768064832258739e-06,
+      "loss": 0.6347,
+      "num_input_tokens_seen": 69600235,
+      "step": 3271
+    },
+    {
+      "epoch": 0.39343473817110564,
+      "grad_norm": 1.9541891967771021,
+      "learning_rate": 2.7673455342214334e-06,
+      "loss": 0.8101,
+      "num_input_tokens_seen": 69616945,
+      "step": 3272
+    },
+    {
+      "epoch": 0.39355498106174475,
+      "grad_norm": 2.150477408039174,
+      "learning_rate": 2.7666261197784198e-06,
+      "loss": 0.7571,
+      "num_input_tokens_seen": 69635480,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3936752239523838,
+      "grad_norm": 2.3944691953020665,
+      "learning_rate": 2.7659065890388336e-06,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 69651200,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3937954668430229,
+      "grad_norm": 4.83584702275678,
+      "learning_rate": 2.765186942111827e-06,
+      "loss": 0.8442,
+      "num_input_tokens_seen": 69667530,
+      "step": 3275
+    },
+    {
+      "epoch": 0.393915709733662,
+      "grad_norm": 1.8070150301717234,
+      "learning_rate": 2.764467179106569e-06,
+      "loss": 0.8216,
+      "num_input_tokens_seen": 69687955,
+      "step": 3276
+    },
+    {
+      "epoch": 0.3940359526243011,
+      "grad_norm": 2.325226101712111,
+      "learning_rate": 2.763747300132249e-06,
+      "loss": 0.7577,
+      "num_input_tokens_seen": 69705115,
+      "step": 3277
+    },
+    {
+      "epoch": 0.3941561955149402,
+      "grad_norm": 1.653860604550936,
+      "learning_rate": 2.7630273052980704e-06,
+      "loss": 0.8619,
+      "num_input_tokens_seen": 69725425,
+      "step": 3278
+    },
+    {
+      "epoch": 0.39427643840557924,
+      "grad_norm": 2.6488817239056406,
+      "learning_rate": 2.762307194713256e-06,
+      "loss": 0.6706,
+      "num_input_tokens_seen": 69742175,
+      "step": 3279
+    },
+    {
+      "epoch": 0.39439668129621835,
+      "grad_norm": 1.97649731825759,
+      "learning_rate": 2.7615869684870458e-06,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 69761205,
+      "step": 3280
+    },
+    {
+      "epoch": 0.39451692418685746,
+      "grad_norm": 2.426535129401768,
+      "learning_rate": 2.7608666267286986e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 69781155,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3946371670774965,
+      "grad_norm": 2.3596359702074126,
+      "learning_rate": 2.7601461695474888e-06,
+      "loss": 0.8561,
+      "num_input_tokens_seen": 69797640,
+      "step": 3282
+    },
+    {
+      "epoch": 0.39475740996813563,
+      "grad_norm": 1.5514249441449879,
+      "learning_rate": 2.75942559705271e-06,
+      "loss": 0.7615,
+      "num_input_tokens_seen": 69817095,
+      "step": 3283
+    },
+    {
+      "epoch": 0.39487765285877474,
+      "grad_norm": 2.6687006281010945,
+      "learning_rate": 2.7587049093536713e-06,
+      "loss": 0.8772,
+      "num_input_tokens_seen": 69833145,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3949978957494138,
+      "grad_norm": 1.7031599929084906,
+      "learning_rate": 2.757984106559701e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 69851850,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3951181386400529,
+      "grad_norm": 2.4336544278473333,
+      "learning_rate": 2.757263188780145e-06,
+      "loss": 0.7149,
+      "num_input_tokens_seen": 69873195,
+      "step": 3286
+    },
+    {
+      "epoch": 0.395238381530692,
+      "grad_norm": 1.7044943689675445,
+      "learning_rate": 2.7565421561243654e-06,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 69891080,
+      "step": 3287
+    },
+    {
+      "epoch": 0.3953586244213311,
+      "grad_norm": 2.5191738399306574,
+      "learning_rate": 2.7558210087017413e-06,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 69910735,
+      "step": 3288
+    },
+    {
+      "epoch": 0.3954788673119702,
+      "grad_norm": 2.170256057836073,
+      "learning_rate": 2.7550997466216724e-06,
+      "loss": 0.7386,
+      "num_input_tokens_seen": 69928250,
+      "step": 3289
+    },
+    {
+      "epoch": 0.3955991102026093,
+      "grad_norm": 2.2323220819837166,
+      "learning_rate": 2.7543783699935714e-06,
+      "loss": 0.8041,
+      "num_input_tokens_seen": 69946000,
+      "step": 3290
+    },
+    {
+      "epoch": 0.39571935309324835,
+      "grad_norm": 2.40132013766084,
+      "learning_rate": 2.753656878926872e-06,
+      "loss": 0.8528,
+      "num_input_tokens_seen": 69961600,
+      "step": 3291
+    },
+    {
+      "epoch": 0.39583959598388746,
+      "grad_norm": 1.8374856154153099,
+      "learning_rate": 2.7529352735310226e-06,
+      "loss": 0.737,
+      "num_input_tokens_seen": 69979470,
+      "step": 3292
+    },
+    {
+      "epoch": 0.39595983887452657,
+      "grad_norm": 2.325397445467587,
+      "learning_rate": 2.7522135539154914e-06,
+      "loss": 0.7908,
+      "num_input_tokens_seen": 69997545,
+      "step": 3293
+    },
+    {
+      "epoch": 0.3960800817651656,
+      "grad_norm": 1.131151958426341,
+      "learning_rate": 2.751491720189762e-06,
+      "loss": 0.6817,
+      "num_input_tokens_seen": 70055375,
+      "step": 3294
+    },
+    {
+      "epoch": 0.39620032465580474,
+      "grad_norm": 2.669612614441195,
+      "learning_rate": 2.7507697724633364e-06,
+      "loss": 0.9081,
+      "num_input_tokens_seen": 70071855,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3963205675464438,
+      "grad_norm": 0.8652166978901629,
+      "learning_rate": 2.7500477108457327e-06,
+      "loss": 0.5659,
+      "num_input_tokens_seen": 70123585,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3964408104370829,
+      "grad_norm": 1.8499369610659868,
+      "learning_rate": 2.749325535446488e-06,
+      "loss": 0.7999,
+      "num_input_tokens_seen": 70141115,
+      "step": 3297
+    },
+    {
+      "epoch": 0.396561053327722,
+      "grad_norm": 1.7507039295732016,
+      "learning_rate": 2.7486032463751555e-06,
+      "loss": 0.7539,
+      "num_input_tokens_seen": 70158850,
+      "step": 3298
+    },
+    {
+      "epoch": 0.39668129621836107,
+      "grad_norm": 2.1185347097908513,
+      "learning_rate": 2.7478808437413055e-06,
+      "loss": 0.6919,
+      "num_input_tokens_seen": 70177980,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3968015391090002,
+      "grad_norm": 1.780913118824692,
+      "learning_rate": 2.7471583276545263e-06,
+      "loss": 0.6542,
+      "num_input_tokens_seen": 70198360,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3969217819996393,
+      "grad_norm": 2.0239469972871236,
+      "learning_rate": 2.746435698224423e-06,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 70216080,
+      "step": 3301
+    },
+    {
+      "epoch": 0.39704202489027834,
+      "grad_norm": 0.8377937703109886,
+      "learning_rate": 2.745712955560617e-06,
+      "loss": 0.6431,
+      "num_input_tokens_seen": 70272005,
+      "step": 3302
+    },
+    {
+      "epoch": 0.39716226778091746,
+      "grad_norm": 2.431012901034257,
+      "learning_rate": 2.7449900997727496e-06,
+      "loss": 0.772,
+      "num_input_tokens_seen": 70289835,
+      "step": 3303
+    },
+    {
+      "epoch": 0.39728251067155657,
+      "grad_norm": 1.641958929823671,
+      "learning_rate": 2.7442671309704754e-06,
+      "loss": 0.8325,
+      "num_input_tokens_seen": 70309280,
+      "step": 3304
+    },
+    {
+      "epoch": 0.3974027535621956,
+      "grad_norm": 1.810880890923636,
+      "learning_rate": 2.7435440492634697e-06,
+      "loss": 0.7628,
+      "num_input_tokens_seen": 70328325,
+      "step": 3305
+    },
+    {
+      "epoch": 0.39752299645283473,
+      "grad_norm": 2.36911539646419,
+      "learning_rate": 2.7428208547614223e-06,
+      "loss": 0.6628,
+      "num_input_tokens_seen": 70347540,
+      "step": 3306
+    },
+    {
+      "epoch": 0.39764323934347384,
+      "grad_norm": 2.289460460435136,
+      "learning_rate": 2.742097547574043e-06,
+      "loss": 0.769,
+      "num_input_tokens_seen": 70365485,
+      "step": 3307
+    },
+    {
+      "epoch": 0.3977634822341129,
+      "grad_norm": 17.143966592699762,
+      "learning_rate": 2.741374127811055e-06,
+      "loss": 0.7685,
+      "num_input_tokens_seen": 70383895,
+      "step": 3308
+    },
+    {
+      "epoch": 0.397883725124752,
+      "grad_norm": 2.7658846322547666,
+      "learning_rate": 2.7406505955822016e-06,
+      "loss": 0.6815,
+      "num_input_tokens_seen": 70404640,
+      "step": 3309
+    },
+    {
+      "epoch": 0.39800396801539106,
+      "grad_norm": 2.3642394700964733,
+      "learning_rate": 2.7399269509972415e-06,
+      "loss": 0.6575,
+      "num_input_tokens_seen": 70418515,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3981242109060302,
+      "grad_norm": 2.628236673925707,
+      "learning_rate": 2.7392031941659514e-06,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 70436080,
+      "step": 3311
+    },
+    {
+      "epoch": 0.3982444537966693,
+      "grad_norm": 1.7531765776602481,
+      "learning_rate": 2.7384793251981244e-06,
+      "loss": 0.8506,
+      "num_input_tokens_seen": 70454785,
+      "step": 3312
+    },
+    {
+      "epoch": 0.39836469668730834,
+      "grad_norm": 1.7100082239913483,
+      "learning_rate": 2.737755344203571e-06,
+      "loss": 0.804,
+      "num_input_tokens_seen": 70474455,
+      "step": 3313
+    },
+    {
+      "epoch": 0.39848493957794745,
+      "grad_norm": 1.7263494720033836,
+      "learning_rate": 2.7370312512921177e-06,
+      "loss": 0.7936,
+      "num_input_tokens_seen": 70495955,
+      "step": 3314
+    },
+    {
+      "epoch": 0.39860518246858656,
+      "grad_norm": 2.7479809507713586,
+      "learning_rate": 2.7363070465736106e-06,
+      "loss": 0.76,
+      "num_input_tokens_seen": 70511545,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3987254253592256,
+      "grad_norm": 2.041973070662389,
+      "learning_rate": 2.73558273015791e-06,
+      "loss": 0.8165,
+      "num_input_tokens_seen": 70531095,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3988456682498647,
+      "grad_norm": 2.285318278607479,
+      "learning_rate": 2.734858302154894e-06,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 70552315,
+      "step": 3317
+    },
+    {
+      "epoch": 0.39896591114050384,
+      "grad_norm": 2.1358721970732013,
+      "learning_rate": 2.734133762674457e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 70571625,
+      "step": 3318
+    },
+    {
+      "epoch": 0.3990861540311429,
+      "grad_norm": 2.9805138918900362,
+      "learning_rate": 2.7334091118265124e-06,
+      "loss": 0.7053,
+      "num_input_tokens_seen": 70593240,
+      "step": 3319
+    },
+    {
+      "epoch": 0.399206396921782,
+      "grad_norm": 0.6737915514037378,
+      "learning_rate": 2.732684349720989e-06,
+      "loss": 0.5862,
+      "num_input_tokens_seen": 70660920,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3993266398124211,
+      "grad_norm": 1.9217949778581944,
+      "learning_rate": 2.7319594764678318e-06,
+      "loss": 0.7448,
+      "num_input_tokens_seen": 70682740,
+      "step": 3321
+    },
+    {
+      "epoch": 0.39944688270306017,
+      "grad_norm": 1.7321146337550002,
+      "learning_rate": 2.7312344921770044e-06,
+      "loss": 0.8339,
+      "num_input_tokens_seen": 70704160,
+      "step": 3322
+    },
+    {
+      "epoch": 0.3995671255936993,
+      "grad_norm": 2.2221370787413295,
+      "learning_rate": 2.730509396958486e-06,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 70722705,
+      "step": 3323
+    },
+    {
+      "epoch": 0.3996873684843384,
+      "grad_norm": 1.7741069563217309,
+      "learning_rate": 2.729784190922272e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 70743860,
+      "step": 3324
+    },
+    {
+      "epoch": 0.39980761137497745,
+      "grad_norm": 0.7990992189193481,
+      "learning_rate": 2.729058874178378e-06,
+      "loss": 0.5923,
+      "num_input_tokens_seen": 70814260,
+      "step": 3325
+    },
+    {
+      "epoch": 0.39992785426561656,
+      "grad_norm": 2.0606626556923238,
+      "learning_rate": 2.728333446836831e-06,
+      "loss": 0.6919,
+      "num_input_tokens_seen": 70835260,
+      "step": 3326
+    },
+    {
+      "epoch": 0.4000480971562556,
+      "grad_norm": 2.0781642183073745,
+      "learning_rate": 2.72760790900768e-06,
+      "loss": 0.7261,
+      "num_input_tokens_seen": 70851565,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4001683400468947,
+      "grad_norm": 1.7296033608493382,
+      "learning_rate": 2.726882260800987e-06,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 70870660,
+      "step": 3328
+    },
+    {
+      "epoch": 0.40028858293753383,
+      "grad_norm": 2.2164326382091377,
+      "learning_rate": 2.726156502326834e-06,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 70891680,
+      "step": 3329
+    },
+    {
+      "epoch": 0.4004088258281729,
+      "grad_norm": 0.7209349574618269,
+      "learning_rate": 2.7254306336953165e-06,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 70954480,
+      "step": 3330
+    },
+    {
+      "epoch": 0.400529068718812,
+      "grad_norm": 0.9488210746982167,
+      "learning_rate": 2.7247046550165485e-06,
+      "loss": 0.6233,
+      "num_input_tokens_seen": 71006325,
+      "step": 3331
+    },
+    {
+      "epoch": 0.4006493116094511,
+      "grad_norm": 7.116124344666175,
+      "learning_rate": 2.7239785664006606e-06,
+      "loss": 0.748,
+      "num_input_tokens_seen": 71029585,
+      "step": 3332
+    },
+    {
+      "epoch": 0.40076955450009016,
+      "grad_norm": 1.346597498015084,
+      "learning_rate": 2.7232523679578002e-06,
+      "loss": 0.6645,
+      "num_input_tokens_seen": 71092385,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4008897973907293,
+      "grad_norm": 2.6243993211882013,
+      "learning_rate": 2.7225260597981295e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 71109810,
+      "step": 3334
+    },
+    {
+      "epoch": 0.4010100402813684,
+      "grad_norm": 2.82306580058909,
+      "learning_rate": 2.721799642031831e-06,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 71125700,
+      "step": 3335
+    },
+    {
+      "epoch": 0.40113028317200744,
+      "grad_norm": 1.903367933987269,
+      "learning_rate": 2.7210731147691006e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 71143095,
+      "step": 3336
+    },
+    {
+      "epoch": 0.40125052606264655,
+      "grad_norm": 1.7018305039058201,
+      "learning_rate": 2.720346478120152e-06,
+      "loss": 0.7489,
+      "num_input_tokens_seen": 71162130,
+      "step": 3337
+    },
+    {
+      "epoch": 0.40137076895328566,
+      "grad_norm": 3.3172086444882303,
+      "learning_rate": 2.719619732195215e-06,
+      "loss": 0.7724,
+      "num_input_tokens_seen": 71183490,
+      "step": 3338
+    },
+    {
+      "epoch": 0.4014910118439247,
+      "grad_norm": 1.3452891425888045,
+      "learning_rate": 2.7188928771045377e-06,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 71204530,
+      "step": 3339
+    },
+    {
+      "epoch": 0.4016112547345638,
+      "grad_norm": 2.1125488827217875,
+      "learning_rate": 2.7181659129583815e-06,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 71223840,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4017314976252029,
+      "grad_norm": 2.167538896620741,
+      "learning_rate": 2.717438839867028e-06,
+      "loss": 0.771,
+      "num_input_tokens_seen": 71242740,
+      "step": 3341
+    },
+    {
+      "epoch": 0.401851740515842,
+      "grad_norm": 1.8810478843528182,
+      "learning_rate": 2.716711657940772e-06,
+      "loss": 0.9098,
+      "num_input_tokens_seen": 71263470,
+      "step": 3342
+    },
+    {
+      "epoch": 0.4019719834064811,
+      "grad_norm": 0.8605382082240843,
+      "learning_rate": 2.7159843672899284e-06,
+      "loss": 0.6014,
+      "num_input_tokens_seen": 71327390,
+      "step": 3343
+    },
+    {
+      "epoch": 0.40209222629712016,
+      "grad_norm": 2.427243314264212,
+      "learning_rate": 2.715256968024825e-06,
+      "loss": 0.8108,
+      "num_input_tokens_seen": 71344185,
+      "step": 3344
+    },
+    {
+      "epoch": 0.40221246918775927,
+      "grad_norm": 1.5906958118657648,
+      "learning_rate": 2.714529460255809e-06,
+      "loss": 0.8188,
+      "num_input_tokens_seen": 71364615,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4023327120783984,
+      "grad_norm": 2.1269068837307636,
+      "learning_rate": 2.713801844093241e-06,
+      "loss": 0.699,
+      "num_input_tokens_seen": 71385485,
+      "step": 3346
+    },
+    {
+      "epoch": 0.40245295496903744,
+      "grad_norm": 2.46450221200613,
+      "learning_rate": 2.7130741196475014e-06,
+      "loss": 0.8828,
+      "num_input_tokens_seen": 71403335,
+      "step": 3347
+    },
+    {
+      "epoch": 0.40257319785967655,
+      "grad_norm": 2.135534097827691,
+      "learning_rate": 2.7123462870289843e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 71423105,
+      "step": 3348
+    },
+    {
+      "epoch": 0.40269344075031566,
+      "grad_norm": 2.2663431012150768,
+      "learning_rate": 2.711618346348102e-06,
+      "loss": 0.802,
+      "num_input_tokens_seen": 71443350,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4028136836409547,
+      "grad_norm": 1.794293838293988,
+      "learning_rate": 2.7108902977152825e-06,
+      "loss": 0.6223,
+      "num_input_tokens_seen": 71460970,
+      "step": 3350
+    },
+    {
+      "epoch": 0.4029339265315938,
+      "grad_norm": 2.041535216498876,
+      "learning_rate": 2.7101621412409704e-06,
+      "loss": 0.7408,
+      "num_input_tokens_seen": 71480175,
+      "step": 3351
+    },
+    {
+      "epoch": 0.40305416942223293,
+      "grad_norm": 1.944953994684264,
+      "learning_rate": 2.7094338770356256e-06,
+      "loss": 0.859,
+      "num_input_tokens_seen": 71498980,
+      "step": 3352
+    },
+    {
+      "epoch": 0.403174412312872,
+      "grad_norm": 3.004637121133846,
+      "learning_rate": 2.708705505209726e-06,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 71519475,
+      "step": 3353
+    },
+    {
+      "epoch": 0.4032946552035111,
+      "grad_norm": 11.794291245593165,
+      "learning_rate": 2.7079770258737646e-06,
+      "loss": 0.9034,
+      "num_input_tokens_seen": 71537105,
+      "step": 3354
+    },
+    {
+      "epoch": 0.4034148980941502,
+      "grad_norm": 2.0043669583208135,
+      "learning_rate": 2.707248439138251e-06,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 71553060,
+      "step": 3355
+    },
+    {
+      "epoch": 0.40353514098478926,
+      "grad_norm": 1.9080100617484956,
+      "learning_rate": 2.7065197451137114e-06,
+      "loss": 0.6505,
+      "num_input_tokens_seen": 71574160,
+      "step": 3356
+    },
+    {
+      "epoch": 0.4036553838754284,
+      "grad_norm": 2.701141185746899,
+      "learning_rate": 2.7057909439106894e-06,
+      "loss": 0.6768,
+      "num_input_tokens_seen": 71591735,
+      "step": 3357
+    },
+    {
+      "epoch": 0.40377562676606743,
+      "grad_norm": 2.579293453042113,
+      "learning_rate": 2.7050620356397413e-06,
+      "loss": 0.774,
+      "num_input_tokens_seen": 71610405,
+      "step": 3358
+    },
+    {
+      "epoch": 0.40389586965670654,
+      "grad_norm": 1.7218811201584583,
+      "learning_rate": 2.7043330204114437e-06,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 71628835,
+      "step": 3359
+    },
+    {
+      "epoch": 0.40401611254734565,
+      "grad_norm": 1.9132685637761047,
+      "learning_rate": 2.7036038983363862e-06,
+      "loss": 0.8516,
+      "num_input_tokens_seen": 71645160,
+      "step": 3360
+    },
+    {
+      "epoch": 0.4041363554379847,
+      "grad_norm": 1.5614400481429616,
+      "learning_rate": 2.702874669525177e-06,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 71663360,
+      "step": 3361
+    },
+    {
+      "epoch": 0.4042565983286238,
+      "grad_norm": 1.9208657977896524,
+      "learning_rate": 2.7021453340884394e-06,
+      "loss": 0.6895,
+      "num_input_tokens_seen": 71680805,
+      "step": 3362
+    },
+    {
+      "epoch": 0.40437684121926293,
+      "grad_norm": 2.3903682709742213,
+      "learning_rate": 2.7014158921368125e-06,
+      "loss": 0.728,
+      "num_input_tokens_seen": 71698850,
+      "step": 3363
+    },
+    {
+      "epoch": 0.404497084109902,
+      "grad_norm": 1.9227937481201445,
+      "learning_rate": 2.7006863437809525e-06,
+      "loss": 0.849,
+      "num_input_tokens_seen": 71718440,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4046173270005411,
+      "grad_norm": 1.7791675423717466,
+      "learning_rate": 2.699956689131532e-06,
+      "loss": 0.8822,
+      "num_input_tokens_seen": 71738145,
+      "step": 3365
+    },
+    {
+      "epoch": 0.4047375698911802,
+      "grad_norm": 2.1601317625976537,
+      "learning_rate": 2.699226928299238e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 71755885,
+      "step": 3366
+    },
+    {
+      "epoch": 0.40485781278181926,
+      "grad_norm": 2.227614434519331,
+      "learning_rate": 2.698497061394775e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 71774090,
+      "step": 3367
+    },
+    {
+      "epoch": 0.40497805567245837,
+      "grad_norm": 2.4416102457337,
+      "learning_rate": 2.6977670885288627e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 71795210,
+      "step": 3368
+    },
+    {
+      "epoch": 0.4050982985630975,
+      "grad_norm": 2.1329745140035494,
+      "learning_rate": 2.6970370098122378e-06,
+      "loss": 0.7473,
+      "num_input_tokens_seen": 71811915,
+      "step": 3369
+    },
+    {
+      "epoch": 0.40521854145373654,
+      "grad_norm": 1.5300164233140587,
+      "learning_rate": 2.696306825355653e-06,
+      "loss": 0.8537,
+      "num_input_tokens_seen": 71833020,
+      "step": 3370
+    },
+    {
+      "epoch": 0.40533878434437565,
+      "grad_norm": 2.547414563651443,
+      "learning_rate": 2.6955765352698763e-06,
+      "loss": 0.8463,
+      "num_input_tokens_seen": 71852885,
+      "step": 3371
+    },
+    {
+      "epoch": 0.40545902723501476,
+      "grad_norm": 2.7094369070598305,
+      "learning_rate": 2.6948461396656923e-06,
+      "loss": 0.7298,
+      "num_input_tokens_seen": 71870015,
+      "step": 3372
+    },
+    {
+      "epoch": 0.4055792701256538,
+      "grad_norm": 4.592481967398067,
+      "learning_rate": 2.6941156386539013e-06,
+      "loss": 0.7476,
+      "num_input_tokens_seen": 71889685,
+      "step": 3373
+    },
+    {
+      "epoch": 0.4056995130162929,
+      "grad_norm": 3.3914769032649903,
+      "learning_rate": 2.6933850323453203e-06,
+      "loss": 0.8099,
+      "num_input_tokens_seen": 71907850,
+      "step": 3374
+    },
+    {
+      "epoch": 0.405819755906932,
+      "grad_norm": 1.7676202574110453,
+      "learning_rate": 2.6926543208507806e-06,
+      "loss": 0.7428,
+      "num_input_tokens_seen": 71926250,
+      "step": 3375
+    },
+    {
+      "epoch": 0.4059399987975711,
+      "grad_norm": 5.777958806098189,
+      "learning_rate": 2.6919235042811316e-06,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 71944755,
+      "step": 3376
+    },
+    {
+      "epoch": 0.4060602416882102,
+      "grad_norm": 5.030816000614874,
+      "learning_rate": 2.691192582747237e-06,
+      "loss": 0.7481,
+      "num_input_tokens_seen": 71964105,
+      "step": 3377
+    },
+    {
+      "epoch": 0.40618048457884925,
+      "grad_norm": 1.8463383960981143,
+      "learning_rate": 2.6904615563599765e-06,
+      "loss": 0.7271,
+      "num_input_tokens_seen": 71983625,
+      "step": 3378
+    },
+    {
+      "epoch": 0.40630072746948837,
+      "grad_norm": 6.913559298439623,
+      "learning_rate": 2.6897304252302477e-06,
+      "loss": 0.8278,
+      "num_input_tokens_seen": 72000665,
+      "step": 3379
+    },
+    {
+      "epoch": 0.4064209703601275,
+      "grad_norm": 0.8113896624976898,
+      "learning_rate": 2.6889991894689614e-06,
+      "loss": 0.5665,
+      "num_input_tokens_seen": 72056815,
+      "step": 3380
+    },
+    {
+      "epoch": 0.40654121325076653,
+      "grad_norm": 2.5257474518195138,
+      "learning_rate": 2.6882678491870464e-06,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 72076970,
+      "step": 3381
+    },
+    {
+      "epoch": 0.40666145614140564,
+      "grad_norm": 1.591360558399751,
+      "learning_rate": 2.6875364044954453e-06,
+      "loss": 0.7064,
+      "num_input_tokens_seen": 72096920,
+      "step": 3382
+    },
+    {
+      "epoch": 0.40678169903204475,
+      "grad_norm": 1.6875527925298854,
+      "learning_rate": 2.6868048555051185e-06,
+      "loss": 0.8075,
+      "num_input_tokens_seen": 72118170,
+      "step": 3383
+    },
+    {
+      "epoch": 0.4069019419226838,
+      "grad_norm": 3.0730493759885644,
+      "learning_rate": 2.686073202327041e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 72136890,
+      "step": 3384
+    },
+    {
+      "epoch": 0.4070221848133229,
+      "grad_norm": 1.5875785528798778,
+      "learning_rate": 2.6853414450722043e-06,
+      "loss": 0.7334,
+      "num_input_tokens_seen": 72156275,
+      "step": 3385
+    },
+    {
+      "epoch": 0.40714242770396203,
+      "grad_norm": 1.8434888944843437,
+      "learning_rate": 2.684609583851615e-06,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 72174170,
+      "step": 3386
+    },
+    {
+      "epoch": 0.4072626705946011,
+      "grad_norm": 1.5530172579210408,
+      "learning_rate": 2.683877618776297e-06,
+      "loss": 0.7983,
+      "num_input_tokens_seen": 72196145,
+      "step": 3387
+    },
+    {
+      "epoch": 0.4073829134852402,
+      "grad_norm": 2.4212531539117,
+      "learning_rate": 2.6831455499572876e-06,
+      "loss": 0.7365,
+      "num_input_tokens_seen": 72213800,
+      "step": 3388
+    },
+    {
+      "epoch": 0.40750315637587925,
+      "grad_norm": 2.8054542417657453,
+      "learning_rate": 2.6824133775056415e-06,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 72232325,
+      "step": 3389
+    },
+    {
+      "epoch": 0.40762339926651836,
+      "grad_norm": 1.8316823881615452,
+      "learning_rate": 2.6816811015324284e-06,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 72250095,
+      "step": 3390
+    },
+    {
+      "epoch": 0.40774364215715747,
+      "grad_norm": 0.8035028203670925,
+      "learning_rate": 2.6809487221487343e-06,
+      "loss": 0.6183,
+      "num_input_tokens_seen": 72309300,
+      "step": 3391
+    },
+    {
+      "epoch": 0.4078638850477965,
+      "grad_norm": 3.1388870643945537,
+      "learning_rate": 2.68021623946566e-06,
+      "loss": 0.8218,
+      "num_input_tokens_seen": 72325730,
+      "step": 3392
+    },
+    {
+      "epoch": 0.40798412793843564,
+      "grad_norm": 1.7038207540118961,
+      "learning_rate": 2.679483653594324e-06,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 72347220,
+      "step": 3393
+    },
+    {
+      "epoch": 0.40810437082907475,
+      "grad_norm": 2.3126524594406415,
+      "learning_rate": 2.678750964645857e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 72366020,
+      "step": 3394
+    },
+    {
+      "epoch": 0.4082246137197138,
+      "grad_norm": 3.0687312899024475,
+      "learning_rate": 2.6780181727314094e-06,
+      "loss": 0.8448,
+      "num_input_tokens_seen": 72380645,
+      "step": 3395
+    },
+    {
+      "epoch": 0.4083448566103529,
+      "grad_norm": 1.7945888033474224,
+      "learning_rate": 2.6772852779621435e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 72398225,
+      "step": 3396
+    },
+    {
+      "epoch": 0.408465099500992,
+      "grad_norm": 2.557111457315078,
+      "learning_rate": 2.676552280449239e-06,
+      "loss": 0.8435,
+      "num_input_tokens_seen": 72417830,
+      "step": 3397
+    },
+    {
+      "epoch": 0.4085853423916311,
+      "grad_norm": 2.3227998784196187,
+      "learning_rate": 2.6758191803038917e-06,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 72436045,
+      "step": 3398
+    },
+    {
+      "epoch": 0.4087055852822702,
+      "grad_norm": 1.7948690312614335,
+      "learning_rate": 2.6750859776373125e-06,
+      "loss": 0.8255,
+      "num_input_tokens_seen": 72455220,
+      "step": 3399
+    },
+    {
+      "epoch": 0.4088258281729093,
+      "grad_norm": 1.1402348242436562,
+      "learning_rate": 2.674352672560727e-06,
+      "loss": 0.6191,
+      "num_input_tokens_seen": 72516385,
+      "step": 3400
+    },
+    {
+      "epoch": 0.40894607106354836,
+      "grad_norm": 1.5677046797627825,
+      "learning_rate": 2.673619265185377e-06,
+      "loss": 0.765,
+      "num_input_tokens_seen": 72535945,
+      "step": 3401
+    },
+    {
+      "epoch": 0.40906631395418747,
+      "grad_norm": 1.7747198813534124,
+      "learning_rate": 2.672885755622521e-06,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 72558080,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4091865568448266,
+      "grad_norm": 2.29491756702475,
+      "learning_rate": 2.67215214398343e-06,
+      "loss": 0.6968,
+      "num_input_tokens_seen": 72577815,
+      "step": 3403
+    },
+    {
+      "epoch": 0.40930679973546563,
+      "grad_norm": 2.0762255367772293,
+      "learning_rate": 2.671418430379393e-06,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 72596220,
+      "step": 3404
+    },
+    {
+      "epoch": 0.40942704262610474,
+      "grad_norm": 2.5708911522977465,
+      "learning_rate": 2.670684614921715e-06,
+      "loss": 0.822,
+      "num_input_tokens_seen": 72614915,
+      "step": 3405
+    },
+    {
+      "epoch": 0.4095472855167438,
+      "grad_norm": 2.38551915398837,
+      "learning_rate": 2.6699506977217128e-06,
+      "loss": 0.6906,
+      "num_input_tokens_seen": 72634810,
+      "step": 3406
+    },
+    {
+      "epoch": 0.4096675284073829,
+      "grad_norm": 2.428873090592351,
+      "learning_rate": 2.6692166788907233e-06,
+      "loss": 0.6914,
+      "num_input_tokens_seen": 72654725,
+      "step": 3407
+    },
+    {
+      "epoch": 0.409787771298022,
+      "grad_norm": 1.8644545769128333,
+      "learning_rate": 2.668482558540095e-06,
+      "loss": 0.7601,
+      "num_input_tokens_seen": 72673390,
+      "step": 3408
+    },
+    {
+      "epoch": 0.4099080141886611,
+      "grad_norm": 0.9027901520175735,
+      "learning_rate": 2.6677483367811947e-06,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 72733150,
+      "step": 3409
+    },
+    {
+      "epoch": 0.4100282570793002,
+      "grad_norm": 1.717554401635684,
+      "learning_rate": 2.6670140137254028e-06,
+      "loss": 0.7495,
+      "num_input_tokens_seen": 72752345,
+      "step": 3410
+    },
+    {
+      "epoch": 0.4101484999699393,
+      "grad_norm": 2.3632587715431055,
+      "learning_rate": 2.666279589484115e-06,
+      "loss": 0.8718,
+      "num_input_tokens_seen": 72769965,
+      "step": 3411
+    },
+    {
+      "epoch": 0.41026874286057835,
+      "grad_norm": 1.9324442147914254,
+      "learning_rate": 2.6655450641687435e-06,
+      "loss": 0.8043,
+      "num_input_tokens_seen": 72787250,
+      "step": 3412
+    },
+    {
+      "epoch": 0.41038898575121746,
+      "grad_norm": 1.766564394940495,
+      "learning_rate": 2.664810437890715e-06,
+      "loss": 0.6841,
+      "num_input_tokens_seen": 72808640,
+      "step": 3413
+    },
+    {
+      "epoch": 0.41050922864185657,
+      "grad_norm": 1.9560068747597648,
+      "learning_rate": 2.664075710761471e-06,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 72826455,
+      "step": 3414
+    },
+    {
+      "epoch": 0.4106294715324956,
+      "grad_norm": 2.5234970454556436,
+      "learning_rate": 2.6633408828924697e-06,
+      "loss": 0.7017,
+      "num_input_tokens_seen": 72845040,
+      "step": 3415
+    },
+    {
+      "epoch": 0.41074971442313474,
+      "grad_norm": 1.6584124478998463,
+      "learning_rate": 2.6626059543951844e-06,
+      "loss": 0.6936,
+      "num_input_tokens_seen": 72864720,
+      "step": 3416
+    },
+    {
+      "epoch": 0.41086995731377385,
+      "grad_norm": 1.6685018982143598,
+      "learning_rate": 2.6618709253811027e-06,
+      "loss": 0.8305,
+      "num_input_tokens_seen": 72883895,
+      "step": 3417
+    },
+    {
+      "epoch": 0.4109902002044129,
+      "grad_norm": 1.5649469194109542,
+      "learning_rate": 2.6611357959617277e-06,
+      "loss": 0.8692,
+      "num_input_tokens_seen": 72903235,
+      "step": 3418
+    },
+    {
+      "epoch": 0.411110443095052,
+      "grad_norm": 2.0446665323212616,
+      "learning_rate": 2.660400566248578e-06,
+      "loss": 0.9001,
+      "num_input_tokens_seen": 72921080,
+      "step": 3419
+    },
+    {
+      "epoch": 0.41123068598569107,
+      "grad_norm": 2.700654350729606,
+      "learning_rate": 2.6596652363531876e-06,
+      "loss": 0.6686,
+      "num_input_tokens_seen": 72936675,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4113509288763302,
+      "grad_norm": 1.7712489812446686,
+      "learning_rate": 2.6589298063871055e-06,
+      "loss": 0.7783,
+      "num_input_tokens_seen": 72956570,
+      "step": 3421
+    },
+    {
+      "epoch": 0.4114711717669693,
+      "grad_norm": 2.1367957345825053,
+      "learning_rate": 2.658194276461895e-06,
+      "loss": 0.6995,
+      "num_input_tokens_seen": 72974215,
+      "step": 3422
+    },
+    {
+      "epoch": 0.41159141465760835,
+      "grad_norm": 2.1442624824450736,
+      "learning_rate": 2.6574586466891368e-06,
+      "loss": 0.6713,
+      "num_input_tokens_seen": 72994410,
+      "step": 3423
+    },
+    {
+      "epoch": 0.41171165754824746,
+      "grad_norm": 2.008027261075277,
+      "learning_rate": 2.6567229171804247e-06,
+      "loss": 0.6416,
+      "num_input_tokens_seen": 73012015,
+      "step": 3424
+    },
+    {
+      "epoch": 0.41183190043888657,
+      "grad_norm": 2.374680694092268,
+      "learning_rate": 2.655987088047368e-06,
+      "loss": 0.8729,
+      "num_input_tokens_seen": 73030080,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4119521433295256,
+      "grad_norm": 3.753176033590248,
+      "learning_rate": 2.6552511594015912e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 73050190,
+      "step": 3426
+    },
+    {
+      "epoch": 0.41207238622016473,
+      "grad_norm": 2.716260329618953,
+      "learning_rate": 2.654515131354735e-06,
+      "loss": 0.8486,
+      "num_input_tokens_seen": 73068175,
+      "step": 3427
+    },
+    {
+      "epoch": 0.41219262911080384,
+      "grad_norm": 1.9204232488094897,
+      "learning_rate": 2.653779004018453e-06,
+      "loss": 0.8492,
+      "num_input_tokens_seen": 73088460,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4123128720014429,
+      "grad_norm": 3.0030937122011734,
+      "learning_rate": 2.653042777504417e-06,
+      "loss": 0.8115,
+      "num_input_tokens_seen": 73110770,
+      "step": 3429
+    },
+    {
+      "epoch": 0.412433114892082,
+      "grad_norm": 1.8567094102953141,
+      "learning_rate": 2.65230645192431e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 73130060,
+      "step": 3430
+    },
+    {
+      "epoch": 0.4125533577827211,
+      "grad_norm": 2.7712165325077045,
+      "learning_rate": 2.6515700273898333e-06,
+      "loss": 0.7851,
+      "num_input_tokens_seen": 73147655,
+      "step": 3431
+    },
+    {
+      "epoch": 0.4126736006733602,
+      "grad_norm": 2.1073799195345386,
+      "learning_rate": 2.6508335040127018e-06,
+      "loss": 0.68,
+      "num_input_tokens_seen": 73167070,
+      "step": 3432
+    },
+    {
+      "epoch": 0.4127938435639993,
+      "grad_norm": 2.511661563899412,
+      "learning_rate": 2.650096881904645e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 73187090,
+      "step": 3433
+    },
+    {
+      "epoch": 0.4129140864546384,
+      "grad_norm": 2.7349122445133958,
+      "learning_rate": 2.649360161177408e-06,
+      "loss": 0.5957,
+      "num_input_tokens_seen": 73201870,
+      "step": 3434
+    },
+    {
+      "epoch": 0.41303432934527745,
+      "grad_norm": 2.449098280609841,
+      "learning_rate": 2.6486233419427504e-06,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 73221405,
+      "step": 3435
+    },
+    {
+      "epoch": 0.41315457223591656,
+      "grad_norm": 2.5354329871752146,
+      "learning_rate": 2.647886424312448e-06,
+      "loss": 0.7536,
+      "num_input_tokens_seen": 73240790,
+      "step": 3436
+    },
+    {
+      "epoch": 0.4132748151265556,
+      "grad_norm": 1.7620370453665812,
+      "learning_rate": 2.6471494083982903e-06,
+      "loss": 0.8434,
+      "num_input_tokens_seen": 73259895,
+      "step": 3437
+    },
+    {
+      "epoch": 0.4133950580171947,
+      "grad_norm": 1.7658132694452524,
+      "learning_rate": 2.6464122943120813e-06,
+      "loss": 0.7431,
+      "num_input_tokens_seen": 73279840,
+      "step": 3438
+    },
+    {
+      "epoch": 0.41351530090783384,
+      "grad_norm": 3.6806907473492183,
+      "learning_rate": 2.645675082165642e-06,
+      "loss": 0.8197,
+      "num_input_tokens_seen": 73295770,
+      "step": 3439
+    },
+    {
+      "epoch": 0.4136355437984729,
+      "grad_norm": 3.014424186075861,
+      "learning_rate": 2.644937772070806e-06,
+      "loss": 0.7445,
+      "num_input_tokens_seen": 73313935,
+      "step": 3440
+    },
+    {
+      "epoch": 0.413755786689112,
+      "grad_norm": 2.3055111545007203,
+      "learning_rate": 2.6442003641394225e-06,
+      "loss": 0.8297,
+      "num_input_tokens_seen": 73331250,
+      "step": 3441
+    },
+    {
+      "epoch": 0.4138760295797511,
+      "grad_norm": 1.6922175703552575,
+      "learning_rate": 2.643462858483356e-06,
+      "loss": 0.8313,
+      "num_input_tokens_seen": 73351255,
+      "step": 3442
+    },
+    {
+      "epoch": 0.41399627247039017,
+      "grad_norm": 2.240576647271395,
+      "learning_rate": 2.6427252552144856e-06,
+      "loss": 0.7277,
+      "num_input_tokens_seen": 73369625,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4141165153610293,
+      "grad_norm": 1.932192897198705,
+      "learning_rate": 2.6419875544447044e-06,
+      "loss": 0.7492,
+      "num_input_tokens_seen": 73390745,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4142367582516684,
+      "grad_norm": 1.641507784456658,
+      "learning_rate": 2.6412497562859218e-06,
+      "loss": 0.717,
+      "num_input_tokens_seen": 73411745,
+      "step": 3445
+    },
+    {
+      "epoch": 0.41435700114230745,
+      "grad_norm": 2.2158735478095166,
+      "learning_rate": 2.6405118608500617e-06,
+      "loss": 0.7503,
+      "num_input_tokens_seen": 73430290,
+      "step": 3446
+    },
+    {
+      "epoch": 0.41447724403294656,
+      "grad_norm": 1.8071482376672585,
+      "learning_rate": 2.6397738682490613e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 73450910,
+      "step": 3447
+    },
+    {
+      "epoch": 0.41459748692358567,
+      "grad_norm": 1.6503833926226124,
+      "learning_rate": 2.6390357785948734e-06,
+      "loss": 0.7475,
+      "num_input_tokens_seen": 73467745,
+      "step": 3448
+    },
+    {
+      "epoch": 0.4147177298142247,
+      "grad_norm": 1.8316376925084643,
+      "learning_rate": 2.6382975919994667e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 73488040,
+      "step": 3449
+    },
+    {
+      "epoch": 0.41483797270486383,
+      "grad_norm": 1.783113943258856,
+      "learning_rate": 2.637559308574822e-06,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 73507505,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4149582155955029,
+      "grad_norm": 1.9927709868414059,
+      "learning_rate": 2.6368209284329376e-06,
+      "loss": 0.7173,
+      "num_input_tokens_seen": 73527855,
+      "step": 3451
+    },
+    {
+      "epoch": 0.415078458486142,
+      "grad_norm": 1.9887636836649296,
+      "learning_rate": 2.6360824516858244e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 73545775,
+      "step": 3452
+    },
+    {
+      "epoch": 0.4151987013767811,
+      "grad_norm": 1.6567403783392922,
+      "learning_rate": 2.635343878445509e-06,
+      "loss": 0.8546,
+      "num_input_tokens_seen": 73568780,
+      "step": 3453
+    },
+    {
+      "epoch": 0.41531894426742016,
+      "grad_norm": 2.6656749830158812,
+      "learning_rate": 2.6346052088240326e-06,
+      "loss": 0.702,
+      "num_input_tokens_seen": 73588020,
+      "step": 3454
+    },
+    {
+      "epoch": 0.4154391871580593,
+      "grad_norm": 3.8473370892630014,
+      "learning_rate": 2.633866442933451e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 73604085,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4155594300486984,
+      "grad_norm": 2.2345336269226723,
+      "learning_rate": 2.633127580885833e-06,
+      "loss": 0.8311,
+      "num_input_tokens_seen": 73618305,
+      "step": 3456
+    },
+    {
+      "epoch": 0.41567967293933744,
+      "grad_norm": 2.037844493782449,
+      "learning_rate": 2.632388622793265e-06,
+      "loss": 0.6485,
+      "num_input_tokens_seen": 73637180,
+      "step": 3457
+    },
+    {
+      "epoch": 0.41579991582997655,
+      "grad_norm": 1.850081695004342,
+      "learning_rate": 2.6316495687678457e-06,
+      "loss": 0.6773,
+      "num_input_tokens_seen": 73655550,
+      "step": 3458
+    },
+    {
+      "epoch": 0.41592015872061566,
+      "grad_norm": 2.4370931376098244,
+      "learning_rate": 2.6309104189216887e-06,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 73672835,
+      "step": 3459
+    },
+    {
+      "epoch": 0.4160404016112547,
+      "grad_norm": 2.549368032416684,
+      "learning_rate": 2.6301711733669226e-06,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 73688355,
+      "step": 3460
+    },
+    {
+      "epoch": 0.41616064450189383,
+      "grad_norm": 2.72231647067919,
+      "learning_rate": 2.629431832215691e-06,
+      "loss": 0.7455,
+      "num_input_tokens_seen": 73702880,
+      "step": 3461
+    },
+    {
+      "epoch": 0.41628088739253294,
+      "grad_norm": 4.329501779951195,
+      "learning_rate": 2.628692395580151e-06,
+      "loss": 0.8692,
+      "num_input_tokens_seen": 73722690,
+      "step": 3462
+    },
+    {
+      "epoch": 0.416401130283172,
+      "grad_norm": 1.782933477662036,
+      "learning_rate": 2.6279528635724747e-06,
+      "loss": 0.7911,
+      "num_input_tokens_seen": 73742565,
+      "step": 3463
+    },
+    {
+      "epoch": 0.4165213731738111,
+      "grad_norm": 2.789999562999678,
+      "learning_rate": 2.6272132363048478e-06,
+      "loss": 0.7839,
+      "num_input_tokens_seen": 73759085,
+      "step": 3464
+    },
+    {
+      "epoch": 0.4166416160644502,
+      "grad_norm": 2.2255415727290995,
+      "learning_rate": 2.626473513889472e-06,
+      "loss": 0.6949,
+      "num_input_tokens_seen": 73781185,
+      "step": 3465
+    },
+    {
+      "epoch": 0.41676185895508927,
+      "grad_norm": 1.9701508551618336,
+      "learning_rate": 2.625733696438562e-06,
+      "loss": 0.8217,
+      "num_input_tokens_seen": 73798410,
+      "step": 3466
+    },
+    {
+      "epoch": 0.4168821018457284,
+      "grad_norm": 1.7359499319730178,
+      "learning_rate": 2.6249937840643476e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 73816435,
+      "step": 3467
+    },
+    {
+      "epoch": 0.41700234473636744,
+      "grad_norm": 2.463008962257231,
+      "learning_rate": 2.6242537768790733e-06,
+      "loss": 0.6718,
+      "num_input_tokens_seen": 73835310,
+      "step": 3468
+    },
+    {
+      "epoch": 0.41712258762700655,
+      "grad_norm": 2.094473969489827,
+      "learning_rate": 2.6235136749949975e-06,
+      "loss": 0.6861,
+      "num_input_tokens_seen": 73858480,
+      "step": 3469
+    },
+    {
+      "epoch": 0.41724283051764566,
+      "grad_norm": 2.366165033959666,
+      "learning_rate": 2.6227734785243924e-06,
+      "loss": 0.6131,
+      "num_input_tokens_seen": 73878160,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4173630734082847,
+      "grad_norm": 1.96040709455256,
+      "learning_rate": 2.6220331875795466e-06,
+      "loss": 0.7845,
+      "num_input_tokens_seen": 73897230,
+      "step": 3471
+    },
+    {
+      "epoch": 0.4174833162989238,
+      "grad_norm": 1.6332146106739969,
+      "learning_rate": 2.62129280227276e-06,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 73916950,
+      "step": 3472
+    },
+    {
+      "epoch": 0.41760355918956293,
+      "grad_norm": 7.142381751001548,
+      "learning_rate": 2.62055232271635e-06,
+      "loss": 0.6875,
+      "num_input_tokens_seen": 73943855,
+      "step": 3473
+    },
+    {
+      "epoch": 0.417723802080202,
+      "grad_norm": 2.1875974686545967,
+      "learning_rate": 2.619811749022645e-06,
+      "loss": 0.8795,
+      "num_input_tokens_seen": 73958885,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4178440449708411,
+      "grad_norm": 2.6680836482395422,
+      "learning_rate": 2.6190710813039917e-06,
+      "loss": 0.7098,
+      "num_input_tokens_seen": 73971730,
+      "step": 3475
+    },
+    {
+      "epoch": 0.4179642878614802,
+      "grad_norm": 2.7845246937491948,
+      "learning_rate": 2.618330319672747e-06,
+      "loss": 0.8372,
+      "num_input_tokens_seen": 73990870,
+      "step": 3476
+    },
+    {
+      "epoch": 0.41808453075211927,
+      "grad_norm": 1.9065108652369176,
+      "learning_rate": 2.6175894642412846e-06,
+      "loss": 0.9097,
+      "num_input_tokens_seen": 74004990,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4182047736427584,
+      "grad_norm": 2.0716710654423336,
+      "learning_rate": 2.6168485151219914e-06,
+      "loss": 0.7319,
+      "num_input_tokens_seen": 74024330,
+      "step": 3478
+    },
+    {
+      "epoch": 0.4183250165333975,
+      "grad_norm": 3.2746381334993773,
+      "learning_rate": 2.616107472427269e-06,
+      "loss": 0.702,
+      "num_input_tokens_seen": 74038745,
+      "step": 3479
+    },
+    {
+      "epoch": 0.41844525942403654,
+      "grad_norm": 2.4956682460770603,
+      "learning_rate": 2.6153663362695325e-06,
+      "loss": 0.7632,
+      "num_input_tokens_seen": 74052130,
+      "step": 3480
+    },
+    {
+      "epoch": 0.41856550231467565,
+      "grad_norm": 2.686935105171513,
+      "learning_rate": 2.6146251067612126e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 74067325,
+      "step": 3481
+    },
+    {
+      "epoch": 0.41868574520531476,
+      "grad_norm": 1.8929240802347667,
+      "learning_rate": 2.613883784014752e-06,
+      "loss": 0.8097,
+      "num_input_tokens_seen": 74086080,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4188059880959538,
+      "grad_norm": 3.320399801687348,
+      "learning_rate": 2.6131423681426103e-06,
+      "loss": 0.7729,
+      "num_input_tokens_seen": 74101715,
+      "step": 3483
+    },
+    {
+      "epoch": 0.41892623098659293,
+      "grad_norm": 1.738595311398414,
+      "learning_rate": 2.6124008592572587e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 74125420,
+      "step": 3484
+    },
+    {
+      "epoch": 0.419046473877232,
+      "grad_norm": 2.2315833168098234,
+      "learning_rate": 2.6116592574711835e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 74143440,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4191667167678711,
+      "grad_norm": 2.458298281802693,
+      "learning_rate": 2.610917562896885e-06,
+      "loss": 0.8297,
+      "num_input_tokens_seen": 74162925,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4192869596585102,
+      "grad_norm": 10.526570370811093,
+      "learning_rate": 2.610175775646878e-06,
+      "loss": 0.8334,
+      "num_input_tokens_seen": 74181225,
+      "step": 3487
+    },
+    {
+      "epoch": 0.41940720254914926,
+      "grad_norm": 2.23643778482333,
+      "learning_rate": 2.6094338958336907e-06,
+      "loss": 0.7422,
+      "num_input_tokens_seen": 74199615,
+      "step": 3488
+    },
+    {
+      "epoch": 0.41952744543978837,
+      "grad_norm": 2.1117603086178605,
+      "learning_rate": 2.608691923569867e-06,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 74216210,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4196476883304275,
+      "grad_norm": 1.6016034418373954,
+      "learning_rate": 2.6079498589679616e-06,
+      "loss": 0.7524,
+      "num_input_tokens_seen": 74237020,
+      "step": 3490
+    },
+    {
+      "epoch": 0.41976793122106654,
+      "grad_norm": 1.9207048359616539,
+      "learning_rate": 2.6072077021405465e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 74255575,
+      "step": 3491
+    },
+    {
+      "epoch": 0.41988817411170565,
+      "grad_norm": 1.7194715663200137,
+      "learning_rate": 2.6064654532002054e-06,
+      "loss": 0.6904,
+      "num_input_tokens_seen": 74274305,
+      "step": 3492
+    },
+    {
+      "epoch": 0.42000841700234476,
+      "grad_norm": 1.6324013058500164,
+      "learning_rate": 2.6057231122595375e-06,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 74295335,
+      "step": 3493
+    },
+    {
+      "epoch": 0.4201286598929838,
+      "grad_norm": 1.9937085385972362,
+      "learning_rate": 2.604980679431154e-06,
+      "loss": 0.7276,
+      "num_input_tokens_seen": 74313295,
+      "step": 3494
+    },
+    {
+      "epoch": 0.4202489027836229,
+      "grad_norm": 2.1162838096830985,
+      "learning_rate": 2.604238154827684e-06,
+      "loss": 0.7467,
+      "num_input_tokens_seen": 74329640,
+      "step": 3495
+    },
+    {
+      "epoch": 0.42036914567426203,
+      "grad_norm": 3.4738411531268016,
+      "learning_rate": 2.6034955385617652e-06,
+      "loss": 0.7306,
+      "num_input_tokens_seen": 74347690,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4204893885649011,
+      "grad_norm": 1.4938675897660767,
+      "learning_rate": 2.6027528307460536e-06,
+      "loss": 0.6636,
+      "num_input_tokens_seen": 74411415,
+      "step": 3497
+    },
+    {
+      "epoch": 0.4206096314555402,
+      "grad_norm": 2.1061221610536487,
+      "learning_rate": 2.602010031493217e-06,
+      "loss": 0.8542,
+      "num_input_tokens_seen": 74429365,
+      "step": 3498
+    },
+    {
+      "epoch": 0.42072987434617926,
+      "grad_norm": 2.079517926701649,
+      "learning_rate": 2.6012671409159376e-06,
+      "loss": 0.8688,
+      "num_input_tokens_seen": 74450420,
+      "step": 3499
+    },
+    {
+      "epoch": 0.42085011723681837,
+      "grad_norm": 2.0728196568071477,
+      "learning_rate": 2.6005241591269097e-06,
+      "loss": 0.818,
+      "num_input_tokens_seen": 74469510,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4209703601274575,
+      "grad_norm": 1.640490933828694,
+      "learning_rate": 2.5997810862388454e-06,
+      "loss": 0.7981,
+      "num_input_tokens_seen": 74489070,
+      "step": 3501
+    },
+    {
+      "epoch": 0.42109060301809653,
+      "grad_norm": 3.9024981982776588,
+      "learning_rate": 2.5990379223644666e-06,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 74507690,
+      "step": 3502
+    },
+    {
+      "epoch": 0.42121084590873564,
+      "grad_norm": 2.4385864910852524,
+      "learning_rate": 2.5982946676165112e-06,
+      "loss": 0.7502,
+      "num_input_tokens_seen": 74527180,
+      "step": 3503
+    },
+    {
+      "epoch": 0.42133108879937475,
+      "grad_norm": 0.7853580114040855,
+      "learning_rate": 2.597551322107731e-06,
+      "loss": 0.6048,
+      "num_input_tokens_seen": 74590870,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4214513316900138,
+      "grad_norm": 2.2241793963271643,
+      "learning_rate": 2.5968078859508897e-06,
+      "loss": 0.8761,
+      "num_input_tokens_seen": 74607790,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4215715745806529,
+      "grad_norm": 2.8289086951487823,
+      "learning_rate": 2.5960643592587673e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 74624920,
+      "step": 3506
+    },
+    {
+      "epoch": 0.42169181747129203,
+      "grad_norm": 1.9406197368302194,
+      "learning_rate": 2.5953207421441553e-06,
+      "loss": 0.8188,
+      "num_input_tokens_seen": 74643240,
+      "step": 3507
+    },
+    {
+      "epoch": 0.4218120603619311,
+      "grad_norm": 2.6330363717547542,
+      "learning_rate": 2.59457703471986e-06,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 74661115,
+      "step": 3508
+    },
+    {
+      "epoch": 0.4219323032525702,
+      "grad_norm": 1.8702127292555886,
+      "learning_rate": 2.593833237098701e-06,
+      "loss": 0.8265,
+      "num_input_tokens_seen": 74678435,
+      "step": 3509
+    },
+    {
+      "epoch": 0.4220525461432093,
+      "grad_norm": 2.2151419581980987,
+      "learning_rate": 2.593089349393512e-06,
+      "loss": 0.6328,
+      "num_input_tokens_seen": 74698645,
+      "step": 3510
+    },
+    {
+      "epoch": 0.42217278903384836,
+      "grad_norm": 4.290503011914125,
+      "learning_rate": 2.592345371717141e-06,
+      "loss": 0.8388,
+      "num_input_tokens_seen": 74717895,
+      "step": 3511
+    },
+    {
+      "epoch": 0.42229303192448747,
+      "grad_norm": 2.2050791920259982,
+      "learning_rate": 2.591601304182448e-06,
+      "loss": 0.7029,
+      "num_input_tokens_seen": 74735585,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4224132748151266,
+      "grad_norm": 1.6329320103893197,
+      "learning_rate": 2.5908571469023067e-06,
+      "loss": 0.7848,
+      "num_input_tokens_seen": 74754790,
+      "step": 3513
+    },
+    {
+      "epoch": 0.42253351770576564,
+      "grad_norm": 2.2098417963556685,
+      "learning_rate": 2.5901128999896067e-06,
+      "loss": 0.7544,
+      "num_input_tokens_seen": 74769940,
+      "step": 3514
+    },
+    {
+      "epoch": 0.42265376059640475,
+      "grad_norm": 1.6964963258270764,
+      "learning_rate": 2.5893685635572487e-06,
+      "loss": 0.6849,
+      "num_input_tokens_seen": 74790510,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4227740034870438,
+      "grad_norm": 2.0269413478013485,
+      "learning_rate": 2.5886241377181483e-06,
+      "loss": 0.6947,
+      "num_input_tokens_seen": 74809100,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4228942463776829,
+      "grad_norm": 1.7284362468897982,
+      "learning_rate": 2.587879622585234e-06,
+      "loss": 0.8055,
+      "num_input_tokens_seen": 74827420,
+      "step": 3517
+    },
+    {
+      "epoch": 0.423014489268322,
+      "grad_norm": 2.0690724226656383,
+      "learning_rate": 2.587135018271448e-06,
+      "loss": 0.7567,
+      "num_input_tokens_seen": 74848020,
+      "step": 3518
+    },
+    {
+      "epoch": 0.4231347321589611,
+      "grad_norm": 1.9716679129922656,
+      "learning_rate": 2.5863903248897475e-06,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 74863640,
+      "step": 3519
+    },
+    {
+      "epoch": 0.4232549750496002,
+      "grad_norm": 2.415037916956387,
+      "learning_rate": 2.5856455425531003e-06,
+      "loss": 0.6616,
+      "num_input_tokens_seen": 74884835,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4233752179402393,
+      "grad_norm": 1.8311904524340907,
+      "learning_rate": 2.5849006713744906e-06,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 74903350,
+      "step": 3521
+    },
+    {
+      "epoch": 0.42349546083087836,
+      "grad_norm": 2.5168038229864433,
+      "learning_rate": 2.5841557114669135e-06,
+      "loss": 0.7268,
+      "num_input_tokens_seen": 74919930,
+      "step": 3522
+    },
+    {
+      "epoch": 0.42361570372151747,
+      "grad_norm": 15.38933095735968,
+      "learning_rate": 2.58341066294338e-06,
+      "loss": 0.6724,
+      "num_input_tokens_seen": 74936315,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4237359466121566,
+      "grad_norm": 2.567370607508526,
+      "learning_rate": 2.582665525916912e-06,
+      "loss": 0.8557,
+      "num_input_tokens_seen": 74954690,
+      "step": 3524
+    },
+    {
+      "epoch": 0.42385618950279563,
+      "grad_norm": 2.4248226776410298,
+      "learning_rate": 2.5819203005005475e-06,
+      "loss": 0.8986,
+      "num_input_tokens_seen": 74971745,
+      "step": 3525
+    },
+    {
+      "epoch": 0.42397643239343474,
+      "grad_norm": 1.8552888965772123,
+      "learning_rate": 2.5811749868073355e-06,
+      "loss": 0.777,
+      "num_input_tokens_seen": 74991700,
+      "step": 3526
+    },
+    {
+      "epoch": 0.42409667528407385,
+      "grad_norm": 2.2740155805463145,
+      "learning_rate": 2.5804295849503414e-06,
+      "loss": 0.9069,
+      "num_input_tokens_seen": 75007170,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4242169181747129,
+      "grad_norm": 2.3355500094581076,
+      "learning_rate": 2.5796840950426397e-06,
+      "loss": 0.6543,
+      "num_input_tokens_seen": 75023975,
+      "step": 3528
+    },
+    {
+      "epoch": 0.424337161065352,
+      "grad_norm": 1.831363347479086,
+      "learning_rate": 2.578938517197322e-06,
+      "loss": 0.6523,
+      "num_input_tokens_seen": 75041790,
+      "step": 3529
+    },
+    {
+      "epoch": 0.4244574039559911,
+      "grad_norm": 2.5770680635795666,
+      "learning_rate": 2.578192851527491e-06,
+      "loss": 0.6278,
+      "num_input_tokens_seen": 75060230,
+      "step": 3530
+    },
+    {
+      "epoch": 0.4245776468466302,
+      "grad_norm": 3.521474364436042,
+      "learning_rate": 2.577447098146265e-06,
+      "loss": 0.6805,
+      "num_input_tokens_seen": 75077125,
+      "step": 3531
+    },
+    {
+      "epoch": 0.4246978897372693,
+      "grad_norm": 1.6822951602955885,
+      "learning_rate": 2.5767012571667724e-06,
+      "loss": 0.78,
+      "num_input_tokens_seen": 75096325,
+      "step": 3532
+    },
+    {
+      "epoch": 0.42481813262790835,
+      "grad_norm": 2.007936744264856,
+      "learning_rate": 2.5759553287021587e-06,
+      "loss": 0.6795,
+      "num_input_tokens_seen": 75114375,
+      "step": 3533
+    },
+    {
+      "epoch": 0.42493837551854746,
+      "grad_norm": 2.1073572231889295,
+      "learning_rate": 2.5752093128655786e-06,
+      "loss": 0.7664,
+      "num_input_tokens_seen": 75132340,
+      "step": 3534
+    },
+    {
+      "epoch": 0.4250586184091866,
+      "grad_norm": 2.204380204621364,
+      "learning_rate": 2.574463209770204e-06,
+      "loss": 0.73,
+      "num_input_tokens_seen": 75151375,
+      "step": 3535
+    },
+    {
+      "epoch": 0.42517886129982563,
+      "grad_norm": 1.8377186689738227,
+      "learning_rate": 2.5737170195292165e-06,
+      "loss": 0.7852,
+      "num_input_tokens_seen": 75174430,
+      "step": 3536
+    },
+    {
+      "epoch": 0.42529910419046474,
+      "grad_norm": 2.0979206057256055,
+      "learning_rate": 2.572970742255814e-06,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 75192640,
+      "step": 3537
+    },
+    {
+      "epoch": 0.42541934708110385,
+      "grad_norm": 1.8470224395050348,
+      "learning_rate": 2.5722243780632046e-06,
+      "loss": 0.8093,
+      "num_input_tokens_seen": 75210625,
+      "step": 3538
+    },
+    {
+      "epoch": 0.4255395899717429,
+      "grad_norm": 0.8357680552186567,
+      "learning_rate": 2.5714779270646125e-06,
+      "loss": 0.6544,
+      "num_input_tokens_seen": 75271115,
+      "step": 3539
+    },
+    {
+      "epoch": 0.425659832862382,
+      "grad_norm": 2.379488153454117,
+      "learning_rate": 2.570731389373273e-06,
+      "loss": 0.7611,
+      "num_input_tokens_seen": 75289375,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4257800757530211,
+      "grad_norm": 3.1027385366722586,
+      "learning_rate": 2.5699847651024356e-06,
+      "loss": 0.7652,
+      "num_input_tokens_seen": 75309735,
+      "step": 3541
+    },
+    {
+      "epoch": 0.4259003186436602,
+      "grad_norm": 2.398128664671438,
+      "learning_rate": 2.5692380543653627e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 75327610,
+      "step": 3542
+    },
+    {
+      "epoch": 0.4260205615342993,
+      "grad_norm": 2.1937646833162625,
+      "learning_rate": 2.5684912572753298e-06,
+      "loss": 0.695,
+      "num_input_tokens_seen": 75343005,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4261408044249384,
+      "grad_norm": 1.948438725942936,
+      "learning_rate": 2.5677443739456245e-06,
+      "loss": 0.8421,
+      "num_input_tokens_seen": 75364385,
+      "step": 3544
+    },
+    {
+      "epoch": 0.42626104731557746,
+      "grad_norm": 2.6506802196316026,
+      "learning_rate": 2.5669974044895495e-06,
+      "loss": 0.7887,
+      "num_input_tokens_seen": 75380500,
+      "step": 3545
+    },
+    {
+      "epoch": 0.42638129020621657,
+      "grad_norm": 1.7416516251374892,
+      "learning_rate": 2.5662503490204187e-06,
+      "loss": 0.7834,
+      "num_input_tokens_seen": 75400385,
+      "step": 3546
+    },
+    {
+      "epoch": 0.4265015330968556,
+      "grad_norm": 2.1656062333468022,
+      "learning_rate": 2.5655032076515603e-06,
+      "loss": 0.7597,
+      "num_input_tokens_seen": 75419430,
+      "step": 3547
+    },
+    {
+      "epoch": 0.42662177598749473,
+      "grad_norm": 2.5049421971210832,
+      "learning_rate": 2.564755980496315e-06,
+      "loss": 0.8146,
+      "num_input_tokens_seen": 75439080,
+      "step": 3548
+    },
+    {
+      "epoch": 0.42674201887813384,
+      "grad_norm": 1.8973682687438247,
+      "learning_rate": 2.5640086676680372e-06,
+      "loss": 0.7834,
+      "num_input_tokens_seen": 75460295,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4268622617687729,
+      "grad_norm": 2.354202737052597,
+      "learning_rate": 2.5632612692800923e-06,
+      "loss": 0.8059,
+      "num_input_tokens_seen": 75479080,
+      "step": 3550
+    },
+    {
+      "epoch": 0.426982504659412,
+      "grad_norm": 2.6146409794521417,
+      "learning_rate": 2.5625137854458603e-06,
+      "loss": 0.7371,
+      "num_input_tokens_seen": 75497815,
+      "step": 3551
+    },
+    {
+      "epoch": 0.4271027475500511,
+      "grad_norm": 1.9667034197764448,
+      "learning_rate": 2.561766216278735e-06,
+      "loss": 0.799,
+      "num_input_tokens_seen": 75515130,
+      "step": 3552
+    },
+    {
+      "epoch": 0.4272229904406902,
+      "grad_norm": 2.0130897890201247,
+      "learning_rate": 2.561018561892121e-06,
+      "loss": 0.8034,
+      "num_input_tokens_seen": 75533990,
+      "step": 3553
+    },
+    {
+      "epoch": 0.4273432333313293,
+      "grad_norm": 1.6399691209264875,
+      "learning_rate": 2.5602708223994363e-06,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 75555575,
+      "step": 3554
+    },
+    {
+      "epoch": 0.4274634762219684,
+      "grad_norm": 3.0607598747278395,
+      "learning_rate": 2.559522997914115e-06,
+      "loss": 0.6791,
+      "num_input_tokens_seen": 75574875,
+      "step": 3555
+    },
+    {
+      "epoch": 0.42758371911260745,
+      "grad_norm": 2.0648131201784183,
+      "learning_rate": 2.558775088549599e-06,
+      "loss": 0.8433,
+      "num_input_tokens_seen": 75594175,
+      "step": 3556
+    },
+    {
+      "epoch": 0.42770396200324656,
+      "grad_norm": 3.0557185461790772,
+      "learning_rate": 2.5580270944193467e-06,
+      "loss": 0.6676,
+      "num_input_tokens_seen": 75610715,
+      "step": 3557
+    },
+    {
+      "epoch": 0.4278242048938857,
+      "grad_norm": 1.633482738245169,
+      "learning_rate": 2.557279015636827e-06,
+      "loss": 0.5734,
+      "num_input_tokens_seen": 75670845,
+      "step": 3558
+    },
+    {
+      "epoch": 0.42794444778452473,
+      "grad_norm": 1.0701234787689877,
+      "learning_rate": 2.5565308523155245e-06,
+      "loss": 0.656,
+      "num_input_tokens_seen": 75730165,
+      "step": 3559
+    },
+    {
+      "epoch": 0.42806469067516384,
+      "grad_norm": 3.8148951486982634,
+      "learning_rate": 2.5557826045689336e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 75746125,
+      "step": 3560
+    },
+    {
+      "epoch": 0.4281849335658029,
+      "grad_norm": 1.108997071902187,
+      "learning_rate": 2.5550342725105643e-06,
+      "loss": 0.6081,
+      "num_input_tokens_seen": 75804010,
+      "step": 3561
+    },
+    {
+      "epoch": 0.428305176456442,
+      "grad_norm": 2.0069417929721305,
+      "learning_rate": 2.554285856253936e-06,
+      "loss": 0.8151,
+      "num_input_tokens_seen": 75822565,
+      "step": 3562
+    },
+    {
+      "epoch": 0.4284254193470811,
+      "grad_norm": 3.16011663722786,
+      "learning_rate": 2.553537355912585e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 75842650,
+      "step": 3563
+    },
+    {
+      "epoch": 0.42854566223772017,
+      "grad_norm": 1.9063930268992315,
+      "learning_rate": 2.552788771600057e-06,
+      "loss": 0.8142,
+      "num_input_tokens_seen": 75862680,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4286659051283593,
+      "grad_norm": 1.9796588727113358,
+      "learning_rate": 2.552040103429912e-06,
+      "loss": 0.8146,
+      "num_input_tokens_seen": 75880160,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4287861480189984,
+      "grad_norm": 1.9757256652959743,
+      "learning_rate": 2.551291351515722e-06,
+      "loss": 0.8633,
+      "num_input_tokens_seen": 75896895,
+      "step": 3566
+    },
+    {
+      "epoch": 0.42890639090963745,
+      "grad_norm": 1.53470517289562,
+      "learning_rate": 2.5505425159710726e-06,
+      "loss": 0.8569,
+      "num_input_tokens_seen": 75916425,
+      "step": 3567
+    },
+    {
+      "epoch": 0.42902663380027656,
+      "grad_norm": 1.9348507717518264,
+      "learning_rate": 2.5497935969095607e-06,
+      "loss": 0.8331,
+      "num_input_tokens_seen": 75934765,
+      "step": 3568
+    },
+    {
+      "epoch": 0.42914687669091567,
+      "grad_norm": 2.7428743786368486,
+      "learning_rate": 2.5490445944447976e-06,
+      "loss": 0.6609,
+      "num_input_tokens_seen": 75952980,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4292671195815547,
+      "grad_norm": 2.091621906058011,
+      "learning_rate": 2.5482955086904056e-06,
+      "loss": 0.6594,
+      "num_input_tokens_seen": 75973995,
+      "step": 3570
+    },
+    {
+      "epoch": 0.42938736247219383,
+      "grad_norm": 1.7252700140295156,
+      "learning_rate": 2.547546339760022e-06,
+      "loss": 0.7579,
+      "num_input_tokens_seen": 75993795,
+      "step": 3571
+    },
+    {
+      "epoch": 0.42950760536283294,
+      "grad_norm": 2.022439570264024,
+      "learning_rate": 2.546797087767293e-06,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 76013640,
+      "step": 3572
+    },
+    {
+      "epoch": 0.429627848253472,
+      "grad_norm": 1.857559768547678,
+      "learning_rate": 2.546047752825881e-06,
+      "loss": 0.8731,
+      "num_input_tokens_seen": 76033965,
+      "step": 3573
+    },
+    {
+      "epoch": 0.4297480911441111,
+      "grad_norm": 2.280151426834863,
+      "learning_rate": 2.545298335049459e-06,
+      "loss": 0.9338,
+      "num_input_tokens_seen": 76049240,
+      "step": 3574
+    },
+    {
+      "epoch": 0.4298683340347502,
+      "grad_norm": 2.1349635101751367,
+      "learning_rate": 2.544548834551713e-06,
+      "loss": 0.6479,
+      "num_input_tokens_seen": 76067965,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4299885769253893,
+      "grad_norm": 2.3711969082986077,
+      "learning_rate": 2.543799251446342e-06,
+      "loss": 0.9358,
+      "num_input_tokens_seen": 76081010,
+      "step": 3576
+    },
+    {
+      "epoch": 0.4301088198160284,
+      "grad_norm": 1.6524779215895846,
+      "learning_rate": 2.5430495858470565e-06,
+      "loss": 0.8754,
+      "num_input_tokens_seen": 76100200,
+      "step": 3577
+    },
+    {
+      "epoch": 0.43022906270666744,
+      "grad_norm": 3.5605133985727857,
+      "learning_rate": 2.5422998378675815e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 76117865,
+      "step": 3578
+    },
+    {
+      "epoch": 0.43034930559730655,
+      "grad_norm": 2.6345596470293304,
+      "learning_rate": 2.541550007621651e-06,
+      "loss": 0.8396,
+      "num_input_tokens_seen": 76136075,
+      "step": 3579
+    },
+    {
+      "epoch": 0.43046954848794566,
+      "grad_norm": 1.7355202612383454,
+      "learning_rate": 2.5408000952230156e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 76154585,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4305897913785847,
+      "grad_norm": 1.8773640690264586,
+      "learning_rate": 2.5400501007854357e-06,
+      "loss": 0.8958,
+      "num_input_tokens_seen": 76173750,
+      "step": 3581
+    },
+    {
+      "epoch": 0.43071003426922383,
+      "grad_norm": 1.821732752755133,
+      "learning_rate": 2.539300024422685e-06,
+      "loss": 0.7548,
+      "num_input_tokens_seen": 76191415,
+      "step": 3582
+    },
+    {
+      "epoch": 0.43083027715986294,
+      "grad_norm": 0.8273045872785468,
+      "learning_rate": 2.538549866248549e-06,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 76246115,
+      "step": 3583
+    },
+    {
+      "epoch": 0.430950520050502,
+      "grad_norm": 2.1396259372566644,
+      "learning_rate": 2.5377996263768265e-06,
+      "loss": 0.8156,
+      "num_input_tokens_seen": 76263915,
+      "step": 3584
+    },
+    {
+      "epoch": 0.4310707629411411,
+      "grad_norm": 1.9521410900541882,
+      "learning_rate": 2.5370493049213285e-06,
+      "loss": 0.6778,
+      "num_input_tokens_seen": 76283280,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4311910058317802,
+      "grad_norm": 2.2755428940754165,
+      "learning_rate": 2.536298901995878e-06,
+      "loss": 0.7943,
+      "num_input_tokens_seen": 76302210,
+      "step": 3586
+    },
+    {
+      "epoch": 0.43131124872241927,
+      "grad_norm": 1.8481039970468383,
+      "learning_rate": 2.535548417714311e-06,
+      "loss": 0.798,
+      "num_input_tokens_seen": 76321230,
+      "step": 3587
+    },
+    {
+      "epoch": 0.4314314916130584,
+      "grad_norm": 1.497113061996737,
+      "learning_rate": 2.534797852190474e-06,
+      "loss": 0.8583,
+      "num_input_tokens_seen": 76341130,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4315517345036975,
+      "grad_norm": 3.1375881868608766,
+      "learning_rate": 2.5340472055382287e-06,
+      "loss": 0.8066,
+      "num_input_tokens_seen": 76356880,
+      "step": 3589
+    },
+    {
+      "epoch": 0.43167197739433655,
+      "grad_norm": 2.2734636433953668,
+      "learning_rate": 2.5332964778714463e-06,
+      "loss": 0.8056,
+      "num_input_tokens_seen": 76373785,
+      "step": 3590
+    },
+    {
+      "epoch": 0.43179222028497566,
+      "grad_norm": 1.86802820388375,
+      "learning_rate": 2.5325456693040123e-06,
+      "loss": 0.6638,
+      "num_input_tokens_seen": 76390700,
+      "step": 3591
+    },
+    {
+      "epoch": 0.43191246317561477,
+      "grad_norm": 2.2976409629793046,
+      "learning_rate": 2.531794779949824e-06,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 76408320,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4320327060662538,
+      "grad_norm": 2.094413064199808,
+      "learning_rate": 2.5310438099227907e-06,
+      "loss": 0.8718,
+      "num_input_tokens_seen": 76425305,
+      "step": 3593
+    },
+    {
+      "epoch": 0.43215294895689293,
+      "grad_norm": 1.3208136464733782,
+      "learning_rate": 2.530292759336833e-06,
+      "loss": 0.5576,
+      "num_input_tokens_seen": 76485760,
+      "step": 3594
+    },
+    {
+      "epoch": 0.432273191847532,
+      "grad_norm": 2.7364501627253808,
+      "learning_rate": 2.5295416283058855e-06,
+      "loss": 0.704,
+      "num_input_tokens_seen": 76504345,
+      "step": 3595
+    },
+    {
+      "epoch": 0.4323934347381711,
+      "grad_norm": 1.71068111164653,
+      "learning_rate": 2.5287904169438943e-06,
+      "loss": 0.6573,
+      "num_input_tokens_seen": 76523270,
+      "step": 3596
+    },
+    {
+      "epoch": 0.4325136776288102,
+      "grad_norm": 2.6516865699439744,
+      "learning_rate": 2.528039125364817e-06,
+      "loss": 0.6491,
+      "num_input_tokens_seen": 76541795,
+      "step": 3597
+    },
+    {
+      "epoch": 0.43263392051944927,
+      "grad_norm": 2.1088631903856583,
+      "learning_rate": 2.527287753682624e-06,
+      "loss": 0.7513,
+      "num_input_tokens_seen": 76560310,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4327541634100884,
+      "grad_norm": 3.1322998025485775,
+      "learning_rate": 2.5265363020112986e-06,
+      "loss": 0.6933,
+      "num_input_tokens_seen": 76580350,
+      "step": 3599
+    },
+    {
+      "epoch": 0.4328744063007275,
+      "grad_norm": 2.4521763991357903,
+      "learning_rate": 2.5257847704648348e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 76601300,
+      "step": 3600
+    },
+    {
+      "epoch": 0.43299464919136654,
+      "grad_norm": 2.0146789461852106,
+      "learning_rate": 2.525033159157239e-06,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 76617335,
+      "step": 3601
+    },
+    {
+      "epoch": 0.43311489208200565,
+      "grad_norm": 1.9533350692708533,
+      "learning_rate": 2.52428146820253e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 76635310,
+      "step": 3602
+    },
+    {
+      "epoch": 0.43323513497264476,
+      "grad_norm": 2.1354105268103307,
+      "learning_rate": 2.52352969771474e-06,
+      "loss": 0.8207,
+      "num_input_tokens_seen": 76654255,
+      "step": 3603
+    },
+    {
+      "epoch": 0.4333553778632838,
+      "grad_norm": 18.34475598522582,
+      "learning_rate": 2.5227778478079106e-06,
+      "loss": 0.8761,
+      "num_input_tokens_seen": 76673385,
+      "step": 3604
+    },
+    {
+      "epoch": 0.43347562075392293,
+      "grad_norm": 1.6200273330478285,
+      "learning_rate": 2.522025918596098e-06,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 76691405,
+      "step": 3605
+    },
+    {
+      "epoch": 0.43359586364456204,
+      "grad_norm": 2.4481792368894895,
+      "learning_rate": 2.5212739101933674e-06,
+      "loss": 0.6531,
+      "num_input_tokens_seen": 76714305,
+      "step": 3606
+    },
+    {
+      "epoch": 0.4337161065352011,
+      "grad_norm": 3.8021683757632614,
+      "learning_rate": 2.5205218227138e-06,
+      "loss": 0.867,
+      "num_input_tokens_seen": 76726980,
+      "step": 3607
+    },
+    {
+      "epoch": 0.4338363494258402,
+      "grad_norm": 2.1433950268519095,
+      "learning_rate": 2.519769656271486e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 76744120,
+      "step": 3608
+    },
+    {
+      "epoch": 0.43395659231647926,
+      "grad_norm": 2.689230813678168,
+      "learning_rate": 2.5190174109805285e-06,
+      "loss": 0.6817,
+      "num_input_tokens_seen": 76763665,
+      "step": 3609
+    },
+    {
+      "epoch": 0.43407683520711837,
+      "grad_norm": 2.499055485090714,
+      "learning_rate": 2.518265086955042e-06,
+      "loss": 0.6307,
+      "num_input_tokens_seen": 76781105,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4341970780977575,
+      "grad_norm": 2.8876975409281487,
+      "learning_rate": 2.5175126843091538e-06,
+      "loss": 0.8375,
+      "num_input_tokens_seen": 76800195,
+      "step": 3611
+    },
+    {
+      "epoch": 0.43431732098839654,
+      "grad_norm": 1.9085992647345504,
+      "learning_rate": 2.5167602031570026e-06,
+      "loss": 0.7391,
+      "num_input_tokens_seen": 76820100,
+      "step": 3612
+    },
+    {
+      "epoch": 0.43443756387903565,
+      "grad_norm": 1.9185820414804613,
+      "learning_rate": 2.51600764361274e-06,
+      "loss": 0.731,
+      "num_input_tokens_seen": 76841345,
+      "step": 3613
+    },
+    {
+      "epoch": 0.43455780676967476,
+      "grad_norm": 2.570985670695853,
+      "learning_rate": 2.5152550057905283e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 76860955,
+      "step": 3614
+    },
+    {
+      "epoch": 0.4346780496603138,
+      "grad_norm": 4.317111983263118,
+      "learning_rate": 2.514502289804542e-06,
+      "loss": 0.7602,
+      "num_input_tokens_seen": 76879860,
+      "step": 3615
+    },
+    {
+      "epoch": 0.4347982925509529,
+      "grad_norm": 2.9784663084731866,
+      "learning_rate": 2.5137494957689664e-06,
+      "loss": 0.8964,
+      "num_input_tokens_seen": 76895190,
+      "step": 3616
+    },
+    {
+      "epoch": 0.43491853544159204,
+      "grad_norm": 0.7686802857635239,
+      "learning_rate": 2.5129966237980016e-06,
+      "loss": 0.5976,
+      "num_input_tokens_seen": 76957905,
+      "step": 3617
+    },
+    {
+      "epoch": 0.4350387783322311,
+      "grad_norm": 2.0116047661381584,
+      "learning_rate": 2.5122436740058565e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 76976990,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4351590212228702,
+      "grad_norm": 2.175968598712834,
+      "learning_rate": 2.5114906465067537e-06,
+      "loss": 0.853,
+      "num_input_tokens_seen": 76997695,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4352792641135093,
+      "grad_norm": 2.433118670613242,
+      "learning_rate": 2.510737541414926e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 77016660,
+      "step": 3620
+    },
+    {
+      "epoch": 0.43539950700414837,
+      "grad_norm": 2.5019388654942802,
+      "learning_rate": 2.5099843588446197e-06,
+      "loss": 0.7163,
+      "num_input_tokens_seen": 77034700,
+      "step": 3621
+    },
+    {
+      "epoch": 0.4355197498947875,
+      "grad_norm": 2.034688791974804,
+      "learning_rate": 2.5092310989100916e-06,
+      "loss": 0.6202,
+      "num_input_tokens_seen": 77054290,
+      "step": 3622
+    },
+    {
+      "epoch": 0.4356399927854266,
+      "grad_norm": 2.4958012752962753,
+      "learning_rate": 2.508477761725611e-06,
+      "loss": 0.7466,
+      "num_input_tokens_seen": 77072285,
+      "step": 3623
+    },
+    {
+      "epoch": 0.43576023567606564,
+      "grad_norm": 2.071654594007142,
+      "learning_rate": 2.507724347405458e-06,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 77089955,
+      "step": 3624
+    },
+    {
+      "epoch": 0.43588047856670475,
+      "grad_norm": 1.9638713136687704,
+      "learning_rate": 2.5069708560639243e-06,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 77107585,
+      "step": 3625
+    },
+    {
+      "epoch": 0.4360007214573438,
+      "grad_norm": 2.1562380783606545,
+      "learning_rate": 2.5062172878153158e-06,
+      "loss": 0.6119,
+      "num_input_tokens_seen": 77126580,
+      "step": 3626
+    },
+    {
+      "epoch": 0.4361209643479829,
+      "grad_norm": 1.8232627785844573,
+      "learning_rate": 2.505463642773947e-06,
+      "loss": 0.8696,
+      "num_input_tokens_seen": 77146265,
+      "step": 3627
+    },
+    {
+      "epoch": 0.43624120723862203,
+      "grad_norm": 2.3733395529526455,
+      "learning_rate": 2.5047099210541455e-06,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 77162800,
+      "step": 3628
+    },
+    {
+      "epoch": 0.4363614501292611,
+      "grad_norm": 2.463057211515184,
+      "learning_rate": 2.50395612277025e-06,
+      "loss": 0.8362,
+      "num_input_tokens_seen": 77178375,
+      "step": 3629
+    },
+    {
+      "epoch": 0.4364816930199002,
+      "grad_norm": 3.2358110808639786,
+      "learning_rate": 2.503202248036612e-06,
+      "loss": 0.7287,
+      "num_input_tokens_seen": 77196950,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4366019359105393,
+      "grad_norm": 1.7480023190368053,
+      "learning_rate": 2.5024482969675927e-06,
+      "loss": 0.7301,
+      "num_input_tokens_seen": 77216625,
+      "step": 3631
+    },
+    {
+      "epoch": 0.43672217880117836,
+      "grad_norm": 1.9845791676509061,
+      "learning_rate": 2.501694269677566e-06,
+      "loss": 0.8411,
+      "num_input_tokens_seen": 77234115,
+      "step": 3632
+    },
+    {
+      "epoch": 0.4368424216918175,
+      "grad_norm": 2.912556517794492,
+      "learning_rate": 2.500940166280918e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 77252265,
+      "step": 3633
+    },
+    {
+      "epoch": 0.4369626645824566,
+      "grad_norm": 2.094067371107186,
+      "learning_rate": 2.5001859868920447e-06,
+      "loss": 0.7894,
+      "num_input_tokens_seen": 77271470,
+      "step": 3634
+    },
+    {
+      "epoch": 0.43708290747309564,
+      "grad_norm": 2.5203751968746846,
+      "learning_rate": 2.499431731625355e-06,
+      "loss": 0.7683,
+      "num_input_tokens_seen": 77290215,
+      "step": 3635
+    },
+    {
+      "epoch": 0.43720315036373475,
+      "grad_norm": 2.0647099265792463,
+      "learning_rate": 2.4986774005952686e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 77312310,
+      "step": 3636
+    },
+    {
+      "epoch": 0.43732339325437386,
+      "grad_norm": 2.1886424785256406,
+      "learning_rate": 2.4979229939162175e-06,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 77330810,
+      "step": 3637
+    },
+    {
+      "epoch": 0.4374436361450129,
+      "grad_norm": 2.107284108286668,
+      "learning_rate": 2.4971685117026433e-06,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 77350295,
+      "step": 3638
+    },
+    {
+      "epoch": 0.437563879035652,
+      "grad_norm": 1.6971732145332628,
+      "learning_rate": 2.4964139540690018e-06,
+      "loss": 0.7639,
+      "num_input_tokens_seen": 77373350,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4376841219262911,
+      "grad_norm": 1.9521561129031817,
+      "learning_rate": 2.4956593211297576e-06,
+      "loss": 0.7198,
+      "num_input_tokens_seen": 77390815,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4378043648169302,
+      "grad_norm": 2.1932654525756385,
+      "learning_rate": 2.494904612999389e-06,
+      "loss": 0.7575,
+      "num_input_tokens_seen": 77409245,
+      "step": 3641
+    },
+    {
+      "epoch": 0.4379246077075693,
+      "grad_norm": 0.8179128352244903,
+      "learning_rate": 2.494149829792384e-06,
+      "loss": 0.6056,
+      "num_input_tokens_seen": 77469535,
+      "step": 3642
+    },
+    {
+      "epoch": 0.43804485059820836,
+      "grad_norm": 1.7350490027753196,
+      "learning_rate": 2.4933949716232424e-06,
+      "loss": 0.699,
+      "num_input_tokens_seen": 77486780,
+      "step": 3643
+    },
+    {
+      "epoch": 0.43816509348884747,
+      "grad_norm": 2.2946995033763287,
+      "learning_rate": 2.4926400386064763e-06,
+      "loss": 0.7326,
+      "num_input_tokens_seen": 77504865,
+      "step": 3644
+    },
+    {
+      "epoch": 0.4382853363794866,
+      "grad_norm": 1.9483651769910342,
+      "learning_rate": 2.491885030856608e-06,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 77522680,
+      "step": 3645
+    },
+    {
+      "epoch": 0.43840557927012563,
+      "grad_norm": 22.566946495723524,
+      "learning_rate": 2.4911299484881713e-06,
+      "loss": 0.828,
+      "num_input_tokens_seen": 77539930,
+      "step": 3646
+    },
+    {
+      "epoch": 0.43852582216076474,
+      "grad_norm": 1.5670642119020959,
+      "learning_rate": 2.490374791615712e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 77559675,
+      "step": 3647
+    },
+    {
+      "epoch": 0.43864606505140386,
+      "grad_norm": 2.9081269667873637,
+      "learning_rate": 2.4896195603537867e-06,
+      "loss": 0.7766,
+      "num_input_tokens_seen": 77574005,
+      "step": 3648
+    },
+    {
+      "epoch": 0.4387663079420429,
+      "grad_norm": 2.1272161158330625,
+      "learning_rate": 2.488864254816964e-06,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 77592415,
+      "step": 3649
+    },
+    {
+      "epoch": 0.438886550832682,
+      "grad_norm": 3.5252548198920475,
+      "learning_rate": 2.4881088751198213e-06,
+      "loss": 0.6746,
+      "num_input_tokens_seen": 77610295,
+      "step": 3650
+    },
+    {
+      "epoch": 0.43900679372332113,
+      "grad_norm": 4.572775541434697,
+      "learning_rate": 2.4873534213769517e-06,
+      "loss": 0.6426,
+      "num_input_tokens_seen": 77625245,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4391270366139602,
+      "grad_norm": 1.8663018221376073,
+      "learning_rate": 2.4865978937029547e-06,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 77643945,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4392472795045993,
+      "grad_norm": 1.7993624891513738,
+      "learning_rate": 2.485842292212445e-06,
+      "loss": 0.6652,
+      "num_input_tokens_seen": 77664880,
+      "step": 3653
+    },
+    {
+      "epoch": 0.4393675223952384,
+      "grad_norm": 2.0300455020310286,
+      "learning_rate": 2.485086617020045e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 77683095,
+      "step": 3654
+    },
+    {
+      "epoch": 0.43948776528587746,
+      "grad_norm": 2.288980954422264,
+      "learning_rate": 2.4843308682403903e-06,
+      "loss": 0.8124,
+      "num_input_tokens_seen": 77699730,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4396080081765166,
+      "grad_norm": 1.727497387055552,
+      "learning_rate": 2.483575045988129e-06,
+      "loss": 0.8238,
+      "num_input_tokens_seen": 77716075,
+      "step": 3656
+    },
+    {
+      "epoch": 0.43972825106715563,
+      "grad_norm": 3.714351063851641,
+      "learning_rate": 2.4828191503779177e-06,
+      "loss": 0.8084,
+      "num_input_tokens_seen": 77733895,
+      "step": 3657
+    },
+    {
+      "epoch": 0.43984849395779474,
+      "grad_norm": 2.1617231967722716,
+      "learning_rate": 2.482063181524425e-06,
+      "loss": 0.8881,
+      "num_input_tokens_seen": 77749515,
+      "step": 3658
+    },
+    {
+      "epoch": 0.43996873684843385,
+      "grad_norm": 2.2832959228087484,
+      "learning_rate": 2.481307139542331e-06,
+      "loss": 0.8055,
+      "num_input_tokens_seen": 77766800,
+      "step": 3659
+    },
+    {
+      "epoch": 0.4400889797390729,
+      "grad_norm": 2.0174947780635413,
+      "learning_rate": 2.4805510245463263e-06,
+      "loss": 0.6452,
+      "num_input_tokens_seen": 77786675,
+      "step": 3660
+    },
+    {
+      "epoch": 0.440209222629712,
+      "grad_norm": 3.1793437415229024,
+      "learning_rate": 2.4797948366511137e-06,
+      "loss": 0.5921,
+      "num_input_tokens_seen": 77806105,
+      "step": 3661
+    },
+    {
+      "epoch": 0.4403294655203511,
+      "grad_norm": 2.108618988979303,
+      "learning_rate": 2.479038575971405e-06,
+      "loss": 0.7656,
+      "num_input_tokens_seen": 77824890,
+      "step": 3662
+    },
+    {
+      "epoch": 0.4404497084109902,
+      "grad_norm": 3.037734146247169,
+      "learning_rate": 2.478282242621926e-06,
+      "loss": 0.7168,
+      "num_input_tokens_seen": 77845070,
+      "step": 3663
+    },
+    {
+      "epoch": 0.4405699513016293,
+      "grad_norm": 0.9227886542991343,
+      "learning_rate": 2.4775258367174108e-06,
+      "loss": 0.6361,
+      "num_input_tokens_seen": 77912555,
+      "step": 3664
+    },
+    {
+      "epoch": 0.4406901941922684,
+      "grad_norm": 2.4583004285564085,
+      "learning_rate": 2.476769358372606e-06,
+      "loss": 0.789,
+      "num_input_tokens_seen": 77933925,
+      "step": 3665
+    },
+    {
+      "epoch": 0.44081043708290746,
+      "grad_norm": 2.370898869345898,
+      "learning_rate": 2.4760128077022687e-06,
+      "loss": 0.745,
+      "num_input_tokens_seen": 77951780,
+      "step": 3666
+    },
+    {
+      "epoch": 0.44093067997354657,
+      "grad_norm": 1.5650845197581849,
+      "learning_rate": 2.4752561848211672e-06,
+      "loss": 0.6769,
+      "num_input_tokens_seen": 77973900,
+      "step": 3667
+    },
+    {
+      "epoch": 0.4410509228641857,
+      "grad_norm": 4.884257453505635,
+      "learning_rate": 2.4744994898440797e-06,
+      "loss": 0.7146,
+      "num_input_tokens_seen": 77992410,
+      "step": 3668
+    },
+    {
+      "epoch": 0.44117116575482473,
+      "grad_norm": 2.2495014818216394,
+      "learning_rate": 2.473742722885797e-06,
+      "loss": 0.827,
+      "num_input_tokens_seen": 78011150,
+      "step": 3669
+    },
+    {
+      "epoch": 0.44129140864546385,
+      "grad_norm": 2.26725516168272,
+      "learning_rate": 2.4729858840611197e-06,
+      "loss": 0.6588,
+      "num_input_tokens_seen": 78029780,
+      "step": 3670
+    },
+    {
+      "epoch": 0.4414116515361029,
+      "grad_norm": 2.1326809384514123,
+      "learning_rate": 2.4722289734848605e-06,
+      "loss": 0.7204,
+      "num_input_tokens_seen": 78049965,
+      "step": 3671
+    },
+    {
+      "epoch": 0.441531894426742,
+      "grad_norm": 2.032060004611709,
+      "learning_rate": 2.4714719912718405e-06,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 78066810,
+      "step": 3672
+    },
+    {
+      "epoch": 0.4416521373173811,
+      "grad_norm": 2.002417550788984,
+      "learning_rate": 2.470714937536895e-06,
+      "loss": 0.7916,
+      "num_input_tokens_seen": 78085255,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4417723802080202,
+      "grad_norm": 2.018039938066539,
+      "learning_rate": 2.469957812394868e-06,
+      "loss": 0.7059,
+      "num_input_tokens_seen": 78103785,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4418926230986593,
+      "grad_norm": 2.192917684356724,
+      "learning_rate": 2.4692006159606148e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 78121035,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4420128659892984,
+      "grad_norm": 1.717614081107856,
+      "learning_rate": 2.468443348349e-06,
+      "loss": 0.7824,
+      "num_input_tokens_seen": 78138630,
+      "step": 3676
+    },
+    {
+      "epoch": 0.44213310887993745,
+      "grad_norm": 2.907081507203634,
+      "learning_rate": 2.467686009674903e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 78152800,
+      "step": 3677
+    },
+    {
+      "epoch": 0.44225335177057656,
+      "grad_norm": 2.5397604118202453,
+      "learning_rate": 2.4669286000532085e-06,
+      "loss": 0.8437,
+      "num_input_tokens_seen": 78167825,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4423735946612157,
+      "grad_norm": 3.665972131634199,
+      "learning_rate": 2.466171119598818e-06,
+      "loss": 0.7134,
+      "num_input_tokens_seen": 78187515,
+      "step": 3679
+    },
+    {
+      "epoch": 0.44249383755185473,
+      "grad_norm": 1.9922627420813293,
+      "learning_rate": 2.465413568426639e-06,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 78208185,
+      "step": 3680
+    },
+    {
+      "epoch": 0.44261408044249384,
+      "grad_norm": 1.6473804742234324,
+      "learning_rate": 2.464655946651592e-06,
+      "loss": 0.8107,
+      "num_input_tokens_seen": 78226910,
+      "step": 3681
+    },
+    {
+      "epoch": 0.44273432333313295,
+      "grad_norm": 2.3946407310600355,
+      "learning_rate": 2.4638982543886065e-06,
+      "loss": 0.7965,
+      "num_input_tokens_seen": 78246670,
+      "step": 3682
+    },
+    {
+      "epoch": 0.442854566223772,
+      "grad_norm": 2.3871891828944394,
+      "learning_rate": 2.4631404917526254e-06,
+      "loss": 0.8701,
+      "num_input_tokens_seen": 78263345,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4429748091144111,
+      "grad_norm": 2.0195508221209066,
+      "learning_rate": 2.4623826588585995e-06,
+      "loss": 0.7825,
+      "num_input_tokens_seen": 78283335,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4430950520050502,
+      "grad_norm": 1.4742967155449034,
+      "learning_rate": 2.461624755821492e-06,
+      "loss": 0.8225,
+      "num_input_tokens_seen": 78302535,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4432152948956893,
+      "grad_norm": 1.7813800763465129,
+      "learning_rate": 2.460866782756276e-06,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 78321585,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4433355377863284,
+      "grad_norm": 1.891004880523317,
+      "learning_rate": 2.460108739777936e-06,
+      "loss": 0.8899,
+      "num_input_tokens_seen": 78340440,
+      "step": 3687
+    },
+    {
+      "epoch": 0.44345578067696745,
+      "grad_norm": 2.0281705254109124,
+      "learning_rate": 2.4593506270014656e-06,
+      "loss": 0.7548,
+      "num_input_tokens_seen": 78359130,
+      "step": 3688
+    },
+    {
+      "epoch": 0.44357602356760656,
+      "grad_norm": 4.169041820865772,
+      "learning_rate": 2.45859244454187e-06,
+      "loss": 0.8156,
+      "num_input_tokens_seen": 78378640,
+      "step": 3689
+    },
+    {
+      "epoch": 0.44369626645824567,
+      "grad_norm": 2.0260239058025675,
+      "learning_rate": 2.4578341925141655e-06,
+      "loss": 0.6545,
+      "num_input_tokens_seen": 78397575,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4438165093488847,
+      "grad_norm": 2.7234866321685485,
+      "learning_rate": 2.457075871033378e-06,
+      "loss": 0.7192,
+      "num_input_tokens_seen": 78419170,
+      "step": 3691
+    },
+    {
+      "epoch": 0.44393675223952384,
+      "grad_norm": 2.112027591235452,
+      "learning_rate": 2.4563174802145445e-06,
+      "loss": 0.879,
+      "num_input_tokens_seen": 78436140,
+      "step": 3692
+    },
+    {
+      "epoch": 0.44405699513016295,
+      "grad_norm": 3.7680237600967557,
+      "learning_rate": 2.455559020172712e-06,
+      "loss": 0.4955,
+      "num_input_tokens_seen": 78503215,
+      "step": 3693
+    },
+    {
+      "epoch": 0.444177238020802,
+      "grad_norm": 1.876866831753697,
+      "learning_rate": 2.454800491022938e-06,
+      "loss": 0.8955,
+      "num_input_tokens_seen": 78520510,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4442974809114411,
+      "grad_norm": 1.7125622645485121,
+      "learning_rate": 2.4540418928802913e-06,
+      "loss": 0.8648,
+      "num_input_tokens_seen": 78538965,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4444177238020802,
+      "grad_norm": 2.2758106352468,
+      "learning_rate": 2.4532832258598506e-06,
+      "loss": 0.6614,
+      "num_input_tokens_seen": 78556515,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4445379666927193,
+      "grad_norm": 2.026183457346423,
+      "learning_rate": 2.4525244900767047e-06,
+      "loss": 0.8038,
+      "num_input_tokens_seen": 78577050,
+      "step": 3697
+    },
+    {
+      "epoch": 0.4446582095833584,
+      "grad_norm": 0.8441549179958148,
+      "learning_rate": 2.4517656856459536e-06,
+      "loss": 0.6295,
+      "num_input_tokens_seen": 78642615,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4447784524739975,
+      "grad_norm": 2.020079458676657,
+      "learning_rate": 2.4510068126827073e-06,
+      "loss": 0.6748,
+      "num_input_tokens_seen": 78663335,
+      "step": 3699
+    },
+    {
+      "epoch": 0.44489869536463655,
+      "grad_norm": 2.4130134941921186,
+      "learning_rate": 2.4502478713020854e-06,
+      "loss": 0.8211,
+      "num_input_tokens_seen": 78680830,
+      "step": 3700
+    },
+    {
+      "epoch": 0.44501893825527566,
+      "grad_norm": 16.391931520600966,
+      "learning_rate": 2.44948886161922e-06,
+      "loss": 0.8309,
+      "num_input_tokens_seen": 78699565,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4451391811459148,
+      "grad_norm": 1.6452690252463997,
+      "learning_rate": 2.4487297837492524e-06,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 78718450,
+      "step": 3702
+    },
+    {
+      "epoch": 0.44525942403655383,
+      "grad_norm": 1.9895799845443682,
+      "learning_rate": 2.4479706378073327e-06,
+      "loss": 0.6096,
+      "num_input_tokens_seen": 78736710,
+      "step": 3703
+    },
+    {
+      "epoch": 0.44537966692719294,
+      "grad_norm": 3.7198373861827783,
+      "learning_rate": 2.447211423908623e-06,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 78756475,
+      "step": 3704
+    },
+    {
+      "epoch": 0.445499909817832,
+      "grad_norm": 2.985763883121609,
+      "learning_rate": 2.4464521421682966e-06,
+      "loss": 0.7402,
+      "num_input_tokens_seen": 78773785,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4456201527084711,
+      "grad_norm": 1.413429154065012,
+      "learning_rate": 2.445692792701534e-06,
+      "loss": 0.868,
+      "num_input_tokens_seen": 78794545,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4457403955991102,
+      "grad_norm": 2.6214275513829484,
+      "learning_rate": 2.4449333756235307e-06,
+      "loss": 0.7553,
+      "num_input_tokens_seen": 78810980,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4458606384897493,
+      "grad_norm": 2.5251702152819546,
+      "learning_rate": 2.4441738910494867e-06,
+      "loss": 0.7864,
+      "num_input_tokens_seen": 78825435,
+      "step": 3708
+    },
+    {
+      "epoch": 0.4459808813803884,
+      "grad_norm": 1.9437303643528219,
+      "learning_rate": 2.4434143390946176e-06,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 78843965,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4461011242710275,
+      "grad_norm": 2.1728020176654326,
+      "learning_rate": 2.4426547198741457e-06,
+      "loss": 0.8507,
+      "num_input_tokens_seen": 78861890,
+      "step": 3710
+    },
+    {
+      "epoch": 0.44622136716166655,
+      "grad_norm": 2.582708958514176,
+      "learning_rate": 2.441895033503305e-06,
+      "loss": 0.7471,
+      "num_input_tokens_seen": 78879530,
+      "step": 3711
+    },
+    {
+      "epoch": 0.44634161005230566,
+      "grad_norm": 2.4183288616389196,
+      "learning_rate": 2.4411352800973375e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 78897685,
+      "step": 3712
+    },
+    {
+      "epoch": 0.44646185294294477,
+      "grad_norm": 3.45301780210301,
+      "learning_rate": 2.4403754597715005e-06,
+      "loss": 0.75,
+      "num_input_tokens_seen": 78916850,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4465820958335838,
+      "grad_norm": 3.234722379197625,
+      "learning_rate": 2.4396155726410553e-06,
+      "loss": 0.919,
+      "num_input_tokens_seen": 78935180,
+      "step": 3714
+    },
+    {
+      "epoch": 0.44670233872422294,
+      "grad_norm": 2.770958662360123,
+      "learning_rate": 2.438855618821278e-06,
+      "loss": 0.9143,
+      "num_input_tokens_seen": 78950700,
+      "step": 3715
+    },
+    {
+      "epoch": 0.44682258161486205,
+      "grad_norm": 2.1322103001409896,
+      "learning_rate": 2.4380955984274513e-06,
+      "loss": 0.6712,
+      "num_input_tokens_seen": 78969075,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4469428245055011,
+      "grad_norm": 2.030468718166149,
+      "learning_rate": 2.4373355115748716e-06,
+      "loss": 0.7717,
+      "num_input_tokens_seen": 78989625,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4470630673961402,
+      "grad_norm": 1.9443336452359934,
+      "learning_rate": 2.436575358378842e-06,
+      "loss": 0.723,
+      "num_input_tokens_seen": 79008835,
+      "step": 3718
+    },
+    {
+      "epoch": 0.44718331028677927,
+      "grad_norm": 6.235313351045172,
+      "learning_rate": 2.4358151389546782e-06,
+      "loss": 0.8288,
+      "num_input_tokens_seen": 79025240,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4473035531774184,
+      "grad_norm": 2.356420714200738,
+      "learning_rate": 2.4350548534177035e-06,
+      "loss": 0.753,
+      "num_input_tokens_seen": 79041790,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4474237960680575,
+      "grad_norm": 1.741783537337227,
+      "learning_rate": 2.434294501883254e-06,
+      "loss": 0.6694,
+      "num_input_tokens_seen": 79064605,
+      "step": 3721
+    },
+    {
+      "epoch": 0.44754403895869654,
+      "grad_norm": 1.734957073196937,
+      "learning_rate": 2.4335340844666737e-06,
+      "loss": 0.6583,
+      "num_input_tokens_seen": 79083545,
+      "step": 3722
+    },
+    {
+      "epoch": 0.44766428184933565,
+      "grad_norm": 1.9204499828183228,
+      "learning_rate": 2.4327736012833178e-06,
+      "loss": 0.7117,
+      "num_input_tokens_seen": 79104985,
+      "step": 3723
+    },
+    {
+      "epoch": 0.44778452473997477,
+      "grad_norm": 2.118016722638842,
+      "learning_rate": 2.4320130524485506e-06,
+      "loss": 0.7586,
+      "num_input_tokens_seen": 79123500,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4479047676306138,
+      "grad_norm": 2.22545907522413,
+      "learning_rate": 2.4312524380777466e-06,
+      "loss": 0.7922,
+      "num_input_tokens_seen": 79142720,
+      "step": 3725
+    },
+    {
+      "epoch": 0.44802501052125293,
+      "grad_norm": 5.704156673247014,
+      "learning_rate": 2.4304917582862906e-06,
+      "loss": 0.7659,
+      "num_input_tokens_seen": 79161620,
+      "step": 3726
+    },
+    {
+      "epoch": 0.44814525341189204,
+      "grad_norm": 2.3372078822668563,
+      "learning_rate": 2.4297310131895774e-06,
+      "loss": 0.8781,
+      "num_input_tokens_seen": 79179885,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4482654963025311,
+      "grad_norm": 2.4597314985903473,
+      "learning_rate": 2.428970202903011e-06,
+      "loss": 0.7615,
+      "num_input_tokens_seen": 79197075,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4483857391931702,
+      "grad_norm": 1.8642653413137424,
+      "learning_rate": 2.4282093275420057e-06,
+      "loss": 0.8176,
+      "num_input_tokens_seen": 79215825,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4485059820838093,
+      "grad_norm": 2.668147409563175,
+      "learning_rate": 2.427448387221986e-06,
+      "loss": 0.6877,
+      "num_input_tokens_seen": 79232905,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4486262249744484,
+      "grad_norm": 1.8122634292229343,
+      "learning_rate": 2.426687382058386e-06,
+      "loss": 0.9255,
+      "num_input_tokens_seen": 79250905,
+      "step": 3731
+    },
+    {
+      "epoch": 0.4487464678650875,
+      "grad_norm": 0.9969375684357404,
+      "learning_rate": 2.425926312166649e-06,
+      "loss": 0.6162,
+      "num_input_tokens_seen": 79303500,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4488667107557266,
+      "grad_norm": 2.5817944547085805,
+      "learning_rate": 2.42516517766223e-06,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 79321300,
+      "step": 3733
+    },
+    {
+      "epoch": 0.44898695364636565,
+      "grad_norm": 1.875285052008389,
+      "learning_rate": 2.4244039786605907e-06,
+      "loss": 0.6709,
+      "num_input_tokens_seen": 79342025,
+      "step": 3734
+    },
+    {
+      "epoch": 0.44910719653700476,
+      "grad_norm": 2.3611656241427452,
+      "learning_rate": 2.4236427152772055e-06,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 79360150,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4492274394276438,
+      "grad_norm": 0.9030233703970052,
+      "learning_rate": 2.422881387627557e-06,
+      "loss": 0.6061,
+      "num_input_tokens_seen": 79412320,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4493476823182829,
+      "grad_norm": 1.4718579608470002,
+      "learning_rate": 2.422119995827139e-06,
+      "loss": 0.7644,
+      "num_input_tokens_seen": 79432165,
+      "step": 3737
+    },
+    {
+      "epoch": 0.44946792520892204,
+      "grad_norm": 2.4124441950882725,
+      "learning_rate": 2.4213585399914523e-06,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 79449090,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4495881680995611,
+      "grad_norm": 1.6798794610633387,
+      "learning_rate": 2.4205970202360113e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 79468375,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4497084109902002,
+      "grad_norm": 4.745589986704857,
+      "learning_rate": 2.4198354366763354e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 79486735,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4498286538808393,
+      "grad_norm": 2.217209334774954,
+      "learning_rate": 2.4190737894279587e-06,
+      "loss": 0.7769,
+      "num_input_tokens_seen": 79503825,
+      "step": 3741
+    },
+    {
+      "epoch": 0.44994889677147837,
+      "grad_norm": 2.196018819808298,
+      "learning_rate": 2.4183120786064203e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 79520420,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4500691396621175,
+      "grad_norm": 2.9469151327662706,
+      "learning_rate": 2.417550304327273e-06,
+      "loss": 0.8439,
+      "num_input_tokens_seen": 79538180,
+      "step": 3743
+    },
+    {
+      "epoch": 0.4501893825527566,
+      "grad_norm": 1.6444190561494996,
+      "learning_rate": 2.416788466706076e-06,
+      "loss": 0.7529,
+      "num_input_tokens_seen": 79560610,
+      "step": 3744
+    },
+    {
+      "epoch": 0.45030962544339564,
+      "grad_norm": 6.1292978724456155,
+      "learning_rate": 2.4160265658584e-06,
+      "loss": 0.8693,
+      "num_input_tokens_seen": 79575220,
+      "step": 3745
+    },
+    {
+      "epoch": 0.45042986833403476,
+      "grad_norm": 2.260887095181136,
+      "learning_rate": 2.4152646018998253e-06,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 79593890,
+      "step": 3746
+    },
+    {
+      "epoch": 0.45055011122467387,
+      "grad_norm": 1.76589947857936,
+      "learning_rate": 2.4145025749459407e-06,
+      "loss": 0.7122,
+      "num_input_tokens_seen": 79614635,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4506703541153129,
+      "grad_norm": 2.233100156065296,
+      "learning_rate": 2.413740485112344e-06,
+      "loss": 0.7016,
+      "num_input_tokens_seen": 79632695,
+      "step": 3748
+    },
+    {
+      "epoch": 0.45079059700595203,
+      "grad_norm": 1.6602487327608608,
+      "learning_rate": 2.412978332514646e-06,
+      "loss": 0.818,
+      "num_input_tokens_seen": 79651195,
+      "step": 3749
+    },
+    {
+      "epoch": 0.4509108398965911,
+      "grad_norm": 2.348598629174502,
+      "learning_rate": 2.412216117268462e-06,
+      "loss": 0.712,
+      "num_input_tokens_seen": 79671710,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4510310827872302,
+      "grad_norm": 2.7683436094340053,
+      "learning_rate": 2.4114538394894216e-06,
+      "loss": 0.8268,
+      "num_input_tokens_seen": 79689070,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4511513256778693,
+      "grad_norm": 1.9232511779175339,
+      "learning_rate": 2.4106914992931605e-06,
+      "loss": 0.8305,
+      "num_input_tokens_seen": 79706945,
+      "step": 3752
+    },
+    {
+      "epoch": 0.45127156856850836,
+      "grad_norm": 1.5574925932576171,
+      "learning_rate": 2.409929096795326e-06,
+      "loss": 0.7373,
+      "num_input_tokens_seen": 79727035,
+      "step": 3753
+    },
+    {
+      "epoch": 0.4513918114591475,
+      "grad_norm": 2.0787002980553804,
+      "learning_rate": 2.409166632111573e-06,
+      "loss": 0.786,
+      "num_input_tokens_seen": 79744890,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4515120543497866,
+      "grad_norm": 1.8420448826829345,
+      "learning_rate": 2.4084041053575674e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 79764030,
+      "step": 3755
+    },
+    {
+      "epoch": 0.45163229724042564,
+      "grad_norm": 1.9298420370521412,
+      "learning_rate": 2.4076415166489834e-06,
+      "loss": 0.7165,
+      "num_input_tokens_seen": 79783160,
+      "step": 3756
+    },
+    {
+      "epoch": 0.45175254013106475,
+      "grad_norm": 1.6077993217628501,
+      "learning_rate": 2.406878866101506e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 79801845,
+      "step": 3757
+    },
+    {
+      "epoch": 0.45187278302170386,
+      "grad_norm": 3.165426706991801,
+      "learning_rate": 2.4061161538308273e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 79818410,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4519930259123429,
+      "grad_norm": 2.002430782558665,
+      "learning_rate": 2.4053533799526523e-06,
+      "loss": 0.884,
+      "num_input_tokens_seen": 79833850,
+      "step": 3759
+    },
+    {
+      "epoch": 0.452113268802982,
+      "grad_norm": 1.7448498090995037,
+      "learning_rate": 2.404590544582691e-06,
+      "loss": 0.8562,
+      "num_input_tokens_seen": 79851805,
+      "step": 3760
+    },
+    {
+      "epoch": 0.45223351169362114,
+      "grad_norm": 1.7781375410708375,
+      "learning_rate": 2.403827647836666e-06,
+      "loss": 0.8025,
+      "num_input_tokens_seen": 79872080,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4523537545842602,
+      "grad_norm": 2.096053595024568,
+      "learning_rate": 2.4030646898303075e-06,
+      "loss": 0.6886,
+      "num_input_tokens_seen": 79893290,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4524739974748993,
+      "grad_norm": 2.048966028515164,
+      "learning_rate": 2.4023016706793566e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 79912805,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4525942403655384,
+      "grad_norm": 0.8614607412046528,
+      "learning_rate": 2.401538590499561e-06,
+      "loss": 0.6052,
+      "num_input_tokens_seen": 79972980,
+      "step": 3764
+    },
+    {
+      "epoch": 0.45271448325617747,
+      "grad_norm": 2.3897817943561463,
+      "learning_rate": 2.400775449406682e-06,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 79995895,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4528347261468166,
+      "grad_norm": 2.3967319642451623,
+      "learning_rate": 2.4000122475164846e-06,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 80016180,
+      "step": 3766
+    },
+    {
+      "epoch": 0.45295496903745563,
+      "grad_norm": 1.7532425034180619,
+      "learning_rate": 2.3992489849447484e-06,
+      "loss": 0.8918,
+      "num_input_tokens_seen": 80034355,
+      "step": 3767
+    },
+    {
+      "epoch": 0.45307521192809475,
+      "grad_norm": 2.757451334205097,
+      "learning_rate": 2.3984856618072584e-06,
+      "loss": 0.784,
+      "num_input_tokens_seen": 80054110,
+      "step": 3768
+    },
+    {
+      "epoch": 0.45319545481873386,
+      "grad_norm": 2.017176844906747,
+      "learning_rate": 2.39772227821981e-06,
+      "loss": 0.7341,
+      "num_input_tokens_seen": 80072465,
+      "step": 3769
+    },
+    {
+      "epoch": 0.4533156977093729,
+      "grad_norm": 2.2361428429504406,
+      "learning_rate": 2.3969588342982077e-06,
+      "loss": 0.7498,
+      "num_input_tokens_seen": 80091560,
+      "step": 3770
+    },
+    {
+      "epoch": 0.453435940600012,
+      "grad_norm": 1.7505164110371585,
+      "learning_rate": 2.396195330158267e-06,
+      "loss": 0.7185,
+      "num_input_tokens_seen": 80111170,
+      "step": 3771
+    },
+    {
+      "epoch": 0.45355618349065113,
+      "grad_norm": 2.7928530648112733,
+      "learning_rate": 2.395431765915809e-06,
+      "loss": 0.7878,
+      "num_input_tokens_seen": 80131225,
+      "step": 3772
+    },
+    {
+      "epoch": 0.4536764263812902,
+      "grad_norm": 0.9032585638216502,
+      "learning_rate": 2.394668141686667e-06,
+      "loss": 0.6225,
+      "num_input_tokens_seen": 80192910,
+      "step": 3773
+    },
+    {
+      "epoch": 0.4537966692719293,
+      "grad_norm": 1.9911977149613673,
+      "learning_rate": 2.393904457586681e-06,
+      "loss": 0.6884,
+      "num_input_tokens_seen": 80215380,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4539169121625684,
+      "grad_norm": 2.396334800499764,
+      "learning_rate": 2.3931407137317024e-06,
+      "loss": 0.7528,
+      "num_input_tokens_seen": 80235255,
+      "step": 3775
+    },
+    {
+      "epoch": 0.45403715505320746,
+      "grad_norm": 2.0798927502133946,
+      "learning_rate": 2.3923769102375907e-06,
+      "loss": 0.8479,
+      "num_input_tokens_seen": 80253840,
+      "step": 3776
+    },
+    {
+      "epoch": 0.4541573979438466,
+      "grad_norm": 2.0454314347648315,
+      "learning_rate": 2.391613047220213e-06,
+      "loss": 0.783,
+      "num_input_tokens_seen": 80273460,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4542776408344857,
+      "grad_norm": 139.94738138022433,
+      "learning_rate": 2.390849124795447e-06,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 80289180,
+      "step": 3778
+    },
+    {
+      "epoch": 0.45439788372512474,
+      "grad_norm": 2.4334082567935567,
+      "learning_rate": 2.3900851430791804e-06,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 80306920,
+      "step": 3779
+    },
+    {
+      "epoch": 0.45451812661576385,
+      "grad_norm": 2.3866613316663896,
+      "learning_rate": 2.389321102187307e-06,
+      "loss": 0.8456,
+      "num_input_tokens_seen": 80325420,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4546383695064029,
+      "grad_norm": 1.7062829687281988,
+      "learning_rate": 2.3885570022357326e-06,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 80344270,
+      "step": 3781
+    },
+    {
+      "epoch": 0.454758612397042,
+      "grad_norm": 0.8254330817330938,
+      "learning_rate": 2.38779284334037e-06,
+      "loss": 0.6272,
+      "num_input_tokens_seen": 80408965,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4548788552876811,
+      "grad_norm": 2.9667837032790576,
+      "learning_rate": 2.387028625617141e-06,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 80427900,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4549990981783202,
+      "grad_norm": 8.112005825289943,
+      "learning_rate": 2.3862643491819766e-06,
+      "loss": 0.8462,
+      "num_input_tokens_seen": 80446185,
+      "step": 3784
+    },
+    {
+      "epoch": 0.4551193410689593,
+      "grad_norm": 1.824956936349683,
+      "learning_rate": 2.3855000141508186e-06,
+      "loss": 0.8323,
+      "num_input_tokens_seen": 80466060,
+      "step": 3785
+    },
+    {
+      "epoch": 0.4552395839595984,
+      "grad_norm": 3.843058233263913,
+      "learning_rate": 2.3847356206396143e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 80483090,
+      "step": 3786
+    },
+    {
+      "epoch": 0.45535982685023746,
+      "grad_norm": 2.507757479655883,
+      "learning_rate": 2.3839711687643227e-06,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 80504035,
+      "step": 3787
+    },
+    {
+      "epoch": 0.45548006974087657,
+      "grad_norm": 2.2504358524027532,
+      "learning_rate": 2.3832066586409097e-06,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 80523105,
+      "step": 3788
+    },
+    {
+      "epoch": 0.4556003126315157,
+      "grad_norm": 1.7567314217925787,
+      "learning_rate": 2.3824420903853516e-06,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 80541290,
+      "step": 3789
+    },
+    {
+      "epoch": 0.45572055552215474,
+      "grad_norm": 2.629940388333714,
+      "learning_rate": 2.3816774641136324e-06,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 80558265,
+      "step": 3790
+    },
+    {
+      "epoch": 0.45584079841279385,
+      "grad_norm": 3.1647475183819287,
+      "learning_rate": 2.380912779941745e-06,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 80581105,
+      "step": 3791
+    },
+    {
+      "epoch": 0.45596104130343296,
+      "grad_norm": 1.9372692586238671,
+      "learning_rate": 2.3801480379856918e-06,
+      "loss": 0.8232,
+      "num_input_tokens_seen": 80602535,
+      "step": 3792
+    },
+    {
+      "epoch": 0.456081284194072,
+      "grad_norm": 1.6754338468828784,
+      "learning_rate": 2.379383238361484e-06,
+      "loss": 0.8323,
+      "num_input_tokens_seen": 80621615,
+      "step": 3793
+    },
+    {
+      "epoch": 0.4562015270847111,
+      "grad_norm": 2.1655691707219087,
+      "learning_rate": 2.3786183811851403e-06,
+      "loss": 0.799,
+      "num_input_tokens_seen": 80642040,
+      "step": 3794
+    },
+    {
+      "epoch": 0.45632176997535023,
+      "grad_norm": 1.9917808540285618,
+      "learning_rate": 2.3778534665726892e-06,
+      "loss": 0.8007,
+      "num_input_tokens_seen": 80658590,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4564420128659893,
+      "grad_norm": 1.8134233830419824,
+      "learning_rate": 2.3770884946401677e-06,
+      "loss": 0.7242,
+      "num_input_tokens_seen": 80680060,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4565622557566284,
+      "grad_norm": 1.867800770801713,
+      "learning_rate": 2.3763234655036216e-06,
+      "loss": 0.78,
+      "num_input_tokens_seen": 80698980,
+      "step": 3797
+    },
+    {
+      "epoch": 0.45668249864726745,
+      "grad_norm": 2.247743058879139,
+      "learning_rate": 2.3755583792791046e-06,
+      "loss": 0.8616,
+      "num_input_tokens_seen": 80718570,
+      "step": 3798
+    },
+    {
+      "epoch": 0.45680274153790656,
+      "grad_norm": 2.4839603848860268,
+      "learning_rate": 2.3747932360826803e-06,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 80735220,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4569229844285457,
+      "grad_norm": 1.8296477558558883,
+      "learning_rate": 2.37402803603042e-06,
+      "loss": 0.8129,
+      "num_input_tokens_seen": 80752665,
+      "step": 3800
+    },
+    {
+      "epoch": 0.45704322731918473,
+      "grad_norm": 1.698678050773504,
+      "learning_rate": 2.3732627792384038e-06,
+      "loss": 0.6789,
+      "num_input_tokens_seen": 80773455,
+      "step": 3801
+    },
+    {
+      "epoch": 0.45716347020982384,
+      "grad_norm": 1.9137999886451664,
+      "learning_rate": 2.3724974658227207e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 80793965,
+      "step": 3802
+    },
+    {
+      "epoch": 0.45728371310046295,
+      "grad_norm": 1.9631892776388056,
+      "learning_rate": 2.3717320958994687e-06,
+      "loss": 0.7041,
+      "num_input_tokens_seen": 80811245,
+      "step": 3803
+    },
+    {
+      "epoch": 0.457403955991102,
+      "grad_norm": 3.1920370240549323,
+      "learning_rate": 2.3709666695847534e-06,
+      "loss": 0.6925,
+      "num_input_tokens_seen": 80829145,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4575241988817411,
+      "grad_norm": 2.060836246666886,
+      "learning_rate": 2.370201186994689e-06,
+      "loss": 0.7039,
+      "num_input_tokens_seen": 80852550,
+      "step": 3805
+    },
+    {
+      "epoch": 0.45764444177238023,
+      "grad_norm": 1.7283738489376945,
+      "learning_rate": 2.369435648245399e-06,
+      "loss": 0.6883,
+      "num_input_tokens_seen": 80872485,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4577646846630193,
+      "grad_norm": 2.3010550185635608,
+      "learning_rate": 2.368670053453015e-06,
+      "loss": 0.8464,
+      "num_input_tokens_seen": 80893895,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4578849275536584,
+      "grad_norm": 2.7277756514888947,
+      "learning_rate": 2.3679044027336757e-06,
+      "loss": 0.7334,
+      "num_input_tokens_seen": 80909505,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4580051704442975,
+      "grad_norm": 2.8588057214012985,
+      "learning_rate": 2.3671386962035326e-06,
+      "loss": 0.6884,
+      "num_input_tokens_seen": 80926695,
+      "step": 3809
+    },
+    {
+      "epoch": 0.45812541333493656,
+      "grad_norm": 1.9092598109887318,
+      "learning_rate": 2.36637293397874e-06,
+      "loss": 0.6893,
+      "num_input_tokens_seen": 80943350,
+      "step": 3810
+    },
+    {
+      "epoch": 0.45824565622557567,
+      "grad_norm": 2.8806544266460046,
+      "learning_rate": 2.3656071161754657e-06,
+      "loss": 0.7196,
+      "num_input_tokens_seen": 80958495,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4583658991162148,
+      "grad_norm": 2.7351544236256715,
+      "learning_rate": 2.3648412429098825e-06,
+      "loss": 0.6706,
+      "num_input_tokens_seen": 80976565,
+      "step": 3812
+    },
+    {
+      "epoch": 0.45848614200685384,
+      "grad_norm": 1.87232444334563,
+      "learning_rate": 2.3640753142981725e-06,
+      "loss": 0.8183,
+      "num_input_tokens_seen": 80993740,
+      "step": 3813
+    },
+    {
+      "epoch": 0.45860638489749295,
+      "grad_norm": 2.912523605110929,
+      "learning_rate": 2.3633093304565267e-06,
+      "loss": 0.7004,
+      "num_input_tokens_seen": 81012515,
+      "step": 3814
+    },
+    {
+      "epoch": 0.458726627788132,
+      "grad_norm": 2.2363559468469583,
+      "learning_rate": 2.3625432915011443e-06,
+      "loss": 0.6323,
+      "num_input_tokens_seen": 81034145,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4588468706787711,
+      "grad_norm": 1.6676404979652715,
+      "learning_rate": 2.361777197548233e-06,
+      "loss": 0.6493,
+      "num_input_tokens_seen": 81052695,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4589671135694102,
+      "grad_norm": 2.057095052356832,
+      "learning_rate": 2.3610110487140083e-06,
+      "loss": 0.7465,
+      "num_input_tokens_seen": 81070850,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4590873564600493,
+      "grad_norm": 1.6818940299341805,
+      "learning_rate": 2.3602448451146944e-06,
+      "loss": 0.8064,
+      "num_input_tokens_seen": 81090190,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4592075993506884,
+      "grad_norm": 2.54877569479225,
+      "learning_rate": 2.3594785868665245e-06,
+      "loss": 0.6913,
+      "num_input_tokens_seen": 81106215,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4593278422413275,
+      "grad_norm": 2.372858756803921,
+      "learning_rate": 2.3587122740857386e-06,
+      "loss": 0.7958,
+      "num_input_tokens_seen": 81123035,
+      "step": 3820
+    },
+    {
+      "epoch": 0.45944808513196655,
+      "grad_norm": 3.2960880943826423,
+      "learning_rate": 2.357945906888586e-06,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 81142195,
+      "step": 3821
+    },
+    {
+      "epoch": 0.45956832802260567,
+      "grad_norm": 2.7107916839648376,
+      "learning_rate": 2.3571794853913234e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 81159770,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4596885709132448,
+      "grad_norm": 1.9438614848448112,
+      "learning_rate": 2.3564130097102173e-06,
+      "loss": 0.8496,
+      "num_input_tokens_seen": 81179145,
+      "step": 3823
+    },
+    {
+      "epoch": 0.45980881380388383,
+      "grad_norm": 1.740798754292395,
+      "learning_rate": 2.355646479961541e-06,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 81198175,
+      "step": 3824
+    },
+    {
+      "epoch": 0.45992905669452294,
+      "grad_norm": 2.241563384165876,
+      "learning_rate": 2.354879896261576e-06,
+      "loss": 0.7128,
+      "num_input_tokens_seen": 81218105,
+      "step": 3825
+    },
+    {
+      "epoch": 0.46004929958516205,
+      "grad_norm": 1.9300561725741652,
+      "learning_rate": 2.3541132587266133e-06,
+      "loss": 0.5714,
+      "num_input_tokens_seen": 81240545,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4601695424758011,
+      "grad_norm": 1.89391879044887,
+      "learning_rate": 2.3533465674729515e-06,
+      "loss": 0.6911,
+      "num_input_tokens_seen": 81257495,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4602897853664402,
+      "grad_norm": 2.1205842654379787,
+      "learning_rate": 2.352579822616895e-06,
+      "loss": 0.7311,
+      "num_input_tokens_seen": 81274650,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4604100282570793,
+      "grad_norm": 2.0948566434895315,
+      "learning_rate": 2.351813024274761e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 81295725,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4605302711477184,
+      "grad_norm": 2.552269444440281,
+      "learning_rate": 2.3510461725628693e-06,
+      "loss": 0.7306,
+      "num_input_tokens_seen": 81315910,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4606505140383575,
+      "grad_norm": 2.0580959540314914,
+      "learning_rate": 2.350279267597554e-06,
+      "loss": 0.7004,
+      "num_input_tokens_seen": 81336270,
+      "step": 3831
+    },
+    {
+      "epoch": 0.46077075692899655,
+      "grad_norm": 2.2330723115707527,
+      "learning_rate": 2.349512309495151e-06,
+      "loss": 0.8259,
+      "num_input_tokens_seen": 81354335,
+      "step": 3832
+    },
+    {
+      "epoch": 0.46089099981963566,
+      "grad_norm": 5.726609525964799,
+      "learning_rate": 2.348745298372009e-06,
+      "loss": 0.7503,
+      "num_input_tokens_seen": 81377600,
+      "step": 3833
+    },
+    {
+      "epoch": 0.46101124271027477,
+      "grad_norm": 1.885454183726305,
+      "learning_rate": 2.347978234344483e-06,
+      "loss": 0.7873,
+      "num_input_tokens_seen": 81393525,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4611314856009138,
+      "grad_norm": 2.35352169795574,
+      "learning_rate": 2.3472111175289354e-06,
+      "loss": 0.6922,
+      "num_input_tokens_seen": 81415545,
+      "step": 3835
+    },
+    {
+      "epoch": 0.46125172849155294,
+      "grad_norm": 1.6657784260102466,
+      "learning_rate": 2.3464439480417374e-06,
+      "loss": 0.7095,
+      "num_input_tokens_seen": 81434785,
+      "step": 3836
+    },
+    {
+      "epoch": 0.46137197138219205,
+      "grad_norm": 3.987890317499909,
+      "learning_rate": 2.3456767259992676e-06,
+      "loss": 0.765,
+      "num_input_tokens_seen": 81452150,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4614922142728311,
+      "grad_norm": 2.416955459657737,
+      "learning_rate": 2.344909451517913e-06,
+      "loss": 0.8783,
+      "num_input_tokens_seen": 81469330,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4616124571634702,
+      "grad_norm": 1.7602869456078403,
+      "learning_rate": 2.34414212471407e-06,
+      "loss": 0.8074,
+      "num_input_tokens_seen": 81488845,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4617327000541093,
+      "grad_norm": 2.1242046268862427,
+      "learning_rate": 2.343374745704139e-06,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 81507270,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4618529429447484,
+      "grad_norm": 2.0528273999808495,
+      "learning_rate": 2.342607314604533e-06,
+      "loss": 0.8399,
+      "num_input_tokens_seen": 81526740,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4619731858353875,
+      "grad_norm": 1.7722429331940264,
+      "learning_rate": 2.3418398315316694e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 81544280,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4620934287260266,
+      "grad_norm": 2.690618721842699,
+      "learning_rate": 2.3410722966019755e-06,
+      "loss": 0.7817,
+      "num_input_tokens_seen": 81559115,
+      "step": 3843
+    },
+    {
+      "epoch": 0.46221367161666566,
+      "grad_norm": 1.7863766648521076,
+      "learning_rate": 2.3403047099318844e-06,
+      "loss": 0.659,
+      "num_input_tokens_seen": 81582905,
+      "step": 3844
+    },
+    {
+      "epoch": 0.46233391450730477,
+      "grad_norm": 4.1728082994628535,
+      "learning_rate": 2.3395370716378405e-06,
+      "loss": 0.7426,
+      "num_input_tokens_seen": 81600070,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4624541573979438,
+      "grad_norm": 2.288279965799928,
+      "learning_rate": 2.338769381836292e-06,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 81619400,
+      "step": 3846
+    },
+    {
+      "epoch": 0.46257440028858293,
+      "grad_norm": 2.2435270205080378,
+      "learning_rate": 2.3380016406436984e-06,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 81636600,
+      "step": 3847
+    },
+    {
+      "epoch": 0.46269464317922204,
+      "grad_norm": 2.1006268801042793,
+      "learning_rate": 2.337233848176524e-06,
+      "loss": 0.8105,
+      "num_input_tokens_seen": 81654090,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4628148860698611,
+      "grad_norm": 2.9068527049208135,
+      "learning_rate": 2.3364660045512435e-06,
+      "loss": 0.8172,
+      "num_input_tokens_seen": 81672570,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4629351289605002,
+      "grad_norm": 0.8045715669279327,
+      "learning_rate": 2.335698109884337e-06,
+      "loss": 0.6268,
+      "num_input_tokens_seen": 81737495,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4630553718511393,
+      "grad_norm": 0.8581761167215042,
+      "learning_rate": 2.334930164292294e-06,
+      "loss": 0.6516,
+      "num_input_tokens_seen": 81799765,
+      "step": 3851
+    },
+    {
+      "epoch": 0.4631756147417784,
+      "grad_norm": 2.173689605762162,
+      "learning_rate": 2.334162167891612e-06,
+      "loss": 0.7946,
+      "num_input_tokens_seen": 81816750,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4632958576324175,
+      "grad_norm": 2.812076533210322,
+      "learning_rate": 2.333394120798795e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 81835205,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4634161005230566,
+      "grad_norm": 3.3263053764128316,
+      "learning_rate": 2.332626023130354e-06,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 81853525,
+      "step": 3854
+    },
+    {
+      "epoch": 0.46353634341369565,
+      "grad_norm": 1.8021511067658027,
+      "learning_rate": 2.3318578750028107e-06,
+      "loss": 0.8688,
+      "num_input_tokens_seen": 81871845,
+      "step": 3855
+    },
+    {
+      "epoch": 0.46365658630433476,
+      "grad_norm": 1.8845017657351006,
+      "learning_rate": 2.3310896765326916e-06,
+      "loss": 0.7592,
+      "num_input_tokens_seen": 81892565,
+      "step": 3856
+    },
+    {
+      "epoch": 0.46377682919497387,
+      "grad_norm": 1.6916032474828888,
+      "learning_rate": 2.3303214278365317e-06,
+      "loss": 0.8299,
+      "num_input_tokens_seen": 81914155,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4638970720856129,
+      "grad_norm": 1.857656959469593,
+      "learning_rate": 2.3295531290308733e-06,
+      "loss": 0.8228,
+      "num_input_tokens_seen": 81932025,
+      "step": 3858
+    },
+    {
+      "epoch": 0.46401731497625204,
+      "grad_norm": 3.8669411752125904,
+      "learning_rate": 2.3287847802322678e-06,
+      "loss": 0.757,
+      "num_input_tokens_seen": 81947315,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4641375578668911,
+      "grad_norm": 2.0894538407598287,
+      "learning_rate": 2.328016381557272e-06,
+      "loss": 0.8323,
+      "num_input_tokens_seen": 81967630,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4642578007575302,
+      "grad_norm": 2.322639889941078,
+      "learning_rate": 2.3272479331224522e-06,
+      "loss": 0.7583,
+      "num_input_tokens_seen": 81984780,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4643780436481693,
+      "grad_norm": 2.0956056372655714,
+      "learning_rate": 2.3264794350443813e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 82006595,
+      "step": 3862
+    },
+    {
+      "epoch": 0.46449828653880837,
+      "grad_norm": 2.1145336423211387,
+      "learning_rate": 2.32571088743964e-06,
+      "loss": 0.7807,
+      "num_input_tokens_seen": 82027410,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4646185294294475,
+      "grad_norm": 2.543642922154108,
+      "learning_rate": 2.3249422904248152e-06,
+      "loss": 0.74,
+      "num_input_tokens_seen": 82045565,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4647387723200866,
+      "grad_norm": 2.0808119822700957,
+      "learning_rate": 2.324173644116504e-06,
+      "loss": 0.8646,
+      "num_input_tokens_seen": 82068135,
+      "step": 3865
+    },
+    {
+      "epoch": 0.46485901521072565,
+      "grad_norm": 1.877704267265373,
+      "learning_rate": 2.3234049486313083e-06,
+      "loss": 0.8099,
+      "num_input_tokens_seen": 82089305,
+      "step": 3866
+    },
+    {
+      "epoch": 0.46497925810136476,
+      "grad_norm": 2.090849298199001,
+      "learning_rate": 2.322636204085839e-06,
+      "loss": 0.7633,
+      "num_input_tokens_seen": 82109095,
+      "step": 3867
+    },
+    {
+      "epoch": 0.46509950099200387,
+      "grad_norm": 2.2182248601223264,
+      "learning_rate": 2.3218674105967143e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 82127080,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4652197438826429,
+      "grad_norm": 1.882616032141527,
+      "learning_rate": 2.3210985682805593e-06,
+      "loss": 0.8291,
+      "num_input_tokens_seen": 82148580,
+      "step": 3869
+    },
+    {
+      "epoch": 0.46533998677328203,
+      "grad_norm": 2.658482988574435,
+      "learning_rate": 2.320329677254007e-06,
+      "loss": 0.6847,
+      "num_input_tokens_seen": 82165630,
+      "step": 3870
+    },
+    {
+      "epoch": 0.46546022966392114,
+      "grad_norm": 2.9410593200771435,
+      "learning_rate": 2.319560737633697e-06,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 82184070,
+      "step": 3871
+    },
+    {
+      "epoch": 0.4655804725545602,
+      "grad_norm": 1.579394409184122,
+      "learning_rate": 2.3187917495362775e-06,
+      "loss": 0.6788,
+      "num_input_tokens_seen": 82208200,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4657007154451993,
+      "grad_norm": 2.614772487432404,
+      "learning_rate": 2.318022713078403e-06,
+      "loss": 0.7564,
+      "num_input_tokens_seen": 82222500,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4658209583358384,
+      "grad_norm": 4.3244900004509175,
+      "learning_rate": 2.3172536283767354e-06,
+      "loss": 0.8414,
+      "num_input_tokens_seen": 82235980,
+      "step": 3874
+    },
+    {
+      "epoch": 0.4659412012264775,
+      "grad_norm": 2.197490117156266,
+      "learning_rate": 2.3164844955479447e-06,
+      "loss": 0.8109,
+      "num_input_tokens_seen": 82251510,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4660614441171166,
+      "grad_norm": 2.291343684695357,
+      "learning_rate": 2.315715314708708e-06,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 82273120,
+      "step": 3876
+    },
+    {
+      "epoch": 0.46618168700775564,
+      "grad_norm": 1.8805075195723835,
+      "learning_rate": 2.314946085975709e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 82291820,
+      "step": 3877
+    },
+    {
+      "epoch": 0.46630192989839475,
+      "grad_norm": 1.816090353056169,
+      "learning_rate": 2.3141768094656393e-06,
+      "loss": 0.8214,
+      "num_input_tokens_seen": 82310115,
+      "step": 3878
+    },
+    {
+      "epoch": 0.46642217278903386,
+      "grad_norm": 2.4714507697822126,
+      "learning_rate": 2.313407485295197e-06,
+      "loss": 0.8162,
+      "num_input_tokens_seen": 82326425,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4665424156796729,
+      "grad_norm": 1.730216099059028,
+      "learning_rate": 2.312638113581088e-06,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 82346630,
+      "step": 3880
+    },
+    {
+      "epoch": 0.46666265857031203,
+      "grad_norm": 3.3854740760973914,
+      "learning_rate": 2.311868694440027e-06,
+      "loss": 0.7803,
+      "num_input_tokens_seen": 82360770,
+      "step": 3881
+    },
+    {
+      "epoch": 0.46678290146095114,
+      "grad_norm": 0.7424071643105359,
+      "learning_rate": 2.311099227988732e-06,
+      "loss": 0.6427,
+      "num_input_tokens_seen": 82432415,
+      "step": 3882
+    },
+    {
+      "epoch": 0.4669031443515902,
+      "grad_norm": 2.999664269308277,
+      "learning_rate": 2.310329714343932e-06,
+      "loss": 0.8409,
+      "num_input_tokens_seen": 82448285,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4670233872422293,
+      "grad_norm": 2.047351590002029,
+      "learning_rate": 2.3095601536223605e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 82464915,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4671436301328684,
+      "grad_norm": 2.5253079723960954,
+      "learning_rate": 2.3087905459407607e-06,
+      "loss": 0.7418,
+      "num_input_tokens_seen": 82483575,
+      "step": 3885
+    },
+    {
+      "epoch": 0.46726387302350747,
+      "grad_norm": 0.8625044948753109,
+      "learning_rate": 2.3080208914158795e-06,
+      "loss": 0.6787,
+      "num_input_tokens_seen": 82546295,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4673841159141466,
+      "grad_norm": 2.1081035558333787,
+      "learning_rate": 2.3072511901644753e-06,
+      "loss": 0.7083,
+      "num_input_tokens_seen": 82565085,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4675043588047857,
+      "grad_norm": 2.451618076372208,
+      "learning_rate": 2.306481442303308e-06,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 82584380,
+      "step": 3888
+    },
+    {
+      "epoch": 0.46762460169542475,
+      "grad_norm": 1.7214738650356083,
+      "learning_rate": 2.3057116479491515e-06,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 82603510,
+      "step": 3889
+    },
+    {
+      "epoch": 0.46774484458606386,
+      "grad_norm": 3.3682413642582505,
+      "learning_rate": 2.30494180721878e-06,
+      "loss": 0.7644,
+      "num_input_tokens_seen": 82620570,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4678650874767029,
+      "grad_norm": 1.8958691148341253,
+      "learning_rate": 2.3041719202289794e-06,
+      "loss": 0.8897,
+      "num_input_tokens_seen": 82636465,
+      "step": 3891
+    },
+    {
+      "epoch": 0.467985330367342,
+      "grad_norm": 1.725798958955176,
+      "learning_rate": 2.30340198709654e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 82656020,
+      "step": 3892
+    },
+    {
+      "epoch": 0.46810557325798113,
+      "grad_norm": 2.623375700332464,
+      "learning_rate": 2.3026320079382605e-06,
+      "loss": 0.7489,
+      "num_input_tokens_seen": 82672675,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4682258161486202,
+      "grad_norm": 1.8846448949013452,
+      "learning_rate": 2.3018619828709454e-06,
+      "loss": 0.7676,
+      "num_input_tokens_seen": 82693935,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4683460590392593,
+      "grad_norm": 2.3363588650319977,
+      "learning_rate": 2.3010919120114084e-06,
+      "loss": 0.811,
+      "num_input_tokens_seen": 82710185,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4684663019298984,
+      "grad_norm": 2.431443103580903,
+      "learning_rate": 2.3003217954764672e-06,
+      "loss": 0.654,
+      "num_input_tokens_seen": 82724610,
+      "step": 3896
+    },
+    {
+      "epoch": 0.46858654482053747,
+      "grad_norm": 2.08968901187863,
+      "learning_rate": 2.299551633382949e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 82744640,
+      "step": 3897
+    },
+    {
+      "epoch": 0.4687067877111766,
+      "grad_norm": 2.8573269069951373,
+      "learning_rate": 2.298781425847685e-06,
+      "loss": 0.8466,
+      "num_input_tokens_seen": 82762160,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4688270306018157,
+      "grad_norm": 2.6821777141700838,
+      "learning_rate": 2.2980111729875173e-06,
+      "loss": 0.6679,
+      "num_input_tokens_seen": 82778130,
+      "step": 3899
+    },
+    {
+      "epoch": 0.46894727349245474,
+      "grad_norm": 1.720827964546958,
+      "learning_rate": 2.2972408749192917e-06,
+      "loss": 0.8213,
+      "num_input_tokens_seen": 82795580,
+      "step": 3900
+    },
+    {
+      "epoch": 0.46906751638309385,
+      "grad_norm": 1.8366391718665733,
+      "learning_rate": 2.2964705317598613e-06,
+      "loss": 0.6695,
+      "num_input_tokens_seen": 82813400,
+      "step": 3901
+    },
+    {
+      "epoch": 0.46918775927373296,
+      "grad_norm": 2.317822908815742,
+      "learning_rate": 2.2957001436260866e-06,
+      "loss": 0.7948,
+      "num_input_tokens_seen": 82830180,
+      "step": 3902
+    },
+    {
+      "epoch": 0.469308002164372,
+      "grad_norm": 2.344160572719799,
+      "learning_rate": 2.294929710634836e-06,
+      "loss": 0.7182,
+      "num_input_tokens_seen": 82847990,
+      "step": 3903
+    },
+    {
+      "epoch": 0.46942824505501113,
+      "grad_norm": 2.3645641591555546,
+      "learning_rate": 2.294159232902982e-06,
+      "loss": 0.6058,
+      "num_input_tokens_seen": 82868815,
+      "step": 3904
+    },
+    {
+      "epoch": 0.46954848794565024,
+      "grad_norm": 1.9159592729954404,
+      "learning_rate": 2.2933887105474067e-06,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 82886710,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4696687308362893,
+      "grad_norm": 3.244506162913963,
+      "learning_rate": 2.2926181436849974e-06,
+      "loss": 0.8062,
+      "num_input_tokens_seen": 82905785,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4697889737269284,
+      "grad_norm": 1.803376994550711,
+      "learning_rate": 2.291847532432648e-06,
+      "loss": 0.7256,
+      "num_input_tokens_seen": 82925225,
+      "step": 3907
+    },
+    {
+      "epoch": 0.46990921661756746,
+      "grad_norm": 2.418841835450677,
+      "learning_rate": 2.2910768769072603e-06,
+      "loss": 0.898,
+      "num_input_tokens_seen": 82943725,
+      "step": 3908
+    },
+    {
+      "epoch": 0.47002945950820657,
+      "grad_norm": 2.279320742691122,
+      "learning_rate": 2.2903061772257417e-06,
+      "loss": 0.756,
+      "num_input_tokens_seen": 82961430,
+      "step": 3909
+    },
+    {
+      "epoch": 0.4701497023988457,
+      "grad_norm": 1.7397273815456287,
+      "learning_rate": 2.289535433505007e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 82982505,
+      "step": 3910
+    },
+    {
+      "epoch": 0.47026994528948474,
+      "grad_norm": 5.749203446013382,
+      "learning_rate": 2.2887646458619767e-06,
+      "loss": 0.6271,
+      "num_input_tokens_seen": 83003590,
+      "step": 3911
+    },
+    {
+      "epoch": 0.47039018818012385,
+      "grad_norm": 2.194035243044513,
+      "learning_rate": 2.2879938144135792e-06,
+      "loss": 0.7645,
+      "num_input_tokens_seen": 83019415,
+      "step": 3912
+    },
+    {
+      "epoch": 0.47051043107076296,
+      "grad_norm": 2.3539811010527805,
+      "learning_rate": 2.2872229392767496e-06,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 83039240,
+      "step": 3913
+    },
+    {
+      "epoch": 0.470630673961402,
+      "grad_norm": 1.5649595653330772,
+      "learning_rate": 2.286452020568428e-06,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 83057035,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4707509168520411,
+      "grad_norm": 2.039568447136854,
+      "learning_rate": 2.2856810584055637e-06,
+      "loss": 0.7339,
+      "num_input_tokens_seen": 83074290,
+      "step": 3915
+    },
+    {
+      "epoch": 0.47087115974268023,
+      "grad_norm": 3.0708734225523733,
+      "learning_rate": 2.2849100529051085e-06,
+      "loss": 0.676,
+      "num_input_tokens_seen": 83100945,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4709914026333193,
+      "grad_norm": 2.8641044071739445,
+      "learning_rate": 2.284139004184026e-06,
+      "loss": 0.7961,
+      "num_input_tokens_seen": 83117895,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4711116455239584,
+      "grad_norm": 2.5565827131555814,
+      "learning_rate": 2.2833679123592814e-06,
+      "loss": 0.7364,
+      "num_input_tokens_seen": 83134875,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4712318884145975,
+      "grad_norm": 2.3514728520517383,
+      "learning_rate": 2.2825967775478508e-06,
+      "loss": 0.6414,
+      "num_input_tokens_seen": 83155695,
+      "step": 3919
+    },
+    {
+      "epoch": 0.47135213130523657,
+      "grad_norm": 2.805163767773821,
+      "learning_rate": 2.281825599866713e-06,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 83173925,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4714723741958757,
+      "grad_norm": 1.7072142845197524,
+      "learning_rate": 2.281054379432856e-06,
+      "loss": 0.7858,
+      "num_input_tokens_seen": 83192680,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4715926170865148,
+      "grad_norm": 1.7755542004236387,
+      "learning_rate": 2.2802831163632735e-06,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 83211120,
+      "step": 3922
+    },
+    {
+      "epoch": 0.47171285997715384,
+      "grad_norm": 1.684280722739875,
+      "learning_rate": 2.279511810774965e-06,
+      "loss": 0.7364,
+      "num_input_tokens_seen": 83232370,
+      "step": 3923
+    },
+    {
+      "epoch": 0.47183310286779295,
+      "grad_norm": 2.210677256333796,
+      "learning_rate": 2.2787404627849364e-06,
+      "loss": 0.7129,
+      "num_input_tokens_seen": 83251300,
+      "step": 3924
+    },
+    {
+      "epoch": 0.471953345758432,
+      "grad_norm": 2.431902189779806,
+      "learning_rate": 2.277969072510202e-06,
+      "loss": 0.7855,
+      "num_input_tokens_seen": 83270000,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4720735886490711,
+      "grad_norm": 1.6858464378133364,
+      "learning_rate": 2.27719764006778e-06,
+      "loss": 0.8079,
+      "num_input_tokens_seen": 83288550,
+      "step": 3926
+    },
+    {
+      "epoch": 0.47219383153971023,
+      "grad_norm": 1.8263414488137624,
+      "learning_rate": 2.2764261655746965e-06,
+      "loss": 0.7809,
+      "num_input_tokens_seen": 83305765,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4723140744303493,
+      "grad_norm": 1.6280657967675856,
+      "learning_rate": 2.2756546491479832e-06,
+      "loss": 0.7526,
+      "num_input_tokens_seen": 83326400,
+      "step": 3928
+    },
+    {
+      "epoch": 0.4724343173209884,
+      "grad_norm": 2.6485678907979313,
+      "learning_rate": 2.2748830909046793e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 83343885,
+      "step": 3929
+    },
+    {
+      "epoch": 0.4725545602116275,
+      "grad_norm": 2.170293108173598,
+      "learning_rate": 2.2741114909618283e-06,
+      "loss": 0.6688,
+      "num_input_tokens_seen": 83359500,
+      "step": 3930
+    },
+    {
+      "epoch": 0.47267480310226656,
+      "grad_norm": 1.954242188692323,
+      "learning_rate": 2.2733398494364828e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 83378465,
+      "step": 3931
+    },
+    {
+      "epoch": 0.47279504599290567,
+      "grad_norm": 2.8175475748822705,
+      "learning_rate": 2.2725681664456986e-06,
+      "loss": 0.8427,
+      "num_input_tokens_seen": 83396750,
+      "step": 3932
+    },
+    {
+      "epoch": 0.4729152888835448,
+      "grad_norm": 3.3686721436493974,
+      "learning_rate": 2.271796442106541e-06,
+      "loss": 0.65,
+      "num_input_tokens_seen": 83415825,
+      "step": 3933
+    },
+    {
+      "epoch": 0.47303553177418384,
+      "grad_norm": 0.8263059994767797,
+      "learning_rate": 2.2710246765360788e-06,
+      "loss": 0.6081,
+      "num_input_tokens_seen": 83475805,
+      "step": 3934
+    },
+    {
+      "epoch": 0.47315577466482295,
+      "grad_norm": 2.207122402478002,
+      "learning_rate": 2.2702528698513894e-06,
+      "loss": 0.7347,
+      "num_input_tokens_seen": 83496650,
+      "step": 3935
+    },
+    {
+      "epoch": 0.47327601755546206,
+      "grad_norm": 3.0425235715313215,
+      "learning_rate": 2.269481022169554e-06,
+      "loss": 0.7815,
+      "num_input_tokens_seen": 83514965,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4733962604461011,
+      "grad_norm": 2.148851477634378,
+      "learning_rate": 2.2687091336076614e-06,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 83534025,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4735165033367402,
+      "grad_norm": 2.088015704959444,
+      "learning_rate": 2.267937204282807e-06,
+      "loss": 0.793,
+      "num_input_tokens_seen": 83550885,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4736367462273793,
+      "grad_norm": 3.482601147055524,
+      "learning_rate": 2.2671652343120926e-06,
+      "loss": 0.7862,
+      "num_input_tokens_seen": 83571080,
+      "step": 3939
+    },
+    {
+      "epoch": 0.4737569891180184,
+      "grad_norm": 1.9590040524768826,
+      "learning_rate": 2.2663932238126236e-06,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 83589360,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4738772320086575,
+      "grad_norm": 2.177625535220625,
+      "learning_rate": 2.265621172901515e-06,
+      "loss": 0.7991,
+      "num_input_tokens_seen": 83612195,
+      "step": 3941
+    },
+    {
+      "epoch": 0.47399747489929656,
+      "grad_norm": 10.273132087407363,
+      "learning_rate": 2.264849081695885e-06,
+      "loss": 0.7148,
+      "num_input_tokens_seen": 83632910,
+      "step": 3942
+    },
+    {
+      "epoch": 0.47411771778993567,
+      "grad_norm": 2.40031148998413,
+      "learning_rate": 2.2640769503128606e-06,
+      "loss": 0.724,
+      "num_input_tokens_seen": 83651440,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4742379606805748,
+      "grad_norm": 1.978315041758226,
+      "learning_rate": 2.2633047788695727e-06,
+      "loss": 0.8163,
+      "num_input_tokens_seen": 83671465,
+      "step": 3944
+    },
+    {
+      "epoch": 0.47435820357121383,
+      "grad_norm": 4.773291690080734,
+      "learning_rate": 2.262532567483159e-06,
+      "loss": 0.637,
+      "num_input_tokens_seen": 83689745,
+      "step": 3945
+    },
+    {
+      "epoch": 0.47447844646185294,
+      "grad_norm": 2.034375632627942,
+      "learning_rate": 2.2617603162707635e-06,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 83709875,
+      "step": 3946
+    },
+    {
+      "epoch": 0.47459868935249205,
+      "grad_norm": 2.6294339345118707,
+      "learning_rate": 2.2609880253495363e-06,
+      "loss": 0.8101,
+      "num_input_tokens_seen": 83729230,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4747189322431311,
+      "grad_norm": 2.0270839386020802,
+      "learning_rate": 2.2602156948366326e-06,
+      "loss": 0.861,
+      "num_input_tokens_seen": 83748125,
+      "step": 3948
+    },
+    {
+      "epoch": 0.4748391751337702,
+      "grad_norm": 2.282539378260951,
+      "learning_rate": 2.2594433248492157e-06,
+      "loss": 0.642,
+      "num_input_tokens_seen": 83766820,
+      "step": 3949
+    },
+    {
+      "epoch": 0.47495941802440933,
+      "grad_norm": 6.6982016438128404,
+      "learning_rate": 2.2586709155044527e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 83787140,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4750796609150484,
+      "grad_norm": 1.758180981032041,
+      "learning_rate": 2.257898466919517e-06,
+      "loss": 0.7524,
+      "num_input_tokens_seen": 83807825,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4751999038056875,
+      "grad_norm": 2.0325777523517807,
+      "learning_rate": 2.2571259792115887e-06,
+      "loss": 0.6643,
+      "num_input_tokens_seen": 83828765,
+      "step": 3952
+    },
+    {
+      "epoch": 0.4753201466963266,
+      "grad_norm": 1.9718263332115764,
+      "learning_rate": 2.2563534524978544e-06,
+      "loss": 0.7886,
+      "num_input_tokens_seen": 83845955,
+      "step": 3953
+    },
+    {
+      "epoch": 0.47544038958696566,
+      "grad_norm": 2.0720144051374256,
+      "learning_rate": 2.255580886895505e-06,
+      "loss": 0.7177,
+      "num_input_tokens_seen": 83867805,
+      "step": 3954
+    },
+    {
+      "epoch": 0.47556063247760477,
+      "grad_norm": 2.3896432691769074,
+      "learning_rate": 2.254808282521738e-06,
+      "loss": 0.724,
+      "num_input_tokens_seen": 83886275,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4756808753682438,
+      "grad_norm": 1.8772941262218292,
+      "learning_rate": 2.2540356394937573e-06,
+      "loss": 0.804,
+      "num_input_tokens_seen": 83904695,
+      "step": 3956
+    },
+    {
+      "epoch": 0.47580111825888294,
+      "grad_norm": 2.68324119875583,
+      "learning_rate": 2.253262957928772e-06,
+      "loss": 0.8334,
+      "num_input_tokens_seen": 83921300,
+      "step": 3957
+    },
+    {
+      "epoch": 0.47592136114952205,
+      "grad_norm": 2.0472759145508306,
+      "learning_rate": 2.2524902379439976e-06,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 83939690,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4760416040401611,
+      "grad_norm": 0.7756131613398751,
+      "learning_rate": 2.251717479656655e-06,
+      "loss": 0.6541,
+      "num_input_tokens_seen": 84004205,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4761618469308002,
+      "grad_norm": 1.8651759657432803,
+      "learning_rate": 2.2509446831839704e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 84023365,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4762820898214393,
+      "grad_norm": 7.514273013188539,
+      "learning_rate": 2.250171848643177e-06,
+      "loss": 0.8097,
+      "num_input_tokens_seen": 84040375,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4764023327120784,
+      "grad_norm": 2.5101103493019044,
+      "learning_rate": 2.249398976151513e-06,
+      "loss": 0.8695,
+      "num_input_tokens_seen": 84057645,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4765225756027175,
+      "grad_norm": 3.5166128883694507,
+      "learning_rate": 2.248626065826223e-06,
+      "loss": 0.7886,
+      "num_input_tokens_seen": 84075570,
+      "step": 3963
+    },
+    {
+      "epoch": 0.4766428184933566,
+      "grad_norm": 0.7760054816626786,
+      "learning_rate": 2.247853117784556e-06,
+      "loss": 0.6413,
+      "num_input_tokens_seen": 84136285,
+      "step": 3964
+    },
+    {
+      "epoch": 0.47676306138399566,
+      "grad_norm": 1.9389613937702037,
+      "learning_rate": 2.2470801321437686e-06,
+      "loss": 0.8467,
+      "num_input_tokens_seen": 84158360,
+      "step": 3965
+    },
+    {
+      "epoch": 0.47688330427463477,
+      "grad_norm": 3.4526233067465695,
+      "learning_rate": 2.246307109021121e-06,
+      "loss": 0.6928,
+      "num_input_tokens_seen": 84175485,
+      "step": 3966
+    },
+    {
+      "epoch": 0.4770035471652739,
+      "grad_norm": 1.9243364395237526,
+      "learning_rate": 2.2455340485338817e-06,
+      "loss": 0.8186,
+      "num_input_tokens_seen": 84192840,
+      "step": 3967
+    },
+    {
+      "epoch": 0.47712379005591293,
+      "grad_norm": 2.3013075010670123,
+      "learning_rate": 2.244760950799322e-06,
+      "loss": 0.6767,
+      "num_input_tokens_seen": 84210830,
+      "step": 3968
+    },
+    {
+      "epoch": 0.47724403294655204,
+      "grad_norm": 2.050949645339831,
+      "learning_rate": 2.2439878159347203e-06,
+      "loss": 0.7237,
+      "num_input_tokens_seen": 84229975,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4773642758371911,
+      "grad_norm": 0.8479051670346899,
+      "learning_rate": 2.2432146440573612e-06,
+      "loss": 0.6274,
+      "num_input_tokens_seen": 84295655,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4774845187278302,
+      "grad_norm": 1.8499564084366906,
+      "learning_rate": 2.242441435284534e-06,
+      "loss": 0.6591,
+      "num_input_tokens_seen": 84314250,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4776047616184693,
+      "grad_norm": 2.4479667811541863,
+      "learning_rate": 2.2416681897335337e-06,
+      "loss": 0.85,
+      "num_input_tokens_seen": 84332120,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4777250045091084,
+      "grad_norm": 2.389861872400696,
+      "learning_rate": 2.2408949075216616e-06,
+      "loss": 0.6658,
+      "num_input_tokens_seen": 84350920,
+      "step": 3973
+    },
+    {
+      "epoch": 0.4778452473997475,
+      "grad_norm": 2.154008097914997,
+      "learning_rate": 2.240121588766223e-06,
+      "loss": 0.6394,
+      "num_input_tokens_seen": 84370690,
+      "step": 3974
+    },
+    {
+      "epoch": 0.4779654902903866,
+      "grad_norm": 2.9392093162559845,
+      "learning_rate": 2.239348233584531e-06,
+      "loss": 0.7008,
+      "num_input_tokens_seen": 84391265,
+      "step": 3975
+    },
+    {
+      "epoch": 0.47808573318102565,
+      "grad_norm": 2.0706680432012634,
+      "learning_rate": 2.238574842093901e-06,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 84410180,
+      "step": 3976
+    },
+    {
+      "epoch": 0.47820597607166476,
+      "grad_norm": 2.1677767253567906,
+      "learning_rate": 2.2378014144116583e-06,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 84428710,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4783262189623039,
+      "grad_norm": 2.055705001606301,
+      "learning_rate": 2.2370279506551295e-06,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 84448010,
+      "step": 3978
+    },
+    {
+      "epoch": 0.47844646185294293,
+      "grad_norm": 0.9957046896111164,
+      "learning_rate": 2.2362544509416493e-06,
+      "loss": 0.7057,
+      "num_input_tokens_seen": 84499845,
+      "step": 3979
+    },
+    {
+      "epoch": 0.47856670474358204,
+      "grad_norm": 3.8507252777473995,
+      "learning_rate": 2.2354809153885572e-06,
+      "loss": 0.8246,
+      "num_input_tokens_seen": 84516635,
+      "step": 3980
+    },
+    {
+      "epoch": 0.47868694763422115,
+      "grad_norm": 1.9571372408487326,
+      "learning_rate": 2.234707344113197e-06,
+      "loss": 0.8171,
+      "num_input_tokens_seen": 84534450,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4788071905248602,
+      "grad_norm": 1.7771968260322124,
+      "learning_rate": 2.233933737232919e-06,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 84551950,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4789274334154993,
+      "grad_norm": 2.02263000661705,
+      "learning_rate": 2.2331600948650793e-06,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 84571815,
+      "step": 3983
+    },
+    {
+      "epoch": 0.4790476763061384,
+      "grad_norm": 1.6973146922126954,
+      "learning_rate": 2.2323864171270386e-06,
+      "loss": 0.7969,
+      "num_input_tokens_seen": 84592805,
+      "step": 3984
+    },
+    {
+      "epoch": 0.4791679191967775,
+      "grad_norm": 2.020238730697779,
+      "learning_rate": 2.231612704136164e-06,
+      "loss": 0.7202,
+      "num_input_tokens_seen": 84612895,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4792881620874166,
+      "grad_norm": 2.171744126260886,
+      "learning_rate": 2.230838956009825e-06,
+      "loss": 0.7457,
+      "num_input_tokens_seen": 84628990,
+      "step": 3986
+    },
+    {
+      "epoch": 0.47940840497805565,
+      "grad_norm": 4.12476011126497,
+      "learning_rate": 2.2300651728654003e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 84643970,
+      "step": 3987
+    },
+    {
+      "epoch": 0.47952864786869476,
+      "grad_norm": 0.7654208974200336,
+      "learning_rate": 2.229291354820272e-06,
+      "loss": 0.6326,
+      "num_input_tokens_seen": 84704700,
+      "step": 3988
+    },
+    {
+      "epoch": 0.47964889075933387,
+      "grad_norm": 2.0509428912487686,
+      "learning_rate": 2.228517501991828e-06,
+      "loss": 0.7646,
+      "num_input_tokens_seen": 84723220,
+      "step": 3989
+    },
+    {
+      "epoch": 0.4797691336499729,
+      "grad_norm": 0.8928135228727033,
+      "learning_rate": 2.22774361449746e-06,
+      "loss": 0.6406,
+      "num_input_tokens_seen": 84779420,
+      "step": 3990
+    },
+    {
+      "epoch": 0.47988937654061203,
+      "grad_norm": 3.116934679164819,
+      "learning_rate": 2.2269696924545668e-06,
+      "loss": 0.7034,
+      "num_input_tokens_seen": 84796970,
+      "step": 3991
+    },
+    {
+      "epoch": 0.48000961943125114,
+      "grad_norm": 3.4014357969814744,
+      "learning_rate": 2.226195735980552e-06,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 84813925,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4801298623218902,
+      "grad_norm": 1.9242909249967262,
+      "learning_rate": 2.225421745192823e-06,
+      "loss": 0.737,
+      "num_input_tokens_seen": 84833800,
+      "step": 3993
+    },
+    {
+      "epoch": 0.4802501052125293,
+      "grad_norm": 2.5393244345059065,
+      "learning_rate": 2.2246477202087955e-06,
+      "loss": 0.7824,
+      "num_input_tokens_seen": 84854200,
+      "step": 3994
+    },
+    {
+      "epoch": 0.4803703481031684,
+      "grad_norm": 1.7164779414116171,
+      "learning_rate": 2.2238736611458875e-06,
+      "loss": 0.8264,
+      "num_input_tokens_seen": 84873975,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4804905909938075,
+      "grad_norm": 1.6664675581919821,
+      "learning_rate": 2.2230995681215226e-06,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 84893220,
+      "step": 3996
+    },
+    {
+      "epoch": 0.4806108338844466,
+      "grad_norm": 2.709812950415109,
+      "learning_rate": 2.2223254412531305e-06,
+      "loss": 0.7865,
+      "num_input_tokens_seen": 84910310,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4807310767750857,
+      "grad_norm": 1.901966989291102,
+      "learning_rate": 2.2215512806581458e-06,
+      "loss": 0.8147,
+      "num_input_tokens_seen": 84929090,
+      "step": 3998
+    },
+    {
+      "epoch": 0.48085131966572475,
+      "grad_norm": 1.9683802677453521,
+      "learning_rate": 2.2207770864540085e-06,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 84947785,
+      "step": 3999
+    },
+    {
+      "epoch": 0.48097156255636386,
+      "grad_norm": 2.193141402276056,
+      "learning_rate": 2.2200028587581617e-06,
+      "loss": 0.7184,
+      "num_input_tokens_seen": 84965495,
+      "step": 4000
+    },
+    {
+      "epoch": 0.481091805447003,
+      "grad_norm": 0.8616572911286942,
+      "learning_rate": 2.2192285976880573e-06,
+      "loss": 0.6231,
+      "num_input_tokens_seen": 85029470,
+      "step": 4001
+    },
+    {
+      "epoch": 0.48121204833764203,
+      "grad_norm": 2.1288868970971944,
+      "learning_rate": 2.2184543033611485e-06,
+      "loss": 0.7998,
+      "num_input_tokens_seen": 85050270,
+      "step": 4002
+    },
+    {
+      "epoch": 0.48133229122828114,
+      "grad_norm": 2.3965664991241313,
+      "learning_rate": 2.2176799758948957e-06,
+      "loss": 0.822,
+      "num_input_tokens_seen": 85070150,
+      "step": 4003
+    },
+    {
+      "epoch": 0.4814525341189202,
+      "grad_norm": 3.3311139437323876,
+      "learning_rate": 2.2169056154067635e-06,
+      "loss": 0.7221,
+      "num_input_tokens_seen": 85093790,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4815727770095593,
+      "grad_norm": 2.7847100909440914,
+      "learning_rate": 2.216131222014222e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 85111585,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4816930199001984,
+      "grad_norm": 2.330108599202827,
+      "learning_rate": 2.2153567958347455e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 85127515,
+      "step": 4006
+    },
+    {
+      "epoch": 0.48181326279083747,
+      "grad_norm": 3.725724356300552,
+      "learning_rate": 2.214582336985815e-06,
+      "loss": 0.7929,
+      "num_input_tokens_seen": 85145135,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4819335056814766,
+      "grad_norm": 2.35969221135231,
+      "learning_rate": 2.213807845584914e-06,
+      "loss": 0.6532,
+      "num_input_tokens_seen": 85162850,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4820537485721157,
+      "grad_norm": 2.1874587913433223,
+      "learning_rate": 2.213033321749533e-06,
+      "loss": 0.7905,
+      "num_input_tokens_seen": 85181740,
+      "step": 4009
+    },
+    {
+      "epoch": 0.48217399146275475,
+      "grad_norm": 4.524478683433816,
+      "learning_rate": 2.2122587655971665e-06,
+      "loss": 0.6708,
+      "num_input_tokens_seen": 85196405,
+      "step": 4010
+    },
+    {
+      "epoch": 0.48229423435339386,
+      "grad_norm": 1.7152819901044982,
+      "learning_rate": 2.211484177245314e-06,
+      "loss": 0.6353,
+      "num_input_tokens_seen": 85215715,
+      "step": 4011
+    },
+    {
+      "epoch": 0.48241447724403297,
+      "grad_norm": 2.3709313826938097,
+      "learning_rate": 2.21070955681148e-06,
+      "loss": 0.7279,
+      "num_input_tokens_seen": 85234540,
+      "step": 4012
+    },
+    {
+      "epoch": 0.482534720134672,
+      "grad_norm": 2.1165116851285486,
+      "learning_rate": 2.2099349044131736e-06,
+      "loss": 0.7685,
+      "num_input_tokens_seen": 85255865,
+      "step": 4013
+    },
+    {
+      "epoch": 0.48265496302531113,
+      "grad_norm": 2.1725383969700403,
+      "learning_rate": 2.2091602201679095e-06,
+      "loss": 0.7109,
+      "num_input_tokens_seen": 85275195,
+      "step": 4014
+    },
+    {
+      "epoch": 0.48277520591595025,
+      "grad_norm": 2.4907957430021748,
+      "learning_rate": 2.208385504193206e-06,
+      "loss": 0.8343,
+      "num_input_tokens_seen": 85292415,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4828954488065893,
+      "grad_norm": 2.2364668413140265,
+      "learning_rate": 2.2076107566065873e-06,
+      "loss": 0.8091,
+      "num_input_tokens_seen": 85309920,
+      "step": 4016
+    },
+    {
+      "epoch": 0.4830156916972284,
+      "grad_norm": 2.8206617945332395,
+      "learning_rate": 2.206835977525582e-06,
+      "loss": 0.7474,
+      "num_input_tokens_seen": 85327950,
+      "step": 4017
+    },
+    {
+      "epoch": 0.48313593458786747,
+      "grad_norm": 3.7942707788886505,
+      "learning_rate": 2.206061167067723e-06,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 85345780,
+      "step": 4018
+    },
+    {
+      "epoch": 0.4832561774785066,
+      "grad_norm": 2.543631195380871,
+      "learning_rate": 2.205286325350549e-06,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 85364565,
+      "step": 4019
+    },
+    {
+      "epoch": 0.4833764203691457,
+      "grad_norm": 4.7489850007640815,
+      "learning_rate": 2.2045114524916025e-06,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 85380910,
+      "step": 4020
+    },
+    {
+      "epoch": 0.48349666325978474,
+      "grad_norm": 1.9599063030864785,
+      "learning_rate": 2.2037365486084316e-06,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 85403870,
+      "step": 4021
+    },
+    {
+      "epoch": 0.48361690615042385,
+      "grad_norm": 2.132908766041915,
+      "learning_rate": 2.202961613818588e-06,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 85422590,
+      "step": 4022
+    },
+    {
+      "epoch": 0.48373714904106296,
+      "grad_norm": 2.2826361280887726,
+      "learning_rate": 2.202186648239629e-06,
+      "loss": 0.8243,
+      "num_input_tokens_seen": 85442245,
+      "step": 4023
+    },
+    {
+      "epoch": 0.483857391931702,
+      "grad_norm": 2.144311663967737,
+      "learning_rate": 2.2014116519891166e-06,
+      "loss": 0.7136,
+      "num_input_tokens_seen": 85463945,
+      "step": 4024
+    },
+    {
+      "epoch": 0.48397763482234113,
+      "grad_norm": 1.9919023560553297,
+      "learning_rate": 2.2006366251846167e-06,
+      "loss": 0.778,
+      "num_input_tokens_seen": 85484305,
+      "step": 4025
+    },
+    {
+      "epoch": 0.48409787771298024,
+      "grad_norm": 1.8952301187257234,
+      "learning_rate": 2.1998615679436997e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 85501565,
+      "step": 4026
+    },
+    {
+      "epoch": 0.4842181206036193,
+      "grad_norm": 4.178083829411189,
+      "learning_rate": 2.199086480383942e-06,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 85520660,
+      "step": 4027
+    },
+    {
+      "epoch": 0.4843383634942584,
+      "grad_norm": 3.8547450751253147,
+      "learning_rate": 2.1983113626229234e-06,
+      "loss": 0.6644,
+      "num_input_tokens_seen": 85539630,
+      "step": 4028
+    },
+    {
+      "epoch": 0.4844586063848975,
+      "grad_norm": 2.1334829684527747,
+      "learning_rate": 2.1975362147782293e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 85558545,
+      "step": 4029
+    },
+    {
+      "epoch": 0.48457884927553657,
+      "grad_norm": 0.7764013679281967,
+      "learning_rate": 2.1967610369674476e-06,
+      "loss": 0.5684,
+      "num_input_tokens_seen": 85626230,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4846990921661757,
+      "grad_norm": 3.4520457125485304,
+      "learning_rate": 2.1959858293081743e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 85645085,
+      "step": 4031
+    },
+    {
+      "epoch": 0.4848193350568148,
+      "grad_norm": 2.1264535195242686,
+      "learning_rate": 2.1952105919180056e-06,
+      "loss": 0.7602,
+      "num_input_tokens_seen": 85664060,
+      "step": 4032
+    },
+    {
+      "epoch": 0.48493957794745385,
+      "grad_norm": 2.6980222493951445,
+      "learning_rate": 2.1944353249145456e-06,
+      "loss": 0.6778,
+      "num_input_tokens_seen": 85682890,
+      "step": 4033
+    },
+    {
+      "epoch": 0.48505982083809296,
+      "grad_norm": 1.833618303271678,
+      "learning_rate": 2.193660028415401e-06,
+      "loss": 0.7427,
+      "num_input_tokens_seen": 85703390,
+      "step": 4034
+    },
+    {
+      "epoch": 0.485180063728732,
+      "grad_norm": 1.870785606655225,
+      "learning_rate": 2.192884702538185e-06,
+      "loss": 0.8149,
+      "num_input_tokens_seen": 85723715,
+      "step": 4035
+    },
+    {
+      "epoch": 0.4853003066193711,
+      "grad_norm": 1.9949015439646607,
+      "learning_rate": 2.1921093474005118e-06,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 85743650,
+      "step": 4036
+    },
+    {
+      "epoch": 0.48542054951001024,
+      "grad_norm": 2.0294309289803536,
+      "learning_rate": 2.191333963120004e-06,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 85762350,
+      "step": 4037
+    },
+    {
+      "epoch": 0.4855407924006493,
+      "grad_norm": 2.510524773136468,
+      "learning_rate": 2.190558549814286e-06,
+      "loss": 0.6941,
+      "num_input_tokens_seen": 85782230,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4856610352912884,
+      "grad_norm": 2.5748970904994652,
+      "learning_rate": 2.1897831076009877e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 85801590,
+      "step": 4039
+    },
+    {
+      "epoch": 0.4857812781819275,
+      "grad_norm": 1.8775042628356595,
+      "learning_rate": 2.1890076365977426e-06,
+      "loss": 0.7979,
+      "num_input_tokens_seen": 85821135,
+      "step": 4040
+    },
+    {
+      "epoch": 0.48590152107256657,
+      "grad_norm": 0.8973110757229292,
+      "learning_rate": 2.188232136922189e-06,
+      "loss": 0.5613,
+      "num_input_tokens_seen": 85878975,
+      "step": 4041
+    },
+    {
+      "epoch": 0.4860217639632057,
+      "grad_norm": 1.9957205318414821,
+      "learning_rate": 2.1874566086919704e-06,
+      "loss": 0.7586,
+      "num_input_tokens_seen": 85897570,
+      "step": 4042
+    },
+    {
+      "epoch": 0.4861420068538448,
+      "grad_norm": 2.0863338870701327,
+      "learning_rate": 2.1866810520247334e-06,
+      "loss": 0.8718,
+      "num_input_tokens_seen": 85916160,
+      "step": 4043
+    },
+    {
+      "epoch": 0.48626224974448384,
+      "grad_norm": 2.509750142977052,
+      "learning_rate": 2.1859054670381285e-06,
+      "loss": 0.6482,
+      "num_input_tokens_seen": 85934785,
+      "step": 4044
+    },
+    {
+      "epoch": 0.48638249263512295,
+      "grad_norm": 1.7320036008282027,
+      "learning_rate": 2.1851298538498127e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 85954220,
+      "step": 4045
+    },
+    {
+      "epoch": 0.48650273552576206,
+      "grad_norm": 3.962868497273255,
+      "learning_rate": 2.1843542125774458e-06,
+      "loss": 0.7881,
+      "num_input_tokens_seen": 85974245,
+      "step": 4046
+    },
+    {
+      "epoch": 0.4866229784164011,
+      "grad_norm": 3.3559082693801114,
+      "learning_rate": 2.1835785433386907e-06,
+      "loss": 0.6286,
+      "num_input_tokens_seen": 85992780,
+      "step": 4047
+    },
+    {
+      "epoch": 0.48674322130704023,
+      "grad_norm": 2.1645281281371003,
+      "learning_rate": 2.182802846251216e-06,
+      "loss": 0.648,
+      "num_input_tokens_seen": 86012770,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4868634641976793,
+      "grad_norm": 2.394288272597297,
+      "learning_rate": 2.182027121432696e-06,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 86033115,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4869837070883184,
+      "grad_norm": 3.705333413781921,
+      "learning_rate": 2.1812513690008054e-06,
+      "loss": 0.8106,
+      "num_input_tokens_seen": 86051955,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4871039499789575,
+      "grad_norm": 2.896595064472955,
+      "learning_rate": 2.180475589073227e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 86069375,
+      "step": 4051
+    },
+    {
+      "epoch": 0.48722419286959656,
+      "grad_norm": 1.87295892736607,
+      "learning_rate": 2.1796997817676452e-06,
+      "loss": 0.7319,
+      "num_input_tokens_seen": 86090105,
+      "step": 4052
+    },
+    {
+      "epoch": 0.4873444357602357,
+      "grad_norm": 1.8030410483364006,
+      "learning_rate": 2.1789239472017494e-06,
+      "loss": 0.6706,
+      "num_input_tokens_seen": 86111475,
+      "step": 4053
+    },
+    {
+      "epoch": 0.4874646786508748,
+      "grad_norm": 2.805445121700547,
+      "learning_rate": 2.1781480854932326e-06,
+      "loss": 0.7227,
+      "num_input_tokens_seen": 86130960,
+      "step": 4054
+    },
+    {
+      "epoch": 0.48758492154151384,
+      "grad_norm": 2.1706092976400706,
+      "learning_rate": 2.1773721967597933e-06,
+      "loss": 0.7861,
+      "num_input_tokens_seen": 86149130,
+      "step": 4055
+    },
+    {
+      "epoch": 0.48770516443215295,
+      "grad_norm": 0.9057332053408153,
+      "learning_rate": 2.1765962811191322e-06,
+      "loss": 0.6298,
+      "num_input_tokens_seen": 86203315,
+      "step": 4056
+    },
+    {
+      "epoch": 0.48782540732279206,
+      "grad_norm": 0.9385802643009273,
+      "learning_rate": 2.175820338688956e-06,
+      "loss": 0.684,
+      "num_input_tokens_seen": 86265805,
+      "step": 4057
+    },
+    {
+      "epoch": 0.4879456502134311,
+      "grad_norm": 2.193576747209621,
+      "learning_rate": 2.175044369586974e-06,
+      "loss": 0.836,
+      "num_input_tokens_seen": 86281095,
+      "step": 4058
+    },
+    {
+      "epoch": 0.4880658931040702,
+      "grad_norm": 1.7538532861930534,
+      "learning_rate": 2.174268373930901e-06,
+      "loss": 0.8467,
+      "num_input_tokens_seen": 86298330,
+      "step": 4059
+    },
+    {
+      "epoch": 0.48818613599470934,
+      "grad_norm": 4.719760448090094,
+      "learning_rate": 2.1734923518384537e-06,
+      "loss": 0.7992,
+      "num_input_tokens_seen": 86314655,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4883063788853484,
+      "grad_norm": 2.0171074911475873,
+      "learning_rate": 2.172716303427355e-06,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 86332540,
+      "step": 4061
+    },
+    {
+      "epoch": 0.4884266217759875,
+      "grad_norm": 2.5119551309686465,
+      "learning_rate": 2.17194022881533e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 86348350,
+      "step": 4062
+    },
+    {
+      "epoch": 0.4885468646666266,
+      "grad_norm": 1.7473587416142473,
+      "learning_rate": 2.1711641281201092e-06,
+      "loss": 0.667,
+      "num_input_tokens_seen": 86368000,
+      "step": 4063
+    },
+    {
+      "epoch": 0.48866710755726567,
+      "grad_norm": 2.3072115942313136,
+      "learning_rate": 2.170388001459426e-06,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 86385310,
+      "step": 4064
+    },
+    {
+      "epoch": 0.4887873504479048,
+      "grad_norm": 53.61829120016976,
+      "learning_rate": 2.1696118489510182e-06,
+      "loss": 0.7237,
+      "num_input_tokens_seen": 86405960,
+      "step": 4065
+    },
+    {
+      "epoch": 0.48890759333854383,
+      "grad_norm": 7.055357257253438,
+      "learning_rate": 2.168835670712628e-06,
+      "loss": 0.7189,
+      "num_input_tokens_seen": 86425300,
+      "step": 4066
+    },
+    {
+      "epoch": 0.48902783622918294,
+      "grad_norm": 2.1585127671406954,
+      "learning_rate": 2.168059466862001e-06,
+      "loss": 0.6948,
+      "num_input_tokens_seen": 86443170,
+      "step": 4067
+    },
+    {
+      "epoch": 0.48914807911982205,
+      "grad_norm": 2.407857100025039,
+      "learning_rate": 2.1672832375168867e-06,
+      "loss": 0.8107,
+      "num_input_tokens_seen": 86461165,
+      "step": 4068
+    },
+    {
+      "epoch": 0.4892683220104611,
+      "grad_norm": 2.043004033923676,
+      "learning_rate": 2.1665069827950383e-06,
+      "loss": 0.748,
+      "num_input_tokens_seen": 86478170,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4893885649011002,
+      "grad_norm": 3.653390309101017,
+      "learning_rate": 2.1657307028142126e-06,
+      "loss": 0.8606,
+      "num_input_tokens_seen": 86495430,
+      "step": 4070
+    },
+    {
+      "epoch": 0.48950880779173933,
+      "grad_norm": 2.092922282495773,
+      "learning_rate": 2.164954397692171e-06,
+      "loss": 0.6683,
+      "num_input_tokens_seen": 86514575,
+      "step": 4071
+    },
+    {
+      "epoch": 0.4896290506823784,
+      "grad_norm": 1.101250165544609,
+      "learning_rate": 2.164178067546678e-06,
+      "loss": 0.8276,
+      "num_input_tokens_seen": 86573460,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4897492935730175,
+      "grad_norm": 2.1594961789616023,
+      "learning_rate": 2.163401712495504e-06,
+      "loss": 0.9055,
+      "num_input_tokens_seen": 86590875,
+      "step": 4073
+    },
+    {
+      "epoch": 0.4898695364636566,
+      "grad_norm": 1.8216365870371665,
+      "learning_rate": 2.162625332656419e-06,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 86609545,
+      "step": 4074
+    },
+    {
+      "epoch": 0.48998977935429566,
+      "grad_norm": 2.0084334242948776,
+      "learning_rate": 2.161848928147201e-06,
+      "loss": 0.769,
+      "num_input_tokens_seen": 86629535,
+      "step": 4075
+    },
+    {
+      "epoch": 0.4901100222449348,
+      "grad_norm": 2.162922578281108,
+      "learning_rate": 2.161072499085629e-06,
+      "loss": 0.807,
+      "num_input_tokens_seen": 86648250,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4902302651355739,
+      "grad_norm": 2.3969922174836853,
+      "learning_rate": 2.160296045589487e-06,
+      "loss": 0.8224,
+      "num_input_tokens_seen": 86671430,
+      "step": 4077
+    },
+    {
+      "epoch": 0.49035050802621294,
+      "grad_norm": 1.9511942024447229,
+      "learning_rate": 2.159519567776562e-06,
+      "loss": 0.6905,
+      "num_input_tokens_seen": 86690800,
+      "step": 4078
+    },
+    {
+      "epoch": 0.49047075091685205,
+      "grad_norm": 2.7683633296270123,
+      "learning_rate": 2.158743065764646e-06,
+      "loss": 0.697,
+      "num_input_tokens_seen": 86703955,
+      "step": 4079
+    },
+    {
+      "epoch": 0.4905909938074911,
+      "grad_norm": 2.2170510855719585,
+      "learning_rate": 2.1579665396715326e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 86723315,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4907112366981302,
+      "grad_norm": 3.814077948130425,
+      "learning_rate": 2.157189989615021e-06,
+      "loss": 0.6608,
+      "num_input_tokens_seen": 86741625,
+      "step": 4081
+    },
+    {
+      "epoch": 0.4908314795887693,
+      "grad_norm": 2.184306966192034,
+      "learning_rate": 2.156413415712913e-06,
+      "loss": 0.7445,
+      "num_input_tokens_seen": 86763395,
+      "step": 4082
+    },
+    {
+      "epoch": 0.4909517224794084,
+      "grad_norm": 1.970678818119009,
+      "learning_rate": 2.1556368180830144e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 86784485,
+      "step": 4083
+    },
+    {
+      "epoch": 0.4910719653700475,
+      "grad_norm": 2.44523123345503,
+      "learning_rate": 2.154860196843134e-06,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 86803400,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4911922082606866,
+      "grad_norm": 1.9484428319632539,
+      "learning_rate": 2.154083552111085e-06,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 86822290,
+      "step": 4085
+    },
+    {
+      "epoch": 0.49131245115132566,
+      "grad_norm": 1.9132718309552519,
+      "learning_rate": 2.153306884004683e-06,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 86842275,
+      "step": 4086
+    },
+    {
+      "epoch": 0.49143269404196477,
+      "grad_norm": 2.594276396172624,
+      "learning_rate": 2.152530192641749e-06,
+      "loss": 0.6065,
+      "num_input_tokens_seen": 86856905,
+      "step": 4087
+    },
+    {
+      "epoch": 0.4915529369326039,
+      "grad_norm": 1.8860865038356716,
+      "learning_rate": 2.1517534781401063e-06,
+      "loss": 0.7118,
+      "num_input_tokens_seen": 86874505,
+      "step": 4088
+    },
+    {
+      "epoch": 0.49167317982324293,
+      "grad_norm": 2.3161062018129237,
+      "learning_rate": 2.150976740617581e-06,
+      "loss": 0.687,
+      "num_input_tokens_seen": 86890785,
+      "step": 4089
+    },
+    {
+      "epoch": 0.49179342271388204,
+      "grad_norm": 2.3066504983486418,
+      "learning_rate": 2.1501999801920055e-06,
+      "loss": 0.7204,
+      "num_input_tokens_seen": 86909625,
+      "step": 4090
+    },
+    {
+      "epoch": 0.49191366560452116,
+      "grad_norm": 2.9919550686542973,
+      "learning_rate": 2.1494231969812114e-06,
+      "loss": 0.8104,
+      "num_input_tokens_seen": 86928335,
+      "step": 4091
+    },
+    {
+      "epoch": 0.4920339084951602,
+      "grad_norm": 2.4962750541457406,
+      "learning_rate": 2.1486463911030372e-06,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 86948705,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4921541513857993,
+      "grad_norm": 2.702063246045616,
+      "learning_rate": 2.147869562675324e-06,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 86967395,
+      "step": 4093
+    },
+    {
+      "epoch": 0.49227439427643843,
+      "grad_norm": 15.30632331128531,
+      "learning_rate": 2.147092711815915e-06,
+      "loss": 0.7129,
+      "num_input_tokens_seen": 86986465,
+      "step": 4094
+    },
+    {
+      "epoch": 0.4923946371670775,
+      "grad_norm": 2.7398572650106554,
+      "learning_rate": 2.1463158386426593e-06,
+      "loss": 0.8588,
+      "num_input_tokens_seen": 87003995,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4925148800577166,
+      "grad_norm": 8.738477506568747,
+      "learning_rate": 2.145538943273407e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 87023990,
+      "step": 4096
+    },
+    {
+      "epoch": 0.49263512294835565,
+      "grad_norm": 1.7592507678083682,
+      "learning_rate": 2.144762025826013e-06,
+      "loss": 0.7128,
+      "num_input_tokens_seen": 87042800,
+      "step": 4097
+    },
+    {
+      "epoch": 0.49275536583899476,
+      "grad_norm": 2.176938157652189,
+      "learning_rate": 2.143985086418334e-06,
+      "loss": 0.8634,
+      "num_input_tokens_seen": 87057700,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4928756087296339,
+      "grad_norm": 1.8635329891005548,
+      "learning_rate": 2.1432081251682324e-06,
+      "loss": 0.7652,
+      "num_input_tokens_seen": 87077790,
+      "step": 4099
+    },
+    {
+      "epoch": 0.49299585162027293,
+      "grad_norm": 1.7899470456238205,
+      "learning_rate": 2.142431142193572e-06,
+      "loss": 0.8574,
+      "num_input_tokens_seen": 87095290,
+      "step": 4100
+    },
+    {
+      "epoch": 0.49311609451091204,
+      "grad_norm": 2.8560385404588695,
+      "learning_rate": 2.1416541376122207e-06,
+      "loss": 0.7173,
+      "num_input_tokens_seen": 87115190,
+      "step": 4101
+    },
+    {
+      "epoch": 0.49323633740155115,
+      "grad_norm": 2.022316662730115,
+      "learning_rate": 2.1408771115420488e-06,
+      "loss": 0.7236,
+      "num_input_tokens_seen": 87134770,
+      "step": 4102
+    },
+    {
+      "epoch": 0.4933565802921902,
+      "grad_norm": 2.304175189096094,
+      "learning_rate": 2.140100064100932e-06,
+      "loss": 0.6418,
+      "num_input_tokens_seen": 87150465,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4934768231828293,
+      "grad_norm": 3.1159184717463106,
+      "learning_rate": 2.139322995406746e-06,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 87167820,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4935970660734684,
+      "grad_norm": 3.0053407868676048,
+      "learning_rate": 2.138545905577373e-06,
+      "loss": 0.7994,
+      "num_input_tokens_seen": 87185730,
+      "step": 4105
+    },
+    {
+      "epoch": 0.4937173089641075,
+      "grad_norm": 2.089165838391317,
+      "learning_rate": 2.137768794730696e-06,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 87208900,
+      "step": 4106
+    },
+    {
+      "epoch": 0.4938375518547466,
+      "grad_norm": 2.557553193540661,
+      "learning_rate": 2.1369916629846026e-06,
+      "loss": 0.7975,
+      "num_input_tokens_seen": 87228370,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4939577947453857,
+      "grad_norm": 2.054041484905405,
+      "learning_rate": 2.136214510456982e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 87246545,
+      "step": 4108
+    },
+    {
+      "epoch": 0.49407803763602476,
+      "grad_norm": 1.0363077091313688,
+      "learning_rate": 2.1354373372657296e-06,
+      "loss": 0.7052,
+      "num_input_tokens_seen": 87304705,
+      "step": 4109
+    },
+    {
+      "epoch": 0.49419828052666387,
+      "grad_norm": 1.8600743719006412,
+      "learning_rate": 2.13466014352874e-06,
+      "loss": 0.7018,
+      "num_input_tokens_seen": 87326695,
+      "step": 4110
+    },
+    {
+      "epoch": 0.494318523417303,
+      "grad_norm": 1.9646913734794522,
+      "learning_rate": 2.1338829293639144e-06,
+      "loss": 0.7963,
+      "num_input_tokens_seen": 87346775,
+      "step": 4111
+    },
+    {
+      "epoch": 0.49443876630794203,
+      "grad_norm": 2.6443894618127706,
+      "learning_rate": 2.1331056948891547e-06,
+      "loss": 0.8261,
+      "num_input_tokens_seen": 87363595,
+      "step": 4112
+    },
+    {
+      "epoch": 0.49455900919858115,
+      "grad_norm": 2.6113874791891223,
+      "learning_rate": 2.1323284402223666e-06,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 87379305,
+      "step": 4113
+    },
+    {
+      "epoch": 0.4946792520892202,
+      "grad_norm": 2.227278235725536,
+      "learning_rate": 2.1315511654814597e-06,
+      "loss": 0.8718,
+      "num_input_tokens_seen": 87397435,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4947994949798593,
+      "grad_norm": 1.7866041715286072,
+      "learning_rate": 2.1307738707843456e-06,
+      "loss": 0.7755,
+      "num_input_tokens_seen": 87416820,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4949197378704984,
+      "grad_norm": 2.8598250254887145,
+      "learning_rate": 2.1299965562489385e-06,
+      "loss": 0.6876,
+      "num_input_tokens_seen": 87436345,
+      "step": 4116
+    },
+    {
+      "epoch": 0.4950399807611375,
+      "grad_norm": 1.5440965830243054,
+      "learning_rate": 2.129219221993158e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 87460850,
+      "step": 4117
+    },
+    {
+      "epoch": 0.4951602236517766,
+      "grad_norm": 0.8636435645482484,
+      "learning_rate": 2.128441868134924e-06,
+      "loss": 0.645,
+      "num_input_tokens_seen": 87522505,
+      "step": 4118
+    },
+    {
+      "epoch": 0.4952804665424157,
+      "grad_norm": 2.4562523951896575,
+      "learning_rate": 2.1276644947921606e-06,
+      "loss": 0.8261,
+      "num_input_tokens_seen": 87541140,
+      "step": 4119
+    },
+    {
+      "epoch": 0.49540070943305475,
+      "grad_norm": 2.1970344098413483,
+      "learning_rate": 2.126887102082795e-06,
+      "loss": 0.8229,
+      "num_input_tokens_seen": 87560885,
+      "step": 4120
+    },
+    {
+      "epoch": 0.49552095232369386,
+      "grad_norm": 5.336959057457993,
+      "learning_rate": 2.126109690124757e-06,
+      "loss": 0.6976,
+      "num_input_tokens_seen": 87581420,
+      "step": 4121
+    },
+    {
+      "epoch": 0.495641195214333,
+      "grad_norm": 1.939125325047874,
+      "learning_rate": 2.1253322590359786e-06,
+      "loss": 0.7107,
+      "num_input_tokens_seen": 87600475,
+      "step": 4122
+    },
+    {
+      "epoch": 0.49576143810497203,
+      "grad_norm": 2.6349657660950188,
+      "learning_rate": 2.124554808934397e-06,
+      "loss": 0.7356,
+      "num_input_tokens_seen": 87620775,
+      "step": 4123
+    },
+    {
+      "epoch": 0.49588168099561114,
+      "grad_norm": 2.454464280317344,
+      "learning_rate": 2.123777339937949e-06,
+      "loss": 0.7294,
+      "num_input_tokens_seen": 87641460,
+      "step": 4124
+    },
+    {
+      "epoch": 0.49600192388625025,
+      "grad_norm": 1.9384378019179296,
+      "learning_rate": 2.122999852164578e-06,
+      "loss": 0.866,
+      "num_input_tokens_seen": 87661800,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4961221667768893,
+      "grad_norm": 2.4404306942510847,
+      "learning_rate": 2.122222345732227e-06,
+      "loss": 0.5704,
+      "num_input_tokens_seen": 87681435,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4962424096675284,
+      "grad_norm": 2.6704386941591354,
+      "learning_rate": 2.1214448207588434e-06,
+      "loss": 0.828,
+      "num_input_tokens_seen": 87699795,
+      "step": 4127
+    },
+    {
+      "epoch": 0.49636265255816747,
+      "grad_norm": 3.7554765886963457,
+      "learning_rate": 2.120667277362376e-06,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 87718230,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4964828954488066,
+      "grad_norm": 2.3844792465803053,
+      "learning_rate": 2.1198897156607796e-06,
+      "loss": 0.8418,
+      "num_input_tokens_seen": 87735305,
+      "step": 4129
+    },
+    {
+      "epoch": 0.4966031383394457,
+      "grad_norm": 4.067079400434869,
+      "learning_rate": 2.119112135772008e-06,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 87753085,
+      "step": 4130
+    },
+    {
+      "epoch": 0.49672338123008475,
+      "grad_norm": 1.8424606088018047,
+      "learning_rate": 2.1183345378140206e-06,
+      "loss": 0.7425,
+      "num_input_tokens_seen": 87772550,
+      "step": 4131
+    },
+    {
+      "epoch": 0.49684362412072386,
+      "grad_norm": 0.982729697096362,
+      "learning_rate": 2.117556921904778e-06,
+      "loss": 0.6602,
+      "num_input_tokens_seen": 87833710,
+      "step": 4132
+    },
+    {
+      "epoch": 0.49696386701136297,
+      "grad_norm": 6.820503470932465,
+      "learning_rate": 2.1167792881622437e-06,
+      "loss": 0.7222,
+      "num_input_tokens_seen": 87852450,
+      "step": 4133
+    },
+    {
+      "epoch": 0.497084109902002,
+      "grad_norm": 2.4952973125165623,
+      "learning_rate": 2.116001636704384e-06,
+      "loss": 0.8009,
+      "num_input_tokens_seen": 87872555,
+      "step": 4134
+    },
+    {
+      "epoch": 0.49720435279264114,
+      "grad_norm": 2.110032017168059,
+      "learning_rate": 2.1152239676491685e-06,
+      "loss": 0.8069,
+      "num_input_tokens_seen": 87890380,
+      "step": 4135
+    },
+    {
+      "epoch": 0.49732459568328025,
+      "grad_norm": 13.799250221120738,
+      "learning_rate": 2.1144462811145685e-06,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 87909120,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4974448385739193,
+      "grad_norm": 2.466884498108347,
+      "learning_rate": 2.1136685772185587e-06,
+      "loss": 0.7511,
+      "num_input_tokens_seen": 87927865,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4975650814645584,
+      "grad_norm": 1.7264769270221016,
+      "learning_rate": 2.1128908560791163e-06,
+      "loss": 0.7722,
+      "num_input_tokens_seen": 87947415,
+      "step": 4138
+    },
+    {
+      "epoch": 0.4976853243551975,
+      "grad_norm": 3.3374382461582988,
+      "learning_rate": 2.1121131178142203e-06,
+      "loss": 0.7786,
+      "num_input_tokens_seen": 87966500,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4978055672458366,
+      "grad_norm": 2.1306517597334236,
+      "learning_rate": 2.1113353625418544e-06,
+      "loss": 0.8156,
+      "num_input_tokens_seen": 87984770,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4979258101364757,
+      "grad_norm": 2.5348808693605753,
+      "learning_rate": 2.1105575903800017e-06,
+      "loss": 0.7886,
+      "num_input_tokens_seen": 88003210,
+      "step": 4141
+    },
+    {
+      "epoch": 0.4980460530271148,
+      "grad_norm": 1.855583746941516,
+      "learning_rate": 2.1097798014466502e-06,
+      "loss": 0.846,
+      "num_input_tokens_seen": 88022530,
+      "step": 4142
+    },
+    {
+      "epoch": 0.49816629591775385,
+      "grad_norm": 2.489770694251267,
+      "learning_rate": 2.109001995859791e-06,
+      "loss": 0.5809,
+      "num_input_tokens_seen": 88041150,
+      "step": 4143
+    },
+    {
+      "epoch": 0.49828653880839296,
+      "grad_norm": 0.7857729028641107,
+      "learning_rate": 2.108224173737415e-06,
+      "loss": 0.6299,
+      "num_input_tokens_seen": 88104170,
+      "step": 4144
+    },
+    {
+      "epoch": 0.498406781699032,
+      "grad_norm": 3.8151541126502857,
+      "learning_rate": 2.1074463351975183e-06,
+      "loss": 0.7579,
+      "num_input_tokens_seen": 88122775,
+      "step": 4145
+    },
+    {
+      "epoch": 0.49852702458967113,
+      "grad_norm": 2.2004693243475564,
+      "learning_rate": 2.1066684803580977e-06,
+      "loss": 0.7108,
+      "num_input_tokens_seen": 88142720,
+      "step": 4146
+    },
+    {
+      "epoch": 0.49864726748031024,
+      "grad_norm": 1.591730928756832,
+      "learning_rate": 2.1058906093371536e-06,
+      "loss": 0.7049,
+      "num_input_tokens_seen": 88160955,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4987675103709493,
+      "grad_norm": 0.7027569594949499,
+      "learning_rate": 2.1051127222526883e-06,
+      "loss": 0.645,
+      "num_input_tokens_seen": 88232790,
+      "step": 4148
+    },
+    {
+      "epoch": 0.4988877532615884,
+      "grad_norm": 1.8119690419736563,
+      "learning_rate": 2.104334819222707e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 88252880,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4990079961522275,
+      "grad_norm": 2.5851918849174464,
+      "learning_rate": 2.1035569003652156e-06,
+      "loss": 0.6268,
+      "num_input_tokens_seen": 88271230,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4991282390428666,
+      "grad_norm": 2.307667873597635,
+      "learning_rate": 2.1027789657982255e-06,
+      "loss": 0.813,
+      "num_input_tokens_seen": 88285165,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4992484819335057,
+      "grad_norm": 1.9705379995932804,
+      "learning_rate": 2.102001015639748e-06,
+      "loss": 0.768,
+      "num_input_tokens_seen": 88302105,
+      "step": 4152
+    },
+    {
+      "epoch": 0.4993687248241448,
+      "grad_norm": 1.9656643552084356,
+      "learning_rate": 2.101223050007797e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 88320375,
+      "step": 4153
+    },
+    {
+      "epoch": 0.49948896771478385,
+      "grad_norm": 0.8792469235928111,
+      "learning_rate": 2.10044506902039e-06,
+      "loss": 0.5815,
+      "num_input_tokens_seen": 88376175,
+      "step": 4154
+    },
+    {
+      "epoch": 0.49960921060542296,
+      "grad_norm": 1.0007593181860748,
+      "learning_rate": 2.099667072795546e-06,
+      "loss": 0.7138,
+      "num_input_tokens_seen": 88438015,
+      "step": 4155
+    },
+    {
+      "epoch": 0.49972945349606207,
+      "grad_norm": 2.3208804633713,
+      "learning_rate": 2.0988890614512864e-06,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 88457625,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4998496963867011,
+      "grad_norm": 2.0836713806088243,
+      "learning_rate": 2.098111035105635e-06,
+      "loss": 0.8279,
+      "num_input_tokens_seen": 88475770,
+      "step": 4157
+    },
+    {
+      "epoch": 0.49996993927734024,
+      "grad_norm": 2.190331275261078,
+      "learning_rate": 2.0973329938766172e-06,
+      "loss": 0.7283,
+      "num_input_tokens_seen": 88492920,
+      "step": 4158
+    },
+    {
+      "epoch": 0.5000901821679793,
+      "grad_norm": 3.094957335929545,
+      "learning_rate": 2.0965549378822618e-06,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 88513930,
+      "step": 4159
+    },
+    {
+      "epoch": 0.5002104250586185,
+      "grad_norm": 3.703561861804555,
+      "learning_rate": 2.095776867240599e-06,
+      "loss": 0.8331,
+      "num_input_tokens_seen": 88530640,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5003306679492575,
+      "grad_norm": 3.85246810192429,
+      "learning_rate": 2.094998782069661e-06,
+      "loss": 0.8244,
+      "num_input_tokens_seen": 88548065,
+      "step": 4161
+    },
+    {
+      "epoch": 0.5004509108398966,
+      "grad_norm": 2.2560065038443073,
+      "learning_rate": 2.0942206824874845e-06,
+      "loss": 0.743,
+      "num_input_tokens_seen": 88570560,
+      "step": 4162
+    },
+    {
+      "epoch": 0.5005711537305357,
+      "grad_norm": 5.897810970123794,
+      "learning_rate": 2.093442568612105e-06,
+      "loss": 0.7934,
+      "num_input_tokens_seen": 88588085,
+      "step": 4163
+    },
+    {
+      "epoch": 0.5006913966211748,
+      "grad_norm": 2.8639942633022284,
+      "learning_rate": 2.0926644405615613e-06,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 88608705,
+      "step": 4164
+    },
+    {
+      "epoch": 0.5008116395118138,
+      "grad_norm": 2.233737780371001,
+      "learning_rate": 2.091886298453897e-06,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 88626610,
+      "step": 4165
+    },
+    {
+      "epoch": 0.500931882402453,
+      "grad_norm": 3.3045622628933837,
+      "learning_rate": 2.091108142407153e-06,
+      "loss": 0.7345,
+      "num_input_tokens_seen": 88645070,
+      "step": 4166
+    },
+    {
+      "epoch": 0.5010521252930921,
+      "grad_norm": 0.9508381683313963,
+      "learning_rate": 2.090329972539377e-06,
+      "loss": 0.6933,
+      "num_input_tokens_seen": 88703355,
+      "step": 4167
+    },
+    {
+      "epoch": 0.5011723681837311,
+      "grad_norm": 2.559090510837635,
+      "learning_rate": 2.0895517889686155e-06,
+      "loss": 0.6781,
+      "num_input_tokens_seen": 88721040,
+      "step": 4168
+    },
+    {
+      "epoch": 0.5012926110743702,
+      "grad_norm": 0.9205286407463413,
+      "learning_rate": 2.0887735918129194e-06,
+      "loss": 0.6486,
+      "num_input_tokens_seen": 88777325,
+      "step": 4169
+    },
+    {
+      "epoch": 0.5014128539650093,
+      "grad_norm": 11.974238113996211,
+      "learning_rate": 2.0879953811903396e-06,
+      "loss": 0.849,
+      "num_input_tokens_seen": 88791930,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5015330968556484,
+      "grad_norm": 5.20233103910194,
+      "learning_rate": 2.0872171572189305e-06,
+      "loss": 0.7786,
+      "num_input_tokens_seen": 88810975,
+      "step": 4171
+    },
+    {
+      "epoch": 0.5016533397462875,
+      "grad_norm": 2.587323056128286,
+      "learning_rate": 2.0864389200167477e-06,
+      "loss": 0.7649,
+      "num_input_tokens_seen": 88828950,
+      "step": 4172
+    },
+    {
+      "epoch": 0.5017735826369266,
+      "grad_norm": 3.5283126849474202,
+      "learning_rate": 2.0856606697018504e-06,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 88846680,
+      "step": 4173
+    },
+    {
+      "epoch": 0.5018938255275657,
+      "grad_norm": 31.681178241362144,
+      "learning_rate": 2.0848824063922966e-06,
+      "loss": 0.732,
+      "num_input_tokens_seen": 88864360,
+      "step": 4174
+    },
+    {
+      "epoch": 0.5020140684182047,
+      "grad_norm": 5.457804274314919,
+      "learning_rate": 2.0841041302061496e-06,
+      "loss": 0.6969,
+      "num_input_tokens_seen": 88883540,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5021343113088439,
+      "grad_norm": 2.8333727765424563,
+      "learning_rate": 2.0833258412614728e-06,
+      "loss": 0.748,
+      "num_input_tokens_seen": 88902320,
+      "step": 4176
+    },
+    {
+      "epoch": 0.502254554199483,
+      "grad_norm": 3.121180037803544,
+      "learning_rate": 2.0825475396763322e-06,
+      "loss": 0.6662,
+      "num_input_tokens_seen": 88922690,
+      "step": 4177
+    },
+    {
+      "epoch": 0.502374797090122,
+      "grad_norm": 1.7486369089352134,
+      "learning_rate": 2.081769225568796e-06,
+      "loss": 0.6466,
+      "num_input_tokens_seen": 88944860,
+      "step": 4178
+    },
+    {
+      "epoch": 0.5024950399807612,
+      "grad_norm": 2.129365363679905,
+      "learning_rate": 2.0809908990569327e-06,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 88966360,
+      "step": 4179
+    },
+    {
+      "epoch": 0.5026152828714002,
+      "grad_norm": 4.08739389311478,
+      "learning_rate": 2.080212560258814e-06,
+      "loss": 0.7927,
+      "num_input_tokens_seen": 88985345,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5027355257620393,
+      "grad_norm": 2.173552467331007,
+      "learning_rate": 2.0794342092925146e-06,
+      "loss": 0.6672,
+      "num_input_tokens_seen": 89006200,
+      "step": 4181
+    },
+    {
+      "epoch": 0.5028557686526784,
+      "grad_norm": 2.9070891497657763,
+      "learning_rate": 2.078655846276108e-06,
+      "loss": 0.6817,
+      "num_input_tokens_seen": 89026250,
+      "step": 4182
+    },
+    {
+      "epoch": 0.5029760115433175,
+      "grad_norm": 2.7557653643091298,
+      "learning_rate": 2.0778774713276727e-06,
+      "loss": 0.6739,
+      "num_input_tokens_seen": 89045445,
+      "step": 4183
+    },
+    {
+      "epoch": 0.5030962544339566,
+      "grad_norm": 2.4909336152406008,
+      "learning_rate": 2.077099084565287e-06,
+      "loss": 0.6604,
+      "num_input_tokens_seen": 89062570,
+      "step": 4184
+    },
+    {
+      "epoch": 0.5032164973245957,
+      "grad_norm": 3.236765329882337,
+      "learning_rate": 2.0763206861070313e-06,
+      "loss": 0.6406,
+      "num_input_tokens_seen": 89081350,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5033367402152348,
+      "grad_norm": 3.128946858176071,
+      "learning_rate": 2.0755422760709876e-06,
+      "loss": 0.7448,
+      "num_input_tokens_seen": 89098470,
+      "step": 4186
+    },
+    {
+      "epoch": 0.5034569831058738,
+      "grad_norm": 6.150868167371143,
+      "learning_rate": 2.0747638545752417e-06,
+      "loss": 0.7603,
+      "num_input_tokens_seen": 89116750,
+      "step": 4187
+    },
+    {
+      "epoch": 0.503577225996513,
+      "grad_norm": 2.913340756215938,
+      "learning_rate": 2.073985421737878e-06,
+      "loss": 0.8221,
+      "num_input_tokens_seen": 89133780,
+      "step": 4188
+    },
+    {
+      "epoch": 0.5036974688871521,
+      "grad_norm": 6.559936612330904,
+      "learning_rate": 2.0732069776769844e-06,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 89150910,
+      "step": 4189
+    },
+    {
+      "epoch": 0.5038177117777911,
+      "grad_norm": 2.754252862776433,
+      "learning_rate": 2.0724285225106505e-06,
+      "loss": 0.725,
+      "num_input_tokens_seen": 89167195,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5039379546684303,
+      "grad_norm": 2.8177142561799164,
+      "learning_rate": 2.0716500563569677e-06,
+      "loss": 0.7598,
+      "num_input_tokens_seen": 89184455,
+      "step": 4191
+    },
+    {
+      "epoch": 0.5040581975590693,
+      "grad_norm": 3.155598690182525,
+      "learning_rate": 2.070871579334028e-06,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 89203285,
+      "step": 4192
+    },
+    {
+      "epoch": 0.5041784404497084,
+      "grad_norm": 2.849103600292471,
+      "learning_rate": 2.070093091559927e-06,
+      "loss": 0.7204,
+      "num_input_tokens_seen": 89222735,
+      "step": 4193
+    },
+    {
+      "epoch": 0.5042986833403476,
+      "grad_norm": 2.374926093173233,
+      "learning_rate": 2.0693145931527583e-06,
+      "loss": 0.7771,
+      "num_input_tokens_seen": 89238935,
+      "step": 4194
+    },
+    {
+      "epoch": 0.5044189262309866,
+      "grad_norm": 3.5498692524042124,
+      "learning_rate": 2.068536084230622e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 89260520,
+      "step": 4195
+    },
+    {
+      "epoch": 0.5045391691216257,
+      "grad_norm": 3.1018330060739374,
+      "learning_rate": 2.0677575649116155e-06,
+      "loss": 0.8775,
+      "num_input_tokens_seen": 89278815,
+      "step": 4196
+    },
+    {
+      "epoch": 0.5046594120122648,
+      "grad_norm": 5.103455401470044,
+      "learning_rate": 2.0669790353138407e-06,
+      "loss": 0.9209,
+      "num_input_tokens_seen": 89297500,
+      "step": 4197
+    },
+    {
+      "epoch": 0.5047796549029039,
+      "grad_norm": 3.757458134993652,
+      "learning_rate": 2.066200495555399e-06,
+      "loss": 0.7203,
+      "num_input_tokens_seen": 89316920,
+      "step": 4198
+    },
+    {
+      "epoch": 0.5048998977935429,
+      "grad_norm": 2.7735056121167405,
+      "learning_rate": 2.065421945754396e-06,
+      "loss": 0.7561,
+      "num_input_tokens_seen": 89334370,
+      "step": 4199
+    },
+    {
+      "epoch": 0.505020140684182,
+      "grad_norm": 2.4737853612715717,
+      "learning_rate": 2.0646433860289344e-06,
+      "loss": 0.7731,
+      "num_input_tokens_seen": 89353015,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5051403835748212,
+      "grad_norm": 2.4757781453171863,
+      "learning_rate": 2.0638648164971233e-06,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 89371200,
+      "step": 4201
+    },
+    {
+      "epoch": 0.5052606264654602,
+      "grad_norm": 10.480835296454176,
+      "learning_rate": 2.0630862372770697e-06,
+      "loss": 0.8837,
+      "num_input_tokens_seen": 89391020,
+      "step": 4202
+    },
+    {
+      "epoch": 0.5053808693560993,
+      "grad_norm": 3.0684361374878746,
+      "learning_rate": 2.0623076484868846e-06,
+      "loss": 0.7495,
+      "num_input_tokens_seen": 89408195,
+      "step": 4203
+    },
+    {
+      "epoch": 0.5055011122467384,
+      "grad_norm": 0.9356533116447043,
+      "learning_rate": 2.061529050244679e-06,
+      "loss": 0.6841,
+      "num_input_tokens_seen": 89467660,
+      "step": 4204
+    },
+    {
+      "epoch": 0.5056213551373775,
+      "grad_norm": 2.8189475198886607,
+      "learning_rate": 2.060750442668565e-06,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 89485135,
+      "step": 4205
+    },
+    {
+      "epoch": 0.5057415980280165,
+      "grad_norm": 3.7122716323904834,
+      "learning_rate": 2.059971825876657e-06,
+      "loss": 0.6385,
+      "num_input_tokens_seen": 89499365,
+      "step": 4206
+    },
+    {
+      "epoch": 0.5058618409186557,
+      "grad_norm": 2.4775335641735197,
+      "learning_rate": 2.0591931999870713e-06,
+      "loss": 0.7571,
+      "num_input_tokens_seen": 89518010,
+      "step": 4207
+    },
+    {
+      "epoch": 0.5059820838092948,
+      "grad_norm": 0.9226974977271406,
+      "learning_rate": 2.0584145651179234e-06,
+      "loss": 0.6363,
+      "num_input_tokens_seen": 89573440,
+      "step": 4208
+    },
+    {
+      "epoch": 0.5061023266999338,
+      "grad_norm": 2.992854677917813,
+      "learning_rate": 2.0576359213873327e-06,
+      "loss": 0.7768,
+      "num_input_tokens_seen": 89588310,
+      "step": 4209
+    },
+    {
+      "epoch": 0.506222569590573,
+      "grad_norm": 3.279759014846503,
+      "learning_rate": 2.056857268913419e-06,
+      "loss": 0.6963,
+      "num_input_tokens_seen": 89608080,
+      "step": 4210
+    },
+    {
+      "epoch": 0.506342812481212,
+      "grad_norm": 8.527151961052574,
+      "learning_rate": 2.056078607814303e-06,
+      "loss": 0.85,
+      "num_input_tokens_seen": 89623585,
+      "step": 4211
+    },
+    {
+      "epoch": 0.5064630553718511,
+      "grad_norm": 2.4326448439217234,
+      "learning_rate": 2.0552999382081054e-06,
+      "loss": 0.7922,
+      "num_input_tokens_seen": 89644295,
+      "step": 4212
+    },
+    {
+      "epoch": 0.5065832982624903,
+      "grad_norm": 1.855576288503193,
+      "learning_rate": 2.054521260212952e-06,
+      "loss": 0.8532,
+      "num_input_tokens_seen": 89663870,
+      "step": 4213
+    },
+    {
+      "epoch": 0.5067035411531293,
+      "grad_norm": 5.396526033370523,
+      "learning_rate": 2.0537425739469673e-06,
+      "loss": 0.6712,
+      "num_input_tokens_seen": 89682525,
+      "step": 4214
+    },
+    {
+      "epoch": 0.5068237840437684,
+      "grad_norm": 0.9435496088658525,
+      "learning_rate": 2.0529638795282763e-06,
+      "loss": 0.6457,
+      "num_input_tokens_seen": 89742115,
+      "step": 4215
+    },
+    {
+      "epoch": 0.5069440269344075,
+      "grad_norm": 2.4606522737107883,
+      "learning_rate": 2.052185177075007e-06,
+      "loss": 0.7545,
+      "num_input_tokens_seen": 89761405,
+      "step": 4216
+    },
+    {
+      "epoch": 0.5070642698250466,
+      "grad_norm": 1.8464865415622307,
+      "learning_rate": 2.051406466705288e-06,
+      "loss": 0.8219,
+      "num_input_tokens_seen": 89780665,
+      "step": 4217
+    },
+    {
+      "epoch": 0.5071845127156857,
+      "grad_norm": 2.5010700291381553,
+      "learning_rate": 2.050627748537248e-06,
+      "loss": 0.8018,
+      "num_input_tokens_seen": 89799210,
+      "step": 4218
+    },
+    {
+      "epoch": 0.5073047556063248,
+      "grad_norm": 2.6240284567193934,
+      "learning_rate": 2.04984902268902e-06,
+      "loss": 0.6622,
+      "num_input_tokens_seen": 89816130,
+      "step": 4219
+    },
+    {
+      "epoch": 0.5074249984969639,
+      "grad_norm": 2.696130012035538,
+      "learning_rate": 2.049070289278734e-06,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 89834910,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5075452413876029,
+      "grad_norm": 3.9335262363590084,
+      "learning_rate": 2.048291548424525e-06,
+      "loss": 0.6217,
+      "num_input_tokens_seen": 89856250,
+      "step": 4221
+    },
+    {
+      "epoch": 0.5076654842782421,
+      "grad_norm": 5.20491427694413,
+      "learning_rate": 2.047512800244526e-06,
+      "loss": 0.8326,
+      "num_input_tokens_seen": 89871235,
+      "step": 4222
+    },
+    {
+      "epoch": 0.5077857271688812,
+      "grad_norm": 2.5926388069727664,
+      "learning_rate": 2.046734044856873e-06,
+      "loss": 0.783,
+      "num_input_tokens_seen": 89890365,
+      "step": 4223
+    },
+    {
+      "epoch": 0.5079059700595202,
+      "grad_norm": 2.643388562606605,
+      "learning_rate": 2.0459552823797018e-06,
+      "loss": 0.7984,
+      "num_input_tokens_seen": 89908745,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5080262129501594,
+      "grad_norm": 5.759068285360543,
+      "learning_rate": 2.045176512931152e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 89923095,
+      "step": 4225
+    },
+    {
+      "epoch": 0.5081464558407984,
+      "grad_norm": 2.440216555795652,
+      "learning_rate": 2.0443977366293604e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 89940855,
+      "step": 4226
+    },
+    {
+      "epoch": 0.5082666987314375,
+      "grad_norm": 3.3024203056878245,
+      "learning_rate": 2.043618953592468e-06,
+      "loss": 0.7714,
+      "num_input_tokens_seen": 89963030,
+      "step": 4227
+    },
+    {
+      "epoch": 0.5083869416220766,
+      "grad_norm": 2.4467352158198383,
+      "learning_rate": 2.0428401639386144e-06,
+      "loss": 0.8066,
+      "num_input_tokens_seen": 89983315,
+      "step": 4228
+    },
+    {
+      "epoch": 0.5085071845127157,
+      "grad_norm": 0.9282028657554723,
+      "learning_rate": 2.042061367785943e-06,
+      "loss": 0.6527,
+      "num_input_tokens_seen": 90036175,
+      "step": 4229
+    },
+    {
+      "epoch": 0.5086274274033548,
+      "grad_norm": 3.620002678549815,
+      "learning_rate": 2.041282565252594e-06,
+      "loss": 0.7437,
+      "num_input_tokens_seen": 90056060,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5087476702939938,
+      "grad_norm": 2.1149812001619335,
+      "learning_rate": 2.040503756456714e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 90074990,
+      "step": 4231
+    },
+    {
+      "epoch": 0.508867913184633,
+      "grad_norm": 2.9484106170940403,
+      "learning_rate": 2.0397249415164456e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 90092065,
+      "step": 4232
+    },
+    {
+      "epoch": 0.508988156075272,
+      "grad_norm": 2.042825478319309,
+      "learning_rate": 2.0389461205499354e-06,
+      "loss": 0.7995,
+      "num_input_tokens_seen": 90110920,
+      "step": 4233
+    },
+    {
+      "epoch": 0.5091083989659111,
+      "grad_norm": 2.5303272342809384,
+      "learning_rate": 2.0381672936753297e-06,
+      "loss": 0.7227,
+      "num_input_tokens_seen": 90128795,
+      "step": 4234
+    },
+    {
+      "epoch": 0.5092286418565503,
+      "grad_norm": 2.5000064404450244,
+      "learning_rate": 2.037388461010776e-06,
+      "loss": 0.7042,
+      "num_input_tokens_seen": 90148830,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5093488847471893,
+      "grad_norm": 9.663893296080847,
+      "learning_rate": 2.0366096226744225e-06,
+      "loss": 0.6832,
+      "num_input_tokens_seen": 90163745,
+      "step": 4236
+    },
+    {
+      "epoch": 0.5094691276378284,
+      "grad_norm": 1.9980892402011279,
+      "learning_rate": 2.0358307787844183e-06,
+      "loss": 0.7706,
+      "num_input_tokens_seen": 90184140,
+      "step": 4237
+    },
+    {
+      "epoch": 0.5095893705284675,
+      "grad_norm": 3.3981728942634724,
+      "learning_rate": 2.0350519294589134e-06,
+      "loss": 0.7957,
+      "num_input_tokens_seen": 90201900,
+      "step": 4238
+    },
+    {
+      "epoch": 0.5097096134191066,
+      "grad_norm": 1.924500560688671,
+      "learning_rate": 2.0342730748160588e-06,
+      "loss": 0.8234,
+      "num_input_tokens_seen": 90222085,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5098298563097456,
+      "grad_norm": 2.6739341027765677,
+      "learning_rate": 2.0334942149740054e-06,
+      "loss": 0.7038,
+      "num_input_tokens_seen": 90242950,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5099500992003848,
+      "grad_norm": 2.389200320237709,
+      "learning_rate": 2.0327153500509067e-06,
+      "loss": 0.8296,
+      "num_input_tokens_seen": 90261695,
+      "step": 4241
+    },
+    {
+      "epoch": 0.5100703420910239,
+      "grad_norm": 2.0488100064388166,
+      "learning_rate": 2.0319364801649154e-06,
+      "loss": 0.8486,
+      "num_input_tokens_seen": 90279155,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5101905849816629,
+      "grad_norm": 2.986931179423849,
+      "learning_rate": 2.031157605434186e-06,
+      "loss": 0.8018,
+      "num_input_tokens_seen": 90299490,
+      "step": 4243
+    },
+    {
+      "epoch": 0.5103108278723021,
+      "grad_norm": 1.9928428142399661,
+      "learning_rate": 2.0303787259768715e-06,
+      "loss": 0.6288,
+      "num_input_tokens_seen": 90320110,
+      "step": 4244
+    },
+    {
+      "epoch": 0.5104310707629411,
+      "grad_norm": 4.167819751351299,
+      "learning_rate": 2.0295998419111294e-06,
+      "loss": 0.6817,
+      "num_input_tokens_seen": 90337120,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5105513136535802,
+      "grad_norm": 5.911066595048062,
+      "learning_rate": 2.0288209533551144e-06,
+      "loss": 0.7251,
+      "num_input_tokens_seen": 90354940,
+      "step": 4246
+    },
+    {
+      "epoch": 0.5106715565442194,
+      "grad_norm": 2.3408366349855956,
+      "learning_rate": 2.0280420604269834e-06,
+      "loss": 0.7798,
+      "num_input_tokens_seen": 90374200,
+      "step": 4247
+    },
+    {
+      "epoch": 0.5107917994348584,
+      "grad_norm": 0.7985233431917704,
+      "learning_rate": 2.0272631632448945e-06,
+      "loss": 0.6404,
+      "num_input_tokens_seen": 90443045,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5109120423254975,
+      "grad_norm": 3.4141041193505712,
+      "learning_rate": 2.026484261927005e-06,
+      "loss": 0.7376,
+      "num_input_tokens_seen": 90462260,
+      "step": 4249
+    },
+    {
+      "epoch": 0.5110322852161366,
+      "grad_norm": 2.989711923922118,
+      "learning_rate": 2.025705356591475e-06,
+      "loss": 0.7239,
+      "num_input_tokens_seen": 90479670,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5111525281067757,
+      "grad_norm": 0.8871605833744901,
+      "learning_rate": 2.024926447356462e-06,
+      "loss": 0.6188,
+      "num_input_tokens_seen": 90541675,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5112727709974147,
+      "grad_norm": 2.1606374572066414,
+      "learning_rate": 2.024147534340127e-06,
+      "loss": 0.7842,
+      "num_input_tokens_seen": 90559255,
+      "step": 4252
+    },
+    {
+      "epoch": 0.5113930138880539,
+      "grad_norm": 2.477900914295781,
+      "learning_rate": 2.02336861766063e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 90578035,
+      "step": 4253
+    },
+    {
+      "epoch": 0.511513256778693,
+      "grad_norm": 2.143114257516258,
+      "learning_rate": 2.0225896974361327e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 90597860,
+      "step": 4254
+    },
+    {
+      "epoch": 0.511633499669332,
+      "grad_norm": 2.432274467029066,
+      "learning_rate": 2.0218107737847962e-06,
+      "loss": 0.656,
+      "num_input_tokens_seen": 90659625,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5117537425599712,
+      "grad_norm": 2.2495354488202306,
+      "learning_rate": 2.021031846824782e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 90678315,
+      "step": 4256
+    },
+    {
+      "epoch": 0.5118739854506102,
+      "grad_norm": 2.194469637942036,
+      "learning_rate": 2.020252916674254e-06,
+      "loss": 0.8138,
+      "num_input_tokens_seen": 90697150,
+      "step": 4257
+    },
+    {
+      "epoch": 0.5119942283412493,
+      "grad_norm": 2.876658687226483,
+      "learning_rate": 2.019473983451375e-06,
+      "loss": 0.803,
+      "num_input_tokens_seen": 90715290,
+      "step": 4258
+    },
+    {
+      "epoch": 0.5121144712318885,
+      "grad_norm": 3.2773293239250374,
+      "learning_rate": 2.0186950472743076e-06,
+      "loss": 0.7005,
+      "num_input_tokens_seen": 90734915,
+      "step": 4259
+    },
+    {
+      "epoch": 0.5122347141225275,
+      "grad_norm": 1.9817477030496025,
+      "learning_rate": 2.0179161082612162e-06,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 90754025,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5123549570131666,
+      "grad_norm": 5.938833149544115,
+      "learning_rate": 2.017137166530266e-06,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 90773280,
+      "step": 4261
+    },
+    {
+      "epoch": 0.5124751999038056,
+      "grad_norm": 2.609474253535179,
+      "learning_rate": 2.016358222199621e-06,
+      "loss": 0.8007,
+      "num_input_tokens_seen": 90791375,
+      "step": 4262
+    },
+    {
+      "epoch": 0.5125954427944448,
+      "grad_norm": 3.0604825711032912,
+      "learning_rate": 2.015579275387446e-06,
+      "loss": 0.678,
+      "num_input_tokens_seen": 90816415,
+      "step": 4263
+    },
+    {
+      "epoch": 0.5127156856850839,
+      "grad_norm": 3.2214540387017343,
+      "learning_rate": 2.0148003262119085e-06,
+      "loss": 0.6872,
+      "num_input_tokens_seen": 90837105,
+      "step": 4264
+    },
+    {
+      "epoch": 0.5128359285757229,
+      "grad_norm": 2.2730404991139816,
+      "learning_rate": 2.014021374791173e-06,
+      "loss": 0.7649,
+      "num_input_tokens_seen": 90855360,
+      "step": 4265
+    },
+    {
+      "epoch": 0.5129561714663621,
+      "grad_norm": 2.3979203686641064,
+      "learning_rate": 2.013242421243406e-06,
+      "loss": 0.7984,
+      "num_input_tokens_seen": 90873985,
+      "step": 4266
+    },
+    {
+      "epoch": 0.5130764143570011,
+      "grad_norm": 3.8914629187626013,
+      "learning_rate": 2.012463465686774e-06,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 90893455,
+      "step": 4267
+    },
+    {
+      "epoch": 0.5131966572476402,
+      "grad_norm": 1.2787630849663625,
+      "learning_rate": 2.0116845082394442e-06,
+      "loss": 0.598,
+      "num_input_tokens_seen": 90958010,
+      "step": 4268
+    },
+    {
+      "epoch": 0.5133169001382794,
+      "grad_norm": 2.5066625973920234,
+      "learning_rate": 2.0109055490195836e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 90976185,
+      "step": 4269
+    },
+    {
+      "epoch": 0.5134371430289184,
+      "grad_norm": 2.698562216743721,
+      "learning_rate": 2.01012658814536e-06,
+      "loss": 0.6388,
+      "num_input_tokens_seen": 90994380,
+      "step": 4270
+    },
+    {
+      "epoch": 0.5135573859195575,
+      "grad_norm": 2.446307156922511,
+      "learning_rate": 2.009347625734941e-06,
+      "loss": 0.7896,
+      "num_input_tokens_seen": 91014010,
+      "step": 4271
+    },
+    {
+      "epoch": 0.5136776288101966,
+      "grad_norm": 3.0333839865120917,
+      "learning_rate": 2.0085686619064954e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 91030170,
+      "step": 4272
+    },
+    {
+      "epoch": 0.5137978717008357,
+      "grad_norm": 5.422883352406179,
+      "learning_rate": 2.00778969677819e-06,
+      "loss": 0.831,
+      "num_input_tokens_seen": 91046925,
+      "step": 4273
+    },
+    {
+      "epoch": 0.5139181145914747,
+      "grad_norm": 3.0214315908554625,
+      "learning_rate": 2.0070107304681934e-06,
+      "loss": 0.6454,
+      "num_input_tokens_seen": 91065600,
+      "step": 4274
+    },
+    {
+      "epoch": 0.5140383574821139,
+      "grad_norm": 2.1202631617984693,
+      "learning_rate": 2.006231763094675e-06,
+      "loss": 0.7775,
+      "num_input_tokens_seen": 91086340,
+      "step": 4275
+    },
+    {
+      "epoch": 0.514158600372753,
+      "grad_norm": 3.3392471000212978,
+      "learning_rate": 2.0054527947758027e-06,
+      "loss": 0.8599,
+      "num_input_tokens_seen": 91104860,
+      "step": 4276
+    },
+    {
+      "epoch": 0.514278843263392,
+      "grad_norm": 0.8365510281458459,
+      "learning_rate": 2.004673825629746e-06,
+      "loss": 0.5912,
+      "num_input_tokens_seen": 91165360,
+      "step": 4277
+    },
+    {
+      "epoch": 0.5143990861540312,
+      "grad_norm": 2.1751455784980798,
+      "learning_rate": 2.003894855774674e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 91186935,
+      "step": 4278
+    },
+    {
+      "epoch": 0.5145193290446702,
+      "grad_norm": 3.086526452217173,
+      "learning_rate": 2.0031158853287554e-06,
+      "loss": 0.748,
+      "num_input_tokens_seen": 91207090,
+      "step": 4279
+    },
+    {
+      "epoch": 0.5146395719353093,
+      "grad_norm": 2.4887210848649204,
+      "learning_rate": 2.0023369144101593e-06,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 91224980,
+      "step": 4280
+    },
+    {
+      "epoch": 0.5147598148259485,
+      "grad_norm": 2.4324375921213983,
+      "learning_rate": 2.0015579431370555e-06,
+      "loss": 0.7606,
+      "num_input_tokens_seen": 91246380,
+      "step": 4281
+    },
+    {
+      "epoch": 0.5148800577165875,
+      "grad_norm": 17.028705456928314,
+      "learning_rate": 2.000778971627612e-06,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 91265565,
+      "step": 4282
+    },
+    {
+      "epoch": 0.5150003006072266,
+      "grad_norm": 2.840245832403917,
+      "learning_rate": 2e-06,
+      "loss": 0.8942,
+      "num_input_tokens_seen": 91282880,
+      "step": 4283
+    },
+    {
+      "epoch": 0.5151205434978657,
+      "grad_norm": 2.124973058379237,
+      "learning_rate": 1.9992210283723878e-06,
+      "loss": 0.8569,
+      "num_input_tokens_seen": 91299840,
+      "step": 4284
+    },
+    {
+      "epoch": 0.5152407863885048,
+      "grad_norm": 2.475802755275781,
+      "learning_rate": 1.998442056862945e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 91322325,
+      "step": 4285
+    },
+    {
+      "epoch": 0.5153610292791438,
+      "grad_norm": 2.6236899410785215,
+      "learning_rate": 1.9976630855898405e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 91339800,
+      "step": 4286
+    },
+    {
+      "epoch": 0.515481272169783,
+      "grad_norm": 2.739997182162152,
+      "learning_rate": 1.996884114671245e-06,
+      "loss": 0.7381,
+      "num_input_tokens_seen": 91359135,
+      "step": 4287
+    },
+    {
+      "epoch": 0.5156015150604221,
+      "grad_norm": 1.5960066151988939,
+      "learning_rate": 1.9961051442253263e-06,
+      "loss": 0.7014,
+      "num_input_tokens_seen": 91379090,
+      "step": 4288
+    },
+    {
+      "epoch": 0.5157217579510611,
+      "grad_norm": 2.242619769685269,
+      "learning_rate": 1.9953261743702543e-06,
+      "loss": 0.7913,
+      "num_input_tokens_seen": 91397910,
+      "step": 4289
+    },
+    {
+      "epoch": 0.5158420008417003,
+      "grad_norm": 1.8925339847482041,
+      "learning_rate": 1.9945472052241967e-06,
+      "loss": 0.7203,
+      "num_input_tokens_seen": 91416535,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5159622437323393,
+      "grad_norm": 3.823087621358085,
+      "learning_rate": 1.993768236905325e-06,
+      "loss": 0.6744,
+      "num_input_tokens_seen": 91436925,
+      "step": 4291
+    },
+    {
+      "epoch": 0.5160824866229784,
+      "grad_norm": 2.4781187606963093,
+      "learning_rate": 1.992989269531807e-06,
+      "loss": 0.6552,
+      "num_input_tokens_seen": 91455455,
+      "step": 4292
+    },
+    {
+      "epoch": 0.5162027295136175,
+      "grad_norm": 3.631057327550693,
+      "learning_rate": 1.9922103032218104e-06,
+      "loss": 0.6757,
+      "num_input_tokens_seen": 91471980,
+      "step": 4293
+    },
+    {
+      "epoch": 0.5163229724042566,
+      "grad_norm": 1.6061926751872557,
+      "learning_rate": 1.991431338093505e-06,
+      "loss": 0.8034,
+      "num_input_tokens_seen": 91494055,
+      "step": 4294
+    },
+    {
+      "epoch": 0.5164432152948957,
+      "grad_norm": 2.757170702023657,
+      "learning_rate": 1.9906523742650587e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 91512635,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5165634581855347,
+      "grad_norm": 2.4268439601317127,
+      "learning_rate": 1.98987341185464e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 91532305,
+      "step": 4296
+    },
+    {
+      "epoch": 0.5166837010761739,
+      "grad_norm": 1.7062827603353805,
+      "learning_rate": 1.9890944509804166e-06,
+      "loss": 0.7998,
+      "num_input_tokens_seen": 91552125,
+      "step": 4297
+    },
+    {
+      "epoch": 0.516803943966813,
+      "grad_norm": 2.607260410747192,
+      "learning_rate": 1.9883154917605556e-06,
+      "loss": 0.7642,
+      "num_input_tokens_seen": 91571055,
+      "step": 4298
+    },
+    {
+      "epoch": 0.516924186857452,
+      "grad_norm": 2.708475364401952,
+      "learning_rate": 1.9875365343132262e-06,
+      "loss": 0.8188,
+      "num_input_tokens_seen": 91587895,
+      "step": 4299
+    },
+    {
+      "epoch": 0.5170444297480912,
+      "grad_norm": 2.667900023407321,
+      "learning_rate": 1.9867575787565946e-06,
+      "loss": 0.8405,
+      "num_input_tokens_seen": 91602275,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5171646726387302,
+      "grad_norm": 4.990030954550193,
+      "learning_rate": 1.9859786252088275e-06,
+      "loss": 0.8577,
+      "num_input_tokens_seen": 91619175,
+      "step": 4301
+    },
+    {
+      "epoch": 0.5172849155293693,
+      "grad_norm": 3.047202311858612,
+      "learning_rate": 1.9851996737880914e-06,
+      "loss": 0.6628,
+      "num_input_tokens_seen": 91634080,
+      "step": 4302
+    },
+    {
+      "epoch": 0.5174051584200084,
+      "grad_norm": 3.739120074222469,
+      "learning_rate": 1.9844207246125537e-06,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 91650380,
+      "step": 4303
+    },
+    {
+      "epoch": 0.5175254013106475,
+      "grad_norm": 2.4108688483548226,
+      "learning_rate": 1.9836417778003794e-06,
+      "loss": 0.6864,
+      "num_input_tokens_seen": 91672745,
+      "step": 4304
+    },
+    {
+      "epoch": 0.5176456442012866,
+      "grad_norm": 0.8169154214823664,
+      "learning_rate": 1.9828628334697347e-06,
+      "loss": 0.6249,
+      "num_input_tokens_seen": 91739675,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5177658870919257,
+      "grad_norm": 0.8102509857012096,
+      "learning_rate": 1.9820838917387836e-06,
+      "loss": 0.597,
+      "num_input_tokens_seen": 91800265,
+      "step": 4306
+    },
+    {
+      "epoch": 0.5178861299825648,
+      "grad_norm": 1.6899432779489791,
+      "learning_rate": 1.9813049527256923e-06,
+      "loss": 0.8221,
+      "num_input_tokens_seen": 91820380,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5180063728732038,
+      "grad_norm": 4.034563796011272,
+      "learning_rate": 1.9805260165486252e-06,
+      "loss": 0.8274,
+      "num_input_tokens_seen": 91839470,
+      "step": 4308
+    },
+    {
+      "epoch": 0.518126615763843,
+      "grad_norm": 2.2360450308254514,
+      "learning_rate": 1.979747083325746e-06,
+      "loss": 0.8644,
+      "num_input_tokens_seen": 91858890,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5182468586544821,
+      "grad_norm": 2.7616577243437024,
+      "learning_rate": 1.9789681531752177e-06,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 91878830,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5183671015451211,
+      "grad_norm": 1.821260709369397,
+      "learning_rate": 1.978189226215204e-06,
+      "loss": 0.7255,
+      "num_input_tokens_seen": 91899095,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5184873444357603,
+      "grad_norm": 2.9685653504103025,
+      "learning_rate": 1.9774103025638675e-06,
+      "loss": 0.766,
+      "num_input_tokens_seen": 91916940,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5186075873263993,
+      "grad_norm": 1.6913024463278183,
+      "learning_rate": 1.97663138233937e-06,
+      "loss": 0.7578,
+      "num_input_tokens_seen": 91937525,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5187278302170384,
+      "grad_norm": 2.991297168858125,
+      "learning_rate": 1.9758524656598724e-06,
+      "loss": 0.6839,
+      "num_input_tokens_seen": 91953225,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5188480731076776,
+      "grad_norm": 21.56889296111843,
+      "learning_rate": 1.9750735526435377e-06,
+      "loss": 0.6998,
+      "num_input_tokens_seen": 91969890,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5189683159983166,
+      "grad_norm": 3.460673895993,
+      "learning_rate": 1.974294643408525e-06,
+      "loss": 0.7953,
+      "num_input_tokens_seen": 91987405,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5190885588889557,
+      "grad_norm": 2.409393181049484,
+      "learning_rate": 1.9735157380729947e-06,
+      "loss": 0.6662,
+      "num_input_tokens_seen": 92007535,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5192088017795948,
+      "grad_norm": 2.8245472072811784,
+      "learning_rate": 1.9727368367551053e-06,
+      "loss": 0.8355,
+      "num_input_tokens_seen": 92025805,
+      "step": 4318
+    },
+    {
+      "epoch": 0.5193290446702339,
+      "grad_norm": 2.454149967834717,
+      "learning_rate": 1.9719579395730164e-06,
+      "loss": 0.6882,
+      "num_input_tokens_seen": 92044900,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5194492875608729,
+      "grad_norm": 4.000719061143896,
+      "learning_rate": 1.971179046644886e-06,
+      "loss": 0.9226,
+      "num_input_tokens_seen": 92058640,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5195695304515121,
+      "grad_norm": 2.4336888395977296,
+      "learning_rate": 1.970400158088871e-06,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 92077100,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5196897733421512,
+      "grad_norm": 2.7565022621310002,
+      "learning_rate": 1.969621274023128e-06,
+      "loss": 0.8674,
+      "num_input_tokens_seen": 92095470,
+      "step": 4322
+    },
+    {
+      "epoch": 0.5198100162327902,
+      "grad_norm": 2.8388527492490403,
+      "learning_rate": 1.968842394565814e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 92116055,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5199302591234293,
+      "grad_norm": 8.03449850856374,
+      "learning_rate": 1.968063519835085e-06,
+      "loss": 0.7186,
+      "num_input_tokens_seen": 92135485,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5200505020140684,
+      "grad_norm": 3.218019743700349,
+      "learning_rate": 1.9672846499490935e-06,
+      "loss": 0.7213,
+      "num_input_tokens_seen": 92154415,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5201707449047075,
+      "grad_norm": 2.360547308356082,
+      "learning_rate": 1.966505785025994e-06,
+      "loss": 0.7157,
+      "num_input_tokens_seen": 92176040,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5202909877953465,
+      "grad_norm": 3.221811797575669,
+      "learning_rate": 1.965726925183941e-06,
+      "loss": 0.7594,
+      "num_input_tokens_seen": 92198865,
+      "step": 4327
+    },
+    {
+      "epoch": 0.5204112306859857,
+      "grad_norm": 3.7145767317145304,
+      "learning_rate": 1.964948070541087e-06,
+      "loss": 0.84,
+      "num_input_tokens_seen": 92217245,
+      "step": 4328
+    },
+    {
+      "epoch": 0.5205314735766248,
+      "grad_norm": 2.8580420021938053,
+      "learning_rate": 1.964169221215582e-06,
+      "loss": 0.6834,
+      "num_input_tokens_seen": 92234730,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5206517164672638,
+      "grad_norm": 2.5171657976586643,
+      "learning_rate": 1.9633903773255777e-06,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 92256765,
+      "step": 4330
+    },
+    {
+      "epoch": 0.520771959357903,
+      "grad_norm": 4.164258559088051,
+      "learning_rate": 1.962611538989224e-06,
+      "loss": 0.7475,
+      "num_input_tokens_seen": 92277275,
+      "step": 4331
+    },
+    {
+      "epoch": 0.520892202248542,
+      "grad_norm": 2.3137798832037504,
+      "learning_rate": 1.9618327063246705e-06,
+      "loss": 0.8498,
+      "num_input_tokens_seen": 92296845,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5210124451391811,
+      "grad_norm": 3.486670104636567,
+      "learning_rate": 1.961053879450065e-06,
+      "loss": 0.7809,
+      "num_input_tokens_seen": 92316115,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5211326880298203,
+      "grad_norm": 0.8254579896791291,
+      "learning_rate": 1.960275058483554e-06,
+      "loss": 0.6308,
+      "num_input_tokens_seen": 92381770,
+      "step": 4334
+    },
+    {
+      "epoch": 0.5212529309204593,
+      "grad_norm": 2.538645826597001,
+      "learning_rate": 1.959496243543286e-06,
+      "loss": 0.8275,
+      "num_input_tokens_seen": 92399370,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5213731738110984,
+      "grad_norm": 4.1577397829328,
+      "learning_rate": 1.9587174347474057e-06,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 92415600,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5214934167017375,
+      "grad_norm": 2.862591332216673,
+      "learning_rate": 1.957938632214058e-06,
+      "loss": 0.8078,
+      "num_input_tokens_seen": 92431000,
+      "step": 4337
+    },
+    {
+      "epoch": 0.5216136595923766,
+      "grad_norm": 2.2818172747199643,
+      "learning_rate": 1.9571598360613854e-06,
+      "loss": 0.8018,
+      "num_input_tokens_seen": 92453595,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5217339024830157,
+      "grad_norm": 2.4460848874558736,
+      "learning_rate": 1.956381046407532e-06,
+      "loss": 0.695,
+      "num_input_tokens_seen": 92473610,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5218541453736548,
+      "grad_norm": 2.339742394442145,
+      "learning_rate": 1.95560226337064e-06,
+      "loss": 0.8515,
+      "num_input_tokens_seen": 92492120,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5219743882642939,
+      "grad_norm": 3.411939574288475,
+      "learning_rate": 1.9548234870688486e-06,
+      "loss": 0.7916,
+      "num_input_tokens_seen": 92512050,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5220946311549329,
+      "grad_norm": 2.1472633873801192,
+      "learning_rate": 1.9540447176202976e-06,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 92533015,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5222148740455721,
+      "grad_norm": 0.8598588637250604,
+      "learning_rate": 1.9532659551431272e-06,
+      "loss": 0.6514,
+      "num_input_tokens_seen": 92599765,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5223351169362112,
+      "grad_norm": 1.8888903945012134,
+      "learning_rate": 1.9524871997554744e-06,
+      "loss": 0.658,
+      "num_input_tokens_seen": 92627245,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5224553598268502,
+      "grad_norm": 2.926895730972248,
+      "learning_rate": 1.951708451575475e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 92644030,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5225756027174894,
+      "grad_norm": 2.7329165232551715,
+      "learning_rate": 1.9509297107212657e-06,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 92660520,
+      "step": 4346
+    },
+    {
+      "epoch": 0.5226958456081284,
+      "grad_norm": 5.009365425022103,
+      "learning_rate": 1.95015097731098e-06,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 92679730,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5228160884987675,
+      "grad_norm": 3.1052224467568603,
+      "learning_rate": 1.9493722514627516e-06,
+      "loss": 0.814,
+      "num_input_tokens_seen": 92696865,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5229363313894067,
+      "grad_norm": 3.2363247045039794,
+      "learning_rate": 1.9485935332947124e-06,
+      "loss": 0.8208,
+      "num_input_tokens_seen": 92714495,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5230565742800457,
+      "grad_norm": 6.421452198815501,
+      "learning_rate": 1.9478148229249926e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 92731725,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5231768171706848,
+      "grad_norm": 6.96818827027789,
+      "learning_rate": 1.9470361204717236e-06,
+      "loss": 0.8216,
+      "num_input_tokens_seen": 92750585,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5232970600613239,
+      "grad_norm": 1.9420980659554217,
+      "learning_rate": 1.9462574260530326e-06,
+      "loss": 0.8055,
+      "num_input_tokens_seen": 92770585,
+      "step": 4352
+    },
+    {
+      "epoch": 0.523417302951963,
+      "grad_norm": 4.144984391168081,
+      "learning_rate": 1.9454787397870477e-06,
+      "loss": 0.8025,
+      "num_input_tokens_seen": 92787625,
+      "step": 4353
+    },
+    {
+      "epoch": 0.523537545842602,
+      "grad_norm": 2.6038864724338526,
+      "learning_rate": 1.944700061791894e-06,
+      "loss": 0.7168,
+      "num_input_tokens_seen": 92805740,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5236577887332411,
+      "grad_norm": 3.5043455093740294,
+      "learning_rate": 1.943921392185698e-06,
+      "loss": 0.6464,
+      "num_input_tokens_seen": 92824085,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5237780316238803,
+      "grad_norm": 16.549584717435856,
+      "learning_rate": 1.9431427310865814e-06,
+      "loss": 0.7674,
+      "num_input_tokens_seen": 92843410,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5238982745145193,
+      "grad_norm": 2.470064852139736,
+      "learning_rate": 1.9423640786126676e-06,
+      "loss": 0.7873,
+      "num_input_tokens_seen": 92861860,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5240185174051584,
+      "grad_norm": 2.2567559949489535,
+      "learning_rate": 1.941585434882076e-06,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 92881430,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5241387602957975,
+      "grad_norm": 2.506739240107256,
+      "learning_rate": 1.940806800012929e-06,
+      "loss": 0.6662,
+      "num_input_tokens_seen": 92901220,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5242590031864366,
+      "grad_norm": 1.9487721000759703,
+      "learning_rate": 1.940028174123343e-06,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 92925830,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5243792460770756,
+      "grad_norm": 0.7067978514468166,
+      "learning_rate": 1.939249557331435e-06,
+      "loss": 0.5567,
+      "num_input_tokens_seen": 92991365,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5244994889677148,
+      "grad_norm": 3.1522807992849913,
+      "learning_rate": 1.938470949755321e-06,
+      "loss": 0.7216,
+      "num_input_tokens_seen": 93010965,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5246197318583539,
+      "grad_norm": 0.9263953067355369,
+      "learning_rate": 1.937692351513115e-06,
+      "loss": 0.6222,
+      "num_input_tokens_seen": 93069680,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5247399747489929,
+      "grad_norm": 4.631820319099404,
+      "learning_rate": 1.93691376272293e-06,
+      "loss": 0.7974,
+      "num_input_tokens_seen": 93087800,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5248602176396321,
+      "grad_norm": 3.4619780138004383,
+      "learning_rate": 1.9361351835028773e-06,
+      "loss": 0.8767,
+      "num_input_tokens_seen": 93104820,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5249804605302711,
+      "grad_norm": 4.573756916794191,
+      "learning_rate": 1.9353566139710654e-06,
+      "loss": 0.8068,
+      "num_input_tokens_seen": 93125200,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5251007034209102,
+      "grad_norm": 2.4023745048422,
+      "learning_rate": 1.9345780542456043e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 93144295,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5252209463115494,
+      "grad_norm": 2.7027041859486407,
+      "learning_rate": 1.933799504444601e-06,
+      "loss": 0.7223,
+      "num_input_tokens_seen": 93162855,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5253411892021884,
+      "grad_norm": 2.9043143765723407,
+      "learning_rate": 1.93302096468616e-06,
+      "loss": 0.7912,
+      "num_input_tokens_seen": 93181725,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5254614320928275,
+      "grad_norm": 3.1500876552118307,
+      "learning_rate": 1.9322424350883843e-06,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 93203280,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5255816749834666,
+      "grad_norm": 2.109027213606183,
+      "learning_rate": 1.9314639157693784e-06,
+      "loss": 0.7743,
+      "num_input_tokens_seen": 93223115,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5257019178741057,
+      "grad_norm": 3.183577255607226,
+      "learning_rate": 1.930685406847242e-06,
+      "loss": 0.7327,
+      "num_input_tokens_seen": 93237410,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5258221607647448,
+      "grad_norm": 10.76185720688949,
+      "learning_rate": 1.929906908440074e-06,
+      "loss": 0.8189,
+      "num_input_tokens_seen": 93257990,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5259424036553839,
+      "grad_norm": 2.224820145878844,
+      "learning_rate": 1.9291284206659717e-06,
+      "loss": 0.6875,
+      "num_input_tokens_seen": 93275895,
+      "step": 4374
+    },
+    {
+      "epoch": 0.526062646546023,
+      "grad_norm": 2.8559530889567992,
+      "learning_rate": 1.9283499436430325e-06,
+      "loss": 0.7111,
+      "num_input_tokens_seen": 93294715,
+      "step": 4375
+    },
+    {
+      "epoch": 0.526182889436662,
+      "grad_norm": 3.751191388187443,
+      "learning_rate": 1.9275714774893497e-06,
+      "loss": 0.8274,
+      "num_input_tokens_seen": 93313890,
+      "step": 4376
+    },
+    {
+      "epoch": 0.5263031323273012,
+      "grad_norm": 2.9434411965998355,
+      "learning_rate": 1.926793022323016e-06,
+      "loss": 0.7354,
+      "num_input_tokens_seen": 93332085,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5264233752179402,
+      "grad_norm": 2.7650948054815605,
+      "learning_rate": 1.926014578262122e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 93349585,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5265436181085793,
+      "grad_norm": 1.9827055758628263,
+      "learning_rate": 1.925236145424758e-06,
+      "loss": 0.8736,
+      "num_input_tokens_seen": 93368125,
+      "step": 4379
+    },
+    {
+      "epoch": 0.5266638609992185,
+      "grad_norm": 0.726743589331589,
+      "learning_rate": 1.924457723929012e-06,
+      "loss": 0.6011,
+      "num_input_tokens_seen": 93438655,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5267841038898575,
+      "grad_norm": 1.758951297155496,
+      "learning_rate": 1.923679313892969e-06,
+      "loss": 0.8214,
+      "num_input_tokens_seen": 93457645,
+      "step": 4381
+    },
+    {
+      "epoch": 0.5269043467804966,
+      "grad_norm": 6.167261646418311,
+      "learning_rate": 1.922900915434713e-06,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 93474955,
+      "step": 4382
+    },
+    {
+      "epoch": 0.5270245896711357,
+      "grad_norm": 3.64843063781641,
+      "learning_rate": 1.922122528672327e-06,
+      "loss": 0.8064,
+      "num_input_tokens_seen": 93493340,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5271448325617748,
+      "grad_norm": 7.284114148568296,
+      "learning_rate": 1.921344153723892e-06,
+      "loss": 0.781,
+      "num_input_tokens_seen": 93509935,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5272650754524139,
+      "grad_norm": 0.9994149535033161,
+      "learning_rate": 1.9205657907074856e-06,
+      "loss": 0.6495,
+      "num_input_tokens_seen": 93575045,
+      "step": 4385
+    },
+    {
+      "epoch": 0.527385318343053,
+      "grad_norm": 4.1900659965811755,
+      "learning_rate": 1.9197874397411853e-06,
+      "loss": 0.6649,
+      "num_input_tokens_seen": 93591395,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5275055612336921,
+      "grad_norm": 4.722446008605398,
+      "learning_rate": 1.919009100943067e-06,
+      "loss": 0.6554,
+      "num_input_tokens_seen": 93606805,
+      "step": 4387
+    },
+    {
+      "epoch": 0.5276258041243311,
+      "grad_norm": 4.956860703755058,
+      "learning_rate": 1.9182307744312043e-06,
+      "loss": 0.6569,
+      "num_input_tokens_seen": 93623630,
+      "step": 4388
+    },
+    {
+      "epoch": 0.5277460470149702,
+      "grad_norm": 2.694895329348522,
+      "learning_rate": 1.9174524603236676e-06,
+      "loss": 0.7595,
+      "num_input_tokens_seen": 93642300,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5278662899056094,
+      "grad_norm": 3.474000544801508,
+      "learning_rate": 1.916674158738527e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 93660925,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5279865327962484,
+      "grad_norm": 2.396671485024513,
+      "learning_rate": 1.9158958697938506e-06,
+      "loss": 0.6016,
+      "num_input_tokens_seen": 93679025,
+      "step": 4391
+    },
+    {
+      "epoch": 0.5281067756868875,
+      "grad_norm": 5.854960807129127,
+      "learning_rate": 1.9151175936077036e-06,
+      "loss": 0.8541,
+      "num_input_tokens_seen": 93693715,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5282270185775266,
+      "grad_norm": 2.304777815690915,
+      "learning_rate": 1.9143393302981507e-06,
+      "loss": 0.7887,
+      "num_input_tokens_seen": 93711120,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5283472614681657,
+      "grad_norm": 3.51146715120032,
+      "learning_rate": 1.9135610799832517e-06,
+      "loss": 0.8285,
+      "num_input_tokens_seen": 93729665,
+      "step": 4394
+    },
+    {
+      "epoch": 0.5284675043588047,
+      "grad_norm": 5.6267238045277,
+      "learning_rate": 1.9127828427810693e-06,
+      "loss": 0.749,
+      "num_input_tokens_seen": 93749950,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5285877472494439,
+      "grad_norm": 2.4455011451216047,
+      "learning_rate": 1.9120046188096607e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 93767715,
+      "step": 4396
+    },
+    {
+      "epoch": 0.528707990140083,
+      "grad_norm": 4.489227879512186,
+      "learning_rate": 1.911226408187081e-06,
+      "loss": 0.746,
+      "num_input_tokens_seen": 93785825,
+      "step": 4397
+    },
+    {
+      "epoch": 0.528828233030722,
+      "grad_norm": 2.579181557626125,
+      "learning_rate": 1.9104482110313843e-06,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 93805135,
+      "step": 4398
+    },
+    {
+      "epoch": 0.5289484759213612,
+      "grad_norm": 2.3864483014280924,
+      "learning_rate": 1.909670027460623e-06,
+      "loss": 0.7413,
+      "num_input_tokens_seen": 93822155,
+      "step": 4399
+    },
+    {
+      "epoch": 0.5290687188120002,
+      "grad_norm": 2.442530661179652,
+      "learning_rate": 1.908891857592847e-06,
+      "loss": 0.7153,
+      "num_input_tokens_seen": 93842945,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5291889617026393,
+      "grad_norm": 2.5753357883674592,
+      "learning_rate": 1.9081137015461038e-06,
+      "loss": 0.8878,
+      "num_input_tokens_seen": 93858740,
+      "step": 4401
+    },
+    {
+      "epoch": 0.5293092045932785,
+      "grad_norm": 2.217216870881063,
+      "learning_rate": 1.9073355594384379e-06,
+      "loss": 0.8977,
+      "num_input_tokens_seen": 93876700,
+      "step": 4402
+    },
+    {
+      "epoch": 0.5294294474839175,
+      "grad_norm": 2.742685642382288,
+      "learning_rate": 1.906557431387895e-06,
+      "loss": 0.801,
+      "num_input_tokens_seen": 93895410,
+      "step": 4403
+    },
+    {
+      "epoch": 0.5295496903745566,
+      "grad_norm": 2.2487668565730914,
+      "learning_rate": 1.9057793175125156e-06,
+      "loss": 0.7741,
+      "num_input_tokens_seen": 93912675,
+      "step": 4404
+    },
+    {
+      "epoch": 0.5296699332651957,
+      "grad_norm": 2.796240715408067,
+      "learning_rate": 1.9050012179303385e-06,
+      "loss": 0.8027,
+      "num_input_tokens_seen": 93930905,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5297901761558348,
+      "grad_norm": 3.2172355322152653,
+      "learning_rate": 1.904223132759401e-06,
+      "loss": 0.683,
+      "num_input_tokens_seen": 93949225,
+      "step": 4406
+    },
+    {
+      "epoch": 0.5299104190464738,
+      "grad_norm": 3.0226313565650127,
+      "learning_rate": 1.9034450621177383e-06,
+      "loss": 0.687,
+      "num_input_tokens_seen": 93967265,
+      "step": 4407
+    },
+    {
+      "epoch": 0.530030661937113,
+      "grad_norm": 3.0912232833601005,
+      "learning_rate": 1.902667006123383e-06,
+      "loss": 0.7052,
+      "num_input_tokens_seen": 93984420,
+      "step": 4408
+    },
+    {
+      "epoch": 0.5301509048277521,
+      "grad_norm": 4.621179612881732,
+      "learning_rate": 1.9018889648943655e-06,
+      "loss": 0.8048,
+      "num_input_tokens_seen": 94003180,
+      "step": 4409
+    },
+    {
+      "epoch": 0.5302711477183911,
+      "grad_norm": 8.082062097438557,
+      "learning_rate": 1.901110938548713e-06,
+      "loss": 0.6784,
+      "num_input_tokens_seen": 94024150,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5303913906090303,
+      "grad_norm": 2.2548373898608003,
+      "learning_rate": 1.900332927204454e-06,
+      "loss": 0.6484,
+      "num_input_tokens_seen": 94042320,
+      "step": 4411
+    },
+    {
+      "epoch": 0.5305116334996693,
+      "grad_norm": 2.3480190408862742,
+      "learning_rate": 1.89955493097961e-06,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 94061345,
+      "step": 4412
+    },
+    {
+      "epoch": 0.5306318763903084,
+      "grad_norm": 2.0355745790844555,
+      "learning_rate": 1.8987769499922032e-06,
+      "loss": 0.7568,
+      "num_input_tokens_seen": 94080035,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5307521192809476,
+      "grad_norm": 2.6908883799514327,
+      "learning_rate": 1.897998984360252e-06,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 94098725,
+      "step": 4414
+    },
+    {
+      "epoch": 0.5308723621715866,
+      "grad_norm": 3.629613308427711,
+      "learning_rate": 1.8972210342017746e-06,
+      "loss": 0.778,
+      "num_input_tokens_seen": 94122185,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5309926050622257,
+      "grad_norm": 1.782050217652631,
+      "learning_rate": 1.8964430996347842e-06,
+      "loss": 0.6626,
+      "num_input_tokens_seen": 94143455,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5311128479528648,
+      "grad_norm": 2.48351731557814,
+      "learning_rate": 1.8956651807772936e-06,
+      "loss": 0.821,
+      "num_input_tokens_seen": 94161210,
+      "step": 4417
+    },
+    {
+      "epoch": 0.5312330908435039,
+      "grad_norm": 2.3983382985017316,
+      "learning_rate": 1.8948872777473115e-06,
+      "loss": 0.829,
+      "num_input_tokens_seen": 94178885,
+      "step": 4418
+    },
+    {
+      "epoch": 0.531353333734143,
+      "grad_norm": 2.398346175339401,
+      "learning_rate": 1.8941093906628462e-06,
+      "loss": 0.6333,
+      "num_input_tokens_seen": 94196390,
+      "step": 4419
+    },
+    {
+      "epoch": 0.531473576624782,
+      "grad_norm": 2.5024207529338316,
+      "learning_rate": 1.8933315196419024e-06,
+      "loss": 0.7117,
+      "num_input_tokens_seen": 94218255,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5315938195154212,
+      "grad_norm": 5.498818319451968,
+      "learning_rate": 1.892553664802482e-06,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 94235395,
+      "step": 4421
+    },
+    {
+      "epoch": 0.5317140624060602,
+      "grad_norm": 17.04042722055294,
+      "learning_rate": 1.8917758262625845e-06,
+      "loss": 0.756,
+      "num_input_tokens_seen": 94255355,
+      "step": 4422
+    },
+    {
+      "epoch": 0.5318343052966993,
+      "grad_norm": 2.3259055292546487,
+      "learning_rate": 1.8909980041402089e-06,
+      "loss": 0.8066,
+      "num_input_tokens_seen": 94273670,
+      "step": 4423
+    },
+    {
+      "epoch": 0.5319545481873384,
+      "grad_norm": 5.70026526891228,
+      "learning_rate": 1.8902201985533494e-06,
+      "loss": 0.6591,
+      "num_input_tokens_seen": 94290655,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5320747910779775,
+      "grad_norm": 2.995304165481853,
+      "learning_rate": 1.8894424096199987e-06,
+      "loss": 0.747,
+      "num_input_tokens_seen": 94309580,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5321950339686166,
+      "grad_norm": 2.574047933770709,
+      "learning_rate": 1.8886646374581459e-06,
+      "loss": 0.8605,
+      "num_input_tokens_seen": 94328525,
+      "step": 4426
+    },
+    {
+      "epoch": 0.5323152768592557,
+      "grad_norm": 2.359365978602531,
+      "learning_rate": 1.8878868821857795e-06,
+      "loss": 0.6995,
+      "num_input_tokens_seen": 94347895,
+      "step": 4427
+    },
+    {
+      "epoch": 0.5324355197498948,
+      "grad_norm": 2.648724418021193,
+      "learning_rate": 1.8871091439208842e-06,
+      "loss": 0.741,
+      "num_input_tokens_seen": 94369225,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5325557626405338,
+      "grad_norm": 5.05780777383036,
+      "learning_rate": 1.8863314227814418e-06,
+      "loss": 0.7712,
+      "num_input_tokens_seen": 94387255,
+      "step": 4429
+    },
+    {
+      "epoch": 0.532676005531173,
+      "grad_norm": 2.732734031889,
+      "learning_rate": 1.8855537188854313e-06,
+      "loss": 0.4789,
+      "num_input_tokens_seen": 94405950,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5327962484218121,
+      "grad_norm": 2.3070939620803133,
+      "learning_rate": 1.8847760323508315e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 94424575,
+      "step": 4431
+    },
+    {
+      "epoch": 0.5329164913124511,
+      "grad_norm": 1.8927473150470764,
+      "learning_rate": 1.883998363295616e-06,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 94441775,
+      "step": 4432
+    },
+    {
+      "epoch": 0.5330367342030903,
+      "grad_norm": 0.98667060284152,
+      "learning_rate": 1.8832207118377565e-06,
+      "loss": 0.6823,
+      "num_input_tokens_seen": 94496865,
+      "step": 4433
+    },
+    {
+      "epoch": 0.5331569770937293,
+      "grad_norm": 3.269748394264652,
+      "learning_rate": 1.882443078095222e-06,
+      "loss": 0.6892,
+      "num_input_tokens_seen": 94515465,
+      "step": 4434
+    },
+    {
+      "epoch": 0.5332772199843684,
+      "grad_norm": 0.8712139923144947,
+      "learning_rate": 1.8816654621859794e-06,
+      "loss": 0.6882,
+      "num_input_tokens_seen": 94574850,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5333974628750076,
+      "grad_norm": 2.257311872312223,
+      "learning_rate": 1.880887864227992e-06,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 94589975,
+      "step": 4436
+    },
+    {
+      "epoch": 0.5335177057656466,
+      "grad_norm": 3.0373034547459032,
+      "learning_rate": 1.8801102843392209e-06,
+      "loss": 0.6486,
+      "num_input_tokens_seen": 94609100,
+      "step": 4437
+    },
+    {
+      "epoch": 0.5336379486562857,
+      "grad_norm": 2.8821916088537316,
+      "learning_rate": 1.8793327226376234e-06,
+      "loss": 0.8491,
+      "num_input_tokens_seen": 94628140,
+      "step": 4438
+    },
+    {
+      "epoch": 0.5337581915469248,
+      "grad_norm": 2.35753029770418,
+      "learning_rate": 1.8785551792411569e-06,
+      "loss": 0.7969,
+      "num_input_tokens_seen": 94646870,
+      "step": 4439
+    },
+    {
+      "epoch": 0.5338784344375639,
+      "grad_norm": 2.438078951391267,
+      "learning_rate": 1.8777776542677733e-06,
+      "loss": 0.8262,
+      "num_input_tokens_seen": 94664640,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5339986773282029,
+      "grad_norm": 3.295531280338399,
+      "learning_rate": 1.877000147835422e-06,
+      "loss": 0.7146,
+      "num_input_tokens_seen": 94684035,
+      "step": 4441
+    },
+    {
+      "epoch": 0.5341189202188421,
+      "grad_norm": 2.644824057408577,
+      "learning_rate": 1.8762226600620504e-06,
+      "loss": 0.8283,
+      "num_input_tokens_seen": 94702370,
+      "step": 4442
+    },
+    {
+      "epoch": 0.5342391631094812,
+      "grad_norm": 7.1359933724098035,
+      "learning_rate": 1.8754451910656031e-06,
+      "loss": 0.5976,
+      "num_input_tokens_seen": 94715990,
+      "step": 4443
+    },
+    {
+      "epoch": 0.5343594060001202,
+      "grad_norm": 2.000797199205094,
+      "learning_rate": 1.8746677409640212e-06,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 94732810,
+      "step": 4444
+    },
+    {
+      "epoch": 0.5344796488907594,
+      "grad_norm": 2.2001629176456166,
+      "learning_rate": 1.8738903098752437e-06,
+      "loss": 0.8386,
+      "num_input_tokens_seen": 94751660,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5345998917813984,
+      "grad_norm": 2.3222989271348458,
+      "learning_rate": 1.8731128979172048e-06,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 94770580,
+      "step": 4446
+    },
+    {
+      "epoch": 0.5347201346720375,
+      "grad_norm": 2.2798175222388437,
+      "learning_rate": 1.8723355052078394e-06,
+      "loss": 0.6559,
+      "num_input_tokens_seen": 94790335,
+      "step": 4447
+    },
+    {
+      "epoch": 0.5348403775626767,
+      "grad_norm": 6.289137556229466,
+      "learning_rate": 1.871558131865076e-06,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 94809110,
+      "step": 4448
+    },
+    {
+      "epoch": 0.5349606204533157,
+      "grad_norm": 4.297781075734802,
+      "learning_rate": 1.8707807780068429e-06,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 94826645,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5350808633439548,
+      "grad_norm": 2.273705193317834,
+      "learning_rate": 1.8700034437510611e-06,
+      "loss": 0.6569,
+      "num_input_tokens_seen": 94846460,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5352011062345938,
+      "grad_norm": 3.877558780218437,
+      "learning_rate": 1.8692261292156549e-06,
+      "loss": 0.8002,
+      "num_input_tokens_seen": 94865415,
+      "step": 4451
+    },
+    {
+      "epoch": 0.535321349125233,
+      "grad_norm": 2.1554374582619804,
+      "learning_rate": 1.8684488345185405e-06,
+      "loss": 0.8179,
+      "num_input_tokens_seen": 94885310,
+      "step": 4452
+    },
+    {
+      "epoch": 0.535441592015872,
+      "grad_norm": 3.642572747656797,
+      "learning_rate": 1.8676715597776336e-06,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 94903375,
+      "step": 4453
+    },
+    {
+      "epoch": 0.5355618349065111,
+      "grad_norm": 1.8366259682792494,
+      "learning_rate": 1.8668943051108455e-06,
+      "loss": 0.7571,
+      "num_input_tokens_seen": 94920400,
+      "step": 4454
+    },
+    {
+      "epoch": 0.5356820777971503,
+      "grad_norm": 2.4695849171586404,
+      "learning_rate": 1.8661170706360856e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 94939285,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5358023206877893,
+      "grad_norm": 2.1278996738246057,
+      "learning_rate": 1.8653398564712598e-06,
+      "loss": 0.8146,
+      "num_input_tokens_seen": 94957950,
+      "step": 4456
+    },
+    {
+      "epoch": 0.5359225635784284,
+      "grad_norm": 1.6637854996793302,
+      "learning_rate": 1.8645626627342708e-06,
+      "loss": 0.8171,
+      "num_input_tokens_seen": 94978435,
+      "step": 4457
+    },
+    {
+      "epoch": 0.5360428064690675,
+      "grad_norm": 2.231268407188041,
+      "learning_rate": 1.8637854895430172e-06,
+      "loss": 0.7995,
+      "num_input_tokens_seen": 94997420,
+      "step": 4458
+    },
+    {
+      "epoch": 0.5361630493597066,
+      "grad_norm": 3.1239707106360344,
+      "learning_rate": 1.8630083370153974e-06,
+      "loss": 0.6809,
+      "num_input_tokens_seen": 95016780,
+      "step": 4459
+    },
+    {
+      "epoch": 0.5362832922503457,
+      "grad_norm": 0.8387809675883565,
+      "learning_rate": 1.8622312052693041e-06,
+      "loss": 0.5895,
+      "num_input_tokens_seen": 95077680,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5364035351409848,
+      "grad_norm": 5.63156720194881,
+      "learning_rate": 1.861454094422627e-06,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 95094070,
+      "step": 4461
+    },
+    {
+      "epoch": 0.5365237780316239,
+      "grad_norm": 5.824647863542392,
+      "learning_rate": 1.8606770045932537e-06,
+      "loss": 0.678,
+      "num_input_tokens_seen": 95112905,
+      "step": 4462
+    },
+    {
+      "epoch": 0.5366440209222629,
+      "grad_norm": 4.120999983008074,
+      "learning_rate": 1.8598999358990684e-06,
+      "loss": 0.8144,
+      "num_input_tokens_seen": 95132480,
+      "step": 4463
+    },
+    {
+      "epoch": 0.5367642638129021,
+      "grad_norm": 2.3812548353314806,
+      "learning_rate": 1.859122888457951e-06,
+      "loss": 0.7897,
+      "num_input_tokens_seen": 95150695,
+      "step": 4464
+    },
+    {
+      "epoch": 0.5368845067035412,
+      "grad_norm": 2.0256451081564997,
+      "learning_rate": 1.85834586238778e-06,
+      "loss": 0.8144,
+      "num_input_tokens_seen": 95169515,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5370047495941802,
+      "grad_norm": 2.067079322522307,
+      "learning_rate": 1.8575688578064277e-06,
+      "loss": 0.738,
+      "num_input_tokens_seen": 95187360,
+      "step": 4466
+    },
+    {
+      "epoch": 0.5371249924848194,
+      "grad_norm": 2.1133976624849606,
+      "learning_rate": 1.8567918748317674e-06,
+      "loss": 0.7562,
+      "num_input_tokens_seen": 95206430,
+      "step": 4467
+    },
+    {
+      "epoch": 0.5372452353754584,
+      "grad_norm": 2.4447912198549564,
+      "learning_rate": 1.8560149135816659e-06,
+      "loss": 0.8199,
+      "num_input_tokens_seen": 95222985,
+      "step": 4468
+    },
+    {
+      "epoch": 0.5373654782660975,
+      "grad_norm": 2.563339531592118,
+      "learning_rate": 1.8552379741739877e-06,
+      "loss": 0.8421,
+      "num_input_tokens_seen": 95240050,
+      "step": 4469
+    },
+    {
+      "epoch": 0.5374857211567367,
+      "grad_norm": 0.9263359863004152,
+      "learning_rate": 1.854461056726593e-06,
+      "loss": 0.5806,
+      "num_input_tokens_seen": 95293710,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5376059640473757,
+      "grad_norm": 3.269741620668846,
+      "learning_rate": 1.853684161357341e-06,
+      "loss": 0.838,
+      "num_input_tokens_seen": 95311090,
+      "step": 4471
+    },
+    {
+      "epoch": 0.5377262069380148,
+      "grad_norm": 3.3909582674245407,
+      "learning_rate": 1.852907288184085e-06,
+      "loss": 0.775,
+      "num_input_tokens_seen": 95329695,
+      "step": 4472
+    },
+    {
+      "epoch": 0.5378464498286539,
+      "grad_norm": 10.617191930527659,
+      "learning_rate": 1.8521304373246766e-06,
+      "loss": 0.7,
+      "num_input_tokens_seen": 95350460,
+      "step": 4473
+    },
+    {
+      "epoch": 0.537966692719293,
+      "grad_norm": 4.418473177654474,
+      "learning_rate": 1.8513536088969626e-06,
+      "loss": 0.8814,
+      "num_input_tokens_seen": 95367845,
+      "step": 4474
+    },
+    {
+      "epoch": 0.538086935609932,
+      "grad_norm": 2.629785559357314,
+      "learning_rate": 1.8505768030187884e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 95387695,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5382071785005712,
+      "grad_norm": 1.9645438551795233,
+      "learning_rate": 1.849800019807995e-06,
+      "loss": 0.7924,
+      "num_input_tokens_seen": 95408640,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5383274213912103,
+      "grad_norm": 2.3472180828043574,
+      "learning_rate": 1.8490232593824186e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 95424815,
+      "step": 4477
+    },
+    {
+      "epoch": 0.5384476642818493,
+      "grad_norm": 1.7984980452169808,
+      "learning_rate": 1.8482465218598935e-06,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 95444480,
+      "step": 4478
+    },
+    {
+      "epoch": 0.5385679071724885,
+      "grad_norm": 1.960688534881271,
+      "learning_rate": 1.8474698073582508e-06,
+      "loss": 0.8346,
+      "num_input_tokens_seen": 95465570,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5386881500631275,
+      "grad_norm": 2.440468062830047,
+      "learning_rate": 1.846693115995317e-06,
+      "loss": 0.8626,
+      "num_input_tokens_seen": 95481925,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5388083929537666,
+      "grad_norm": 1.8743435857275081,
+      "learning_rate": 1.8459164478889158e-06,
+      "loss": 0.8332,
+      "num_input_tokens_seen": 95503040,
+      "step": 4481
+    },
+    {
+      "epoch": 0.5389286358444056,
+      "grad_norm": 2.1722445227544838,
+      "learning_rate": 1.8451398031568658e-06,
+      "loss": 0.7568,
+      "num_input_tokens_seen": 95522385,
+      "step": 4482
+    },
+    {
+      "epoch": 0.5390488787350448,
+      "grad_norm": 1.7483135591373813,
+      "learning_rate": 1.8443631819169856e-06,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 95542830,
+      "step": 4483
+    },
+    {
+      "epoch": 0.5391691216256839,
+      "grad_norm": 3.1262200639516045,
+      "learning_rate": 1.8435865842870868e-06,
+      "loss": 0.8255,
+      "num_input_tokens_seen": 95560490,
+      "step": 4484
+    },
+    {
+      "epoch": 0.5392893645163229,
+      "grad_norm": 2.3602068511998233,
+      "learning_rate": 1.842810010384979e-06,
+      "loss": 0.716,
+      "num_input_tokens_seen": 95580005,
+      "step": 4485
+    },
+    {
+      "epoch": 0.5394096074069621,
+      "grad_norm": 2.444235480450681,
+      "learning_rate": 1.842033460328467e-06,
+      "loss": 0.7137,
+      "num_input_tokens_seen": 95598445,
+      "step": 4486
+    },
+    {
+      "epoch": 0.5395298502976011,
+      "grad_norm": 2.2407812315153977,
+      "learning_rate": 1.8412569342353541e-06,
+      "loss": 0.7481,
+      "num_input_tokens_seen": 95618320,
+      "step": 4487
+    },
+    {
+      "epoch": 0.5396500931882402,
+      "grad_norm": 2.3805927645388314,
+      "learning_rate": 1.840480432223438e-06,
+      "loss": 0.8446,
+      "num_input_tokens_seen": 95637045,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5397703360788794,
+      "grad_norm": 2.6192858027677888,
+      "learning_rate": 1.8397039544105136e-06,
+      "loss": 0.7702,
+      "num_input_tokens_seen": 95655850,
+      "step": 4489
+    },
+    {
+      "epoch": 0.5398905789695184,
+      "grad_norm": 2.23756559739194,
+      "learning_rate": 1.8389275009143707e-06,
+      "loss": 0.6995,
+      "num_input_tokens_seen": 95675310,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5400108218601575,
+      "grad_norm": 2.5963487295155923,
+      "learning_rate": 1.8381510718527988e-06,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 95694640,
+      "step": 4491
+    },
+    {
+      "epoch": 0.5401310647507966,
+      "grad_norm": 2.3130556921590806,
+      "learning_rate": 1.8373746673435812e-06,
+      "loss": 0.6297,
+      "num_input_tokens_seen": 95715385,
+      "step": 4492
+    },
+    {
+      "epoch": 0.5402513076414357,
+      "grad_norm": 3.315596086758766,
+      "learning_rate": 1.8365982875044968e-06,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 95735415,
+      "step": 4493
+    },
+    {
+      "epoch": 0.5403715505320748,
+      "grad_norm": 2.894893736384203,
+      "learning_rate": 1.8358219324533212e-06,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 95755400,
+      "step": 4494
+    },
+    {
+      "epoch": 0.5404917934227139,
+      "grad_norm": 2.3653403758497604,
+      "learning_rate": 1.8350456023078292e-06,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 95777495,
+      "step": 4495
+    },
+    {
+      "epoch": 0.540612036313353,
+      "grad_norm": 2.812174444332911,
+      "learning_rate": 1.8342692971857879e-06,
+      "loss": 0.7732,
+      "num_input_tokens_seen": 95796415,
+      "step": 4496
+    },
+    {
+      "epoch": 0.540732279203992,
+      "grad_norm": 2.8317094289971045,
+      "learning_rate": 1.8334930172049624e-06,
+      "loss": 0.7069,
+      "num_input_tokens_seen": 95816240,
+      "step": 4497
+    },
+    {
+      "epoch": 0.5408525220946312,
+      "grad_norm": 2.522146148860825,
+      "learning_rate": 1.8327167624831134e-06,
+      "loss": 0.7661,
+      "num_input_tokens_seen": 95833690,
+      "step": 4498
+    },
+    {
+      "epoch": 0.5409727649852702,
+      "grad_norm": 1.8753000241432358,
+      "learning_rate": 1.831940533137999e-06,
+      "loss": 0.6998,
+      "num_input_tokens_seen": 95852315,
+      "step": 4499
+    },
+    {
+      "epoch": 0.5410930078759093,
+      "grad_norm": 2.3991958453254223,
+      "learning_rate": 1.8311643292873723e-06,
+      "loss": 0.7149,
+      "num_input_tokens_seen": 95870855,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5412132507665485,
+      "grad_norm": 2.1620078352298546,
+      "learning_rate": 1.8303881510489822e-06,
+      "loss": 0.8782,
+      "num_input_tokens_seen": 95888965,
+      "step": 4501
+    },
+    {
+      "epoch": 0.5413334936571875,
+      "grad_norm": 2.127286289833415,
+      "learning_rate": 1.829611998540574e-06,
+      "loss": 0.6906,
+      "num_input_tokens_seen": 95909890,
+      "step": 4502
+    },
+    {
+      "epoch": 0.5414537365478266,
+      "grad_norm": 10.78141914026094,
+      "learning_rate": 1.828835871879891e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 95928800,
+      "step": 4503
+    },
+    {
+      "epoch": 0.5415739794384657,
+      "grad_norm": 2.41228774844088,
+      "learning_rate": 1.8280597711846703e-06,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 95946760,
+      "step": 4504
+    },
+    {
+      "epoch": 0.5416942223291048,
+      "grad_norm": 2.3437457531770627,
+      "learning_rate": 1.8272836965726455e-06,
+      "loss": 0.8326,
+      "num_input_tokens_seen": 95965415,
+      "step": 4505
+    },
+    {
+      "epoch": 0.5418144652197439,
+      "grad_norm": 4.082541764448734,
+      "learning_rate": 1.8265076481615461e-06,
+      "loss": 0.7748,
+      "num_input_tokens_seen": 95985050,
+      "step": 4506
+    },
+    {
+      "epoch": 0.541934708110383,
+      "grad_norm": 2.4423541465414953,
+      "learning_rate": 1.8257316260690991e-06,
+      "loss": 0.8663,
+      "num_input_tokens_seen": 96002555,
+      "step": 4507
+    },
+    {
+      "epoch": 0.5420549510010221,
+      "grad_norm": 1.6724831441693595,
+      "learning_rate": 1.8249556304130258e-06,
+      "loss": 0.7583,
+      "num_input_tokens_seen": 96023555,
+      "step": 4508
+    },
+    {
+      "epoch": 0.5421751938916611,
+      "grad_norm": 2.6205232693030083,
+      "learning_rate": 1.8241796613110443e-06,
+      "loss": 0.6772,
+      "num_input_tokens_seen": 96042025,
+      "step": 4509
+    },
+    {
+      "epoch": 0.5422954367823003,
+      "grad_norm": 2.394894924883063,
+      "learning_rate": 1.8234037188808676e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 96060505,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5424156796729394,
+      "grad_norm": 2.186440693775744,
+      "learning_rate": 1.822627803240207e-06,
+      "loss": 0.6603,
+      "num_input_tokens_seen": 96082555,
+      "step": 4511
+    },
+    {
+      "epoch": 0.5425359225635784,
+      "grad_norm": 3.1415320784749183,
+      "learning_rate": 1.8218519145067675e-06,
+      "loss": 0.8423,
+      "num_input_tokens_seen": 96097895,
+      "step": 4512
+    },
+    {
+      "epoch": 0.5426561654542175,
+      "grad_norm": 2.4460330056835984,
+      "learning_rate": 1.8210760527982512e-06,
+      "loss": 0.8933,
+      "num_input_tokens_seen": 96117900,
+      "step": 4513
+    },
+    {
+      "epoch": 0.5427764083448566,
+      "grad_norm": 2.6044980489508154,
+      "learning_rate": 1.8203002182323548e-06,
+      "loss": 0.7415,
+      "num_input_tokens_seen": 96135175,
+      "step": 4514
+    },
+    {
+      "epoch": 0.5428966512354957,
+      "grad_norm": 2.4114133682685392,
+      "learning_rate": 1.819524410926773e-06,
+      "loss": 0.7534,
+      "num_input_tokens_seen": 96152575,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5430168941261347,
+      "grad_norm": 3.295945531207163,
+      "learning_rate": 1.8187486309991944e-06,
+      "loss": 0.7673,
+      "num_input_tokens_seen": 96173175,
+      "step": 4516
+    },
+    {
+      "epoch": 0.5431371370167739,
+      "grad_norm": 3.34917098162259,
+      "learning_rate": 1.8179728785673044e-06,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 96191550,
+      "step": 4517
+    },
+    {
+      "epoch": 0.543257379907413,
+      "grad_norm": 2.5901773411606768,
+      "learning_rate": 1.8171971537487834e-06,
+      "loss": 0.7532,
+      "num_input_tokens_seen": 96209920,
+      "step": 4518
+    },
+    {
+      "epoch": 0.543377622798052,
+      "grad_norm": 3.053474104806019,
+      "learning_rate": 1.8164214566613093e-06,
+      "loss": 0.7976,
+      "num_input_tokens_seen": 96228265,
+      "step": 4519
+    },
+    {
+      "epoch": 0.5434978656886912,
+      "grad_norm": 4.070444044362186,
+      "learning_rate": 1.8156457874225547e-06,
+      "loss": 0.6416,
+      "num_input_tokens_seen": 96246445,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5436181085793302,
+      "grad_norm": 2.2413919105501496,
+      "learning_rate": 1.814870146150187e-06,
+      "loss": 0.8002,
+      "num_input_tokens_seen": 96264275,
+      "step": 4521
+    },
+    {
+      "epoch": 0.5437383514699693,
+      "grad_norm": 3.204435880886595,
+      "learning_rate": 1.814094532961871e-06,
+      "loss": 0.7935,
+      "num_input_tokens_seen": 96282570,
+      "step": 4522
+    },
+    {
+      "epoch": 0.5438585943606085,
+      "grad_norm": 4.201664213550183,
+      "learning_rate": 1.8133189479752666e-06,
+      "loss": 0.8269,
+      "num_input_tokens_seen": 96301220,
+      "step": 4523
+    },
+    {
+      "epoch": 0.5439788372512475,
+      "grad_norm": 2.417096310683502,
+      "learning_rate": 1.8125433913080296e-06,
+      "loss": 0.809,
+      "num_input_tokens_seen": 96318640,
+      "step": 4524
+    },
+    {
+      "epoch": 0.5440990801418866,
+      "grad_norm": 2.410189694031847,
+      "learning_rate": 1.811767863077811e-06,
+      "loss": 0.8251,
+      "num_input_tokens_seen": 96337310,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5442193230325257,
+      "grad_norm": 1.7308144573206885,
+      "learning_rate": 1.8109923634022573e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 96357055,
+      "step": 4526
+    },
+    {
+      "epoch": 0.5443395659231648,
+      "grad_norm": 2.3333614558463855,
+      "learning_rate": 1.8102168923990124e-06,
+      "loss": 0.8529,
+      "num_input_tokens_seen": 96370320,
+      "step": 4527
+    },
+    {
+      "epoch": 0.5444598088138038,
+      "grad_norm": 2.381400413939986,
+      "learning_rate": 1.809441450185714e-06,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 96388525,
+      "step": 4528
+    },
+    {
+      "epoch": 0.544580051704443,
+      "grad_norm": 3.6277852093939207,
+      "learning_rate": 1.8086660368799963e-06,
+      "loss": 0.7378,
+      "num_input_tokens_seen": 96406295,
+      "step": 4529
+    },
+    {
+      "epoch": 0.5447002945950821,
+      "grad_norm": 1.8384391980530215,
+      "learning_rate": 1.807890652599488e-06,
+      "loss": 0.7598,
+      "num_input_tokens_seen": 96430400,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5448205374857211,
+      "grad_norm": 3.1386592652514937,
+      "learning_rate": 1.8071152974618156e-06,
+      "loss": 0.8146,
+      "num_input_tokens_seen": 96447920,
+      "step": 4531
+    },
+    {
+      "epoch": 0.5449407803763603,
+      "grad_norm": 3.0048425241233163,
+      "learning_rate": 1.806339971584599e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 96464300,
+      "step": 4532
+    },
+    {
+      "epoch": 0.5450610232669993,
+      "grad_norm": 3.1812703582218136,
+      "learning_rate": 1.805564675085455e-06,
+      "loss": 0.8495,
+      "num_input_tokens_seen": 96483530,
+      "step": 4533
+    },
+    {
+      "epoch": 0.5451812661576384,
+      "grad_norm": 3.0389818427033974,
+      "learning_rate": 1.804789408081994e-06,
+      "loss": 0.8116,
+      "num_input_tokens_seen": 96500500,
+      "step": 4534
+    },
+    {
+      "epoch": 0.5453015090482776,
+      "grad_norm": 0.8047200938155699,
+      "learning_rate": 1.8040141706918258e-06,
+      "loss": 0.6643,
+      "num_input_tokens_seen": 96561460,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5454217519389166,
+      "grad_norm": 3.0918020637606984,
+      "learning_rate": 1.8032389630325525e-06,
+      "loss": 0.7593,
+      "num_input_tokens_seen": 96579930,
+      "step": 4536
+    },
+    {
+      "epoch": 0.5455419948295557,
+      "grad_norm": 1.971746415360197,
+      "learning_rate": 1.8024637852217711e-06,
+      "loss": 0.751,
+      "num_input_tokens_seen": 96599375,
+      "step": 4537
+    },
+    {
+      "epoch": 0.5456622377201948,
+      "grad_norm": 1.890882133359788,
+      "learning_rate": 1.801688637377076e-06,
+      "loss": 0.8409,
+      "num_input_tokens_seen": 96617610,
+      "step": 4538
+    },
+    {
+      "epoch": 0.5457824806108339,
+      "grad_norm": 2.178145395482317,
+      "learning_rate": 1.8009135196160579e-06,
+      "loss": 0.7768,
+      "num_input_tokens_seen": 96636205,
+      "step": 4539
+    },
+    {
+      "epoch": 0.545902723501473,
+      "grad_norm": 4.253572513350592,
+      "learning_rate": 1.8001384320563004e-06,
+      "loss": 0.8348,
+      "num_input_tokens_seen": 96656180,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5460229663921121,
+      "grad_norm": 0.8635187798253072,
+      "learning_rate": 1.7993633748153838e-06,
+      "loss": 0.6133,
+      "num_input_tokens_seen": 96710505,
+      "step": 4541
+    },
+    {
+      "epoch": 0.5461432092827512,
+      "grad_norm": 1.9087505469664325,
+      "learning_rate": 1.7985883480108834e-06,
+      "loss": 0.7198,
+      "num_input_tokens_seen": 96727860,
+      "step": 4542
+    },
+    {
+      "epoch": 0.5462634521733902,
+      "grad_norm": 2.5984113510996187,
+      "learning_rate": 1.797813351760371e-06,
+      "loss": 0.7171,
+      "num_input_tokens_seen": 96749285,
+      "step": 4543
+    },
+    {
+      "epoch": 0.5463836950640293,
+      "grad_norm": 2.004202562171678,
+      "learning_rate": 1.797038386181412e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 96768775,
+      "step": 4544
+    },
+    {
+      "epoch": 0.5465039379546685,
+      "grad_norm": 2.9186098632741073,
+      "learning_rate": 1.7962634513915689e-06,
+      "loss": 0.7374,
+      "num_input_tokens_seen": 96785845,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5466241808453075,
+      "grad_norm": 3.6968052624059795,
+      "learning_rate": 1.7954885475083969e-06,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 96803235,
+      "step": 4546
+    },
+    {
+      "epoch": 0.5467444237359466,
+      "grad_norm": 2.4142627767014875,
+      "learning_rate": 1.7947136746494509e-06,
+      "loss": 0.7345,
+      "num_input_tokens_seen": 96823870,
+      "step": 4547
+    },
+    {
+      "epoch": 0.5468646666265857,
+      "grad_norm": 2.560287438975104,
+      "learning_rate": 1.793938832932277e-06,
+      "loss": 0.8705,
+      "num_input_tokens_seen": 96841700,
+      "step": 4548
+    },
+    {
+      "epoch": 0.5469849095172248,
+      "grad_norm": 2.4788540133021013,
+      "learning_rate": 1.7931640224744185e-06,
+      "loss": 0.692,
+      "num_input_tokens_seen": 96861970,
+      "step": 4549
+    },
+    {
+      "epoch": 0.5471051524078638,
+      "grad_norm": 3.386966434795847,
+      "learning_rate": 1.7923892433934127e-06,
+      "loss": 0.7338,
+      "num_input_tokens_seen": 96882765,
+      "step": 4550
+    },
+    {
+      "epoch": 0.547225395298503,
+      "grad_norm": 2.85734832936579,
+      "learning_rate": 1.7916144958067943e-06,
+      "loss": 0.783,
+      "num_input_tokens_seen": 96900345,
+      "step": 4551
+    },
+    {
+      "epoch": 0.5473456381891421,
+      "grad_norm": 2.3233810216917252,
+      "learning_rate": 1.790839779832091e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 96919800,
+      "step": 4552
+    },
+    {
+      "epoch": 0.5474658810797811,
+      "grad_norm": 2.782166333885925,
+      "learning_rate": 1.790065095586827e-06,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 96939165,
+      "step": 4553
+    },
+    {
+      "epoch": 0.5475861239704203,
+      "grad_norm": 2.45252649003263,
+      "learning_rate": 1.7892904431885198e-06,
+      "loss": 0.7625,
+      "num_input_tokens_seen": 96966060,
+      "step": 4554
+    },
+    {
+      "epoch": 0.5477063668610593,
+      "grad_norm": 2.6418683922961272,
+      "learning_rate": 1.788515822754686e-06,
+      "loss": 0.7549,
+      "num_input_tokens_seen": 96986200,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5478266097516984,
+      "grad_norm": 3.639128229011245,
+      "learning_rate": 1.7877412344028335e-06,
+      "loss": 0.7842,
+      "num_input_tokens_seen": 97005725,
+      "step": 4556
+    },
+    {
+      "epoch": 0.5479468526423376,
+      "grad_norm": 2.66724802219536,
+      "learning_rate": 1.7869666782504673e-06,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 97022025,
+      "step": 4557
+    },
+    {
+      "epoch": 0.5480670955329766,
+      "grad_norm": 2.1979104104713496,
+      "learning_rate": 1.7861921544150862e-06,
+      "loss": 0.6906,
+      "num_input_tokens_seen": 97040595,
+      "step": 4558
+    },
+    {
+      "epoch": 0.5481873384236157,
+      "grad_norm": 5.390991213395312,
+      "learning_rate": 1.7854176630141856e-06,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 97057450,
+      "step": 4559
+    },
+    {
+      "epoch": 0.5483075813142548,
+      "grad_norm": 2.8387650721681896,
+      "learning_rate": 1.784643204165255e-06,
+      "loss": 0.8408,
+      "num_input_tokens_seen": 97076490,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5484278242048939,
+      "grad_norm": 2.129319554576394,
+      "learning_rate": 1.7838687779857788e-06,
+      "loss": 0.7628,
+      "num_input_tokens_seen": 97094085,
+      "step": 4561
+    },
+    {
+      "epoch": 0.5485480670955329,
+      "grad_norm": 6.801854309750558,
+      "learning_rate": 1.7830943845932366e-06,
+      "loss": 0.6376,
+      "num_input_tokens_seen": 97113130,
+      "step": 4562
+    },
+    {
+      "epoch": 0.5486683099861721,
+      "grad_norm": 2.0142045892946583,
+      "learning_rate": 1.7823200241051044e-06,
+      "loss": 0.7422,
+      "num_input_tokens_seen": 97131765,
+      "step": 4563
+    },
+    {
+      "epoch": 0.5487885528768112,
+      "grad_norm": 2.548632231539975,
+      "learning_rate": 1.7815456966388513e-06,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 97150580,
+      "step": 4564
+    },
+    {
+      "epoch": 0.5489087957674502,
+      "grad_norm": 2.328563739640445,
+      "learning_rate": 1.780771402311943e-06,
+      "loss": 0.8162,
+      "num_input_tokens_seen": 97169135,
+      "step": 4565
+    },
+    {
+      "epoch": 0.5490290386580894,
+      "grad_norm": 7.859507943904961,
+      "learning_rate": 1.7799971412418374e-06,
+      "loss": 0.7834,
+      "num_input_tokens_seen": 97190250,
+      "step": 4566
+    },
+    {
+      "epoch": 0.5491492815487284,
+      "grad_norm": 2.65226546756694,
+      "learning_rate": 1.7792229135459918e-06,
+      "loss": 0.7378,
+      "num_input_tokens_seen": 97206620,
+      "step": 4567
+    },
+    {
+      "epoch": 0.5492695244393675,
+      "grad_norm": 0.8319925777710446,
+      "learning_rate": 1.7784487193418542e-06,
+      "loss": 0.6405,
+      "num_input_tokens_seen": 97264190,
+      "step": 4568
+    },
+    {
+      "epoch": 0.5493897673300067,
+      "grad_norm": 4.044800536792994,
+      "learning_rate": 1.7776745587468698e-06,
+      "loss": 0.606,
+      "num_input_tokens_seen": 97281335,
+      "step": 4569
+    },
+    {
+      "epoch": 0.5495100102206457,
+      "grad_norm": 2.619569129379258,
+      "learning_rate": 1.7769004318784772e-06,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 97298700,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5496302531112848,
+      "grad_norm": 1.818039195860528,
+      "learning_rate": 1.7761263388541125e-06,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 97316210,
+      "step": 4571
+    },
+    {
+      "epoch": 0.5497504960019239,
+      "grad_norm": 1.9504988214525416,
+      "learning_rate": 1.7753522797912044e-06,
+      "loss": 0.8335,
+      "num_input_tokens_seen": 97336015,
+      "step": 4572
+    },
+    {
+      "epoch": 0.549870738892563,
+      "grad_norm": 2.4627141351404487,
+      "learning_rate": 1.7745782548071769e-06,
+      "loss": 0.7017,
+      "num_input_tokens_seen": 97352630,
+      "step": 4573
+    },
+    {
+      "epoch": 0.549990981783202,
+      "grad_norm": 1.7276097396731485,
+      "learning_rate": 1.7738042640194482e-06,
+      "loss": 0.7331,
+      "num_input_tokens_seen": 97372015,
+      "step": 4574
+    },
+    {
+      "epoch": 0.5501112246738411,
+      "grad_norm": 1.9824607268051193,
+      "learning_rate": 1.7730303075454335e-06,
+      "loss": 0.7033,
+      "num_input_tokens_seen": 97390625,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5502314675644803,
+      "grad_norm": 2.6737970907316426,
+      "learning_rate": 1.7722563855025402e-06,
+      "loss": 0.8421,
+      "num_input_tokens_seen": 97408375,
+      "step": 4576
+    },
+    {
+      "epoch": 0.5503517104551193,
+      "grad_norm": 2.621610651640046,
+      "learning_rate": 1.7714824980081725e-06,
+      "loss": 0.7012,
+      "num_input_tokens_seen": 97427390,
+      "step": 4577
+    },
+    {
+      "epoch": 0.5504719533457584,
+      "grad_norm": 2.4628521470387637,
+      "learning_rate": 1.7707086451797272e-06,
+      "loss": 0.7359,
+      "num_input_tokens_seen": 97447985,
+      "step": 4578
+    },
+    {
+      "epoch": 0.5505921962363975,
+      "grad_norm": 0.7072043028030115,
+      "learning_rate": 1.7699348271345997e-06,
+      "loss": 0.5395,
+      "num_input_tokens_seen": 97510330,
+      "step": 4579
+    },
+    {
+      "epoch": 0.5507124391270366,
+      "grad_norm": 0.7212288128862477,
+      "learning_rate": 1.7691610439901753e-06,
+      "loss": 0.5641,
+      "num_input_tokens_seen": 97572985,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5508326820176757,
+      "grad_norm": 1.9523288888624444,
+      "learning_rate": 1.768387295863837e-06,
+      "loss": 0.7562,
+      "num_input_tokens_seen": 97591585,
+      "step": 4581
+    },
+    {
+      "epoch": 0.5509529249083148,
+      "grad_norm": 2.277103540951851,
+      "learning_rate": 1.767613582872961e-06,
+      "loss": 0.8372,
+      "num_input_tokens_seen": 97611015,
+      "step": 4582
+    },
+    {
+      "epoch": 0.5510731677989539,
+      "grad_norm": 2.468157347012842,
+      "learning_rate": 1.7668399051349205e-06,
+      "loss": 0.8248,
+      "num_input_tokens_seen": 97630415,
+      "step": 4583
+    },
+    {
+      "epoch": 0.5511934106895929,
+      "grad_norm": 4.3533864891513225,
+      "learning_rate": 1.766066262767081e-06,
+      "loss": 0.8196,
+      "num_input_tokens_seen": 97647975,
+      "step": 4584
+    },
+    {
+      "epoch": 0.5513136535802321,
+      "grad_norm": 3.9866515939481806,
+      "learning_rate": 1.7652926558868035e-06,
+      "loss": 0.7768,
+      "num_input_tokens_seen": 97666340,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5514338964708712,
+      "grad_norm": 3.6574408166005337,
+      "learning_rate": 1.764519084611443e-06,
+      "loss": 0.7028,
+      "num_input_tokens_seen": 97686515,
+      "step": 4586
+    },
+    {
+      "epoch": 0.5515541393615102,
+      "grad_norm": 2.6701065441217184,
+      "learning_rate": 1.7637455490583505e-06,
+      "loss": 0.7719,
+      "num_input_tokens_seen": 97705560,
+      "step": 4587
+    },
+    {
+      "epoch": 0.5516743822521494,
+      "grad_norm": 2.600611771425595,
+      "learning_rate": 1.7629720493448706e-06,
+      "loss": 0.7679,
+      "num_input_tokens_seen": 97722575,
+      "step": 4588
+    },
+    {
+      "epoch": 0.5517946251427884,
+      "grad_norm": 2.7786709025685976,
+      "learning_rate": 1.7621985855883422e-06,
+      "loss": 0.8436,
+      "num_input_tokens_seen": 97738995,
+      "step": 4589
+    },
+    {
+      "epoch": 0.5519148680334275,
+      "grad_norm": 2.1653372250359006,
+      "learning_rate": 1.7614251579060983e-06,
+      "loss": 0.7245,
+      "num_input_tokens_seen": 97757310,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5520351109240667,
+      "grad_norm": 2.1356490786964986,
+      "learning_rate": 1.7606517664154693e-06,
+      "loss": 0.8404,
+      "num_input_tokens_seen": 97779740,
+      "step": 4591
+    },
+    {
+      "epoch": 0.5521553538147057,
+      "grad_norm": 2.009791898315628,
+      "learning_rate": 1.759878411233777e-06,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 97797920,
+      "step": 4592
+    },
+    {
+      "epoch": 0.5522755967053448,
+      "grad_norm": 2.7608515867367402,
+      "learning_rate": 1.7591050924783388e-06,
+      "loss": 0.7529,
+      "num_input_tokens_seen": 97814830,
+      "step": 4593
+    },
+    {
+      "epoch": 0.5523958395959839,
+      "grad_norm": 0.9190290365883461,
+      "learning_rate": 1.7583318102664661e-06,
+      "loss": 0.6283,
+      "num_input_tokens_seen": 97882115,
+      "step": 4594
+    },
+    {
+      "epoch": 0.552516082486623,
+      "grad_norm": 2.1414458485954846,
+      "learning_rate": 1.757558564715466e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 97899910,
+      "step": 4595
+    },
+    {
+      "epoch": 0.552636325377262,
+      "grad_norm": 3.3702286309477674,
+      "learning_rate": 1.756785355942639e-06,
+      "loss": 0.7369,
+      "num_input_tokens_seen": 97916680,
+      "step": 4596
+    },
+    {
+      "epoch": 0.5527565682679012,
+      "grad_norm": 2.3405612207399407,
+      "learning_rate": 1.7560121840652801e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 97935785,
+      "step": 4597
+    },
+    {
+      "epoch": 0.5528768111585403,
+      "grad_norm": 3.4159267622632257,
+      "learning_rate": 1.7552390492006778e-06,
+      "loss": 0.6939,
+      "num_input_tokens_seen": 97953825,
+      "step": 4598
+    },
+    {
+      "epoch": 0.5529970540491793,
+      "grad_norm": 2.393271359955657,
+      "learning_rate": 1.7544659514661184e-06,
+      "loss": 0.6517,
+      "num_input_tokens_seen": 97976635,
+      "step": 4599
+    },
+    {
+      "epoch": 0.5531172969398185,
+      "grad_norm": 2.2223138923828514,
+      "learning_rate": 1.7536928909788786e-06,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 97995660,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5532375398304575,
+      "grad_norm": 0.9079854887168407,
+      "learning_rate": 1.7529198678562317e-06,
+      "loss": 0.6382,
+      "num_input_tokens_seen": 98047025,
+      "step": 4601
+    },
+    {
+      "epoch": 0.5533577827210966,
+      "grad_norm": 2.287128317680286,
+      "learning_rate": 1.7521468822154436e-06,
+      "loss": 0.7744,
+      "num_input_tokens_seen": 98065660,
+      "step": 4602
+    },
+    {
+      "epoch": 0.5534780256117358,
+      "grad_norm": 2.131467136692197,
+      "learning_rate": 1.751373934173777e-06,
+      "loss": 0.7351,
+      "num_input_tokens_seen": 98088125,
+      "step": 4603
+    },
+    {
+      "epoch": 0.5535982685023748,
+      "grad_norm": 2.6375927004399946,
+      "learning_rate": 1.750601023848487e-06,
+      "loss": 0.7277,
+      "num_input_tokens_seen": 98108570,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5537185113930139,
+      "grad_norm": 7.943641189541467,
+      "learning_rate": 1.7498281513568233e-06,
+      "loss": 0.7341,
+      "num_input_tokens_seen": 98128485,
+      "step": 4605
+    },
+    {
+      "epoch": 0.553838754283653,
+      "grad_norm": 1.9589051071615995,
+      "learning_rate": 1.7490553168160292e-06,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 98149275,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5539589971742921,
+      "grad_norm": 2.751107397056313,
+      "learning_rate": 1.748282520343345e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 98168025,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5540792400649311,
+      "grad_norm": 4.7398017392329574,
+      "learning_rate": 1.7475097620560023e-06,
+      "loss": 0.7828,
+      "num_input_tokens_seen": 98187810,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5541994829555702,
+      "grad_norm": 2.32946365279131,
+      "learning_rate": 1.746737042071228e-06,
+      "loss": 0.6955,
+      "num_input_tokens_seen": 98206035,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5543197258462094,
+      "grad_norm": 2.286728538534861,
+      "learning_rate": 1.7459643605062424e-06,
+      "loss": 0.7921,
+      "num_input_tokens_seen": 98223015,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5544399687368484,
+      "grad_norm": 3.617949768722936,
+      "learning_rate": 1.745191717478262e-06,
+      "loss": 0.807,
+      "num_input_tokens_seen": 98241315,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5545602116274875,
+      "grad_norm": 2.142886975268269,
+      "learning_rate": 1.7444191131044952e-06,
+      "loss": 0.7961,
+      "num_input_tokens_seen": 98261310,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5546804545181266,
+      "grad_norm": 2.0224893838038374,
+      "learning_rate": 1.743646547502146e-06,
+      "loss": 0.7178,
+      "num_input_tokens_seen": 98281080,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5548006974087657,
+      "grad_norm": 2.399871588061633,
+      "learning_rate": 1.7428740207884107e-06,
+      "loss": 0.7002,
+      "num_input_tokens_seen": 98301680,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5549209402994048,
+      "grad_norm": 2.1087184573684725,
+      "learning_rate": 1.742101533080483e-06,
+      "loss": 0.6115,
+      "num_input_tokens_seen": 98321320,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5550411831900439,
+      "grad_norm": 2.0228526046229565,
+      "learning_rate": 1.7413290844955475e-06,
+      "loss": 0.7237,
+      "num_input_tokens_seen": 98341070,
+      "step": 4616
+    },
+    {
+      "epoch": 0.555161426080683,
+      "grad_norm": 2.468785719319241,
+      "learning_rate": 1.7405566751507848e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 98358835,
+      "step": 4617
+    },
+    {
+      "epoch": 0.555281668971322,
+      "grad_norm": 1.8245892629041438,
+      "learning_rate": 1.7397843051633668e-06,
+      "loss": 0.6721,
+      "num_input_tokens_seen": 98381250,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5554019118619612,
+      "grad_norm": 1.7072485014369077,
+      "learning_rate": 1.739011974650464e-06,
+      "loss": 0.7074,
+      "num_input_tokens_seen": 98400300,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5555221547526003,
+      "grad_norm": 3.4370309849372958,
+      "learning_rate": 1.7382396837292365e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 98420480,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5556423976432393,
+      "grad_norm": 2.055057315638427,
+      "learning_rate": 1.7374674325168414e-06,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 98440300,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5557626405338785,
+      "grad_norm": 2.6919249385100374,
+      "learning_rate": 1.7366952211304274e-06,
+      "loss": 0.7354,
+      "num_input_tokens_seen": 98457865,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5558828834245175,
+      "grad_norm": 4.059086203916531,
+      "learning_rate": 1.7359230496871392e-06,
+      "loss": 0.8252,
+      "num_input_tokens_seen": 98474160,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5560031263151566,
+      "grad_norm": 1.699107537094162,
+      "learning_rate": 1.7351509183041149e-06,
+      "loss": 0.7433,
+      "num_input_tokens_seen": 98494210,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5561233692057957,
+      "grad_norm": 2.3238791412693143,
+      "learning_rate": 1.7343788270984856e-06,
+      "loss": 0.7222,
+      "num_input_tokens_seen": 98513070,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5562436120964348,
+      "grad_norm": 1.932713218314441,
+      "learning_rate": 1.733606776187376e-06,
+      "loss": 0.7315,
+      "num_input_tokens_seen": 98535215,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5563638549870739,
+      "grad_norm": 2.7704153670930913,
+      "learning_rate": 1.7328347656879076e-06,
+      "loss": 0.7553,
+      "num_input_tokens_seen": 98554795,
+      "step": 4627
+    },
+    {
+      "epoch": 0.556484097877713,
+      "grad_norm": 3.145226618649204,
+      "learning_rate": 1.7320627957171927e-06,
+      "loss": 0.6764,
+      "num_input_tokens_seen": 98569175,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5566043407683521,
+      "grad_norm": 1.8068028262777365,
+      "learning_rate": 1.7312908663923386e-06,
+      "loss": 0.7995,
+      "num_input_tokens_seen": 98585070,
+      "step": 4629
+    },
+    {
+      "epoch": 0.5567245836589911,
+      "grad_norm": 2.630490418375558,
+      "learning_rate": 1.7305189778304463e-06,
+      "loss": 0.6675,
+      "num_input_tokens_seen": 98602965,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5568448265496303,
+      "grad_norm": 2.587094026970457,
+      "learning_rate": 1.729747130148611e-06,
+      "loss": 0.7959,
+      "num_input_tokens_seen": 98621880,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5569650694402694,
+      "grad_norm": 2.4483123974819576,
+      "learning_rate": 1.7289753234639213e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 98640575,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5570853123309084,
+      "grad_norm": 1.92245706126497,
+      "learning_rate": 1.7282035578934596e-06,
+      "loss": 0.7517,
+      "num_input_tokens_seen": 98658460,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5572055552215476,
+      "grad_norm": 1.9486684671640153,
+      "learning_rate": 1.727431833554301e-06,
+      "loss": 0.7814,
+      "num_input_tokens_seen": 98676655,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5573257981121866,
+      "grad_norm": 2.348891844585534,
+      "learning_rate": 1.7266601505635175e-06,
+      "loss": 0.7665,
+      "num_input_tokens_seen": 98693715,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5574460410028257,
+      "grad_norm": 2.399017925617153,
+      "learning_rate": 1.7258885090381717e-06,
+      "loss": 0.757,
+      "num_input_tokens_seen": 98711475,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5575662838934649,
+      "grad_norm": 2.081514680426162,
+      "learning_rate": 1.7251169090953213e-06,
+      "loss": 0.7795,
+      "num_input_tokens_seen": 98731670,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5576865267841039,
+      "grad_norm": 2.856826903110264,
+      "learning_rate": 1.7243453508520168e-06,
+      "loss": 0.758,
+      "num_input_tokens_seen": 98748375,
+      "step": 4638
+    },
+    {
+      "epoch": 0.557806769674743,
+      "grad_norm": 2.4484151050637473,
+      "learning_rate": 1.7235738344253038e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 98761725,
+      "step": 4639
+    },
+    {
+      "epoch": 0.557927012565382,
+      "grad_norm": 2.2418588312995844,
+      "learning_rate": 1.7228023599322204e-06,
+      "loss": 0.8172,
+      "num_input_tokens_seen": 98779750,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5580472554560212,
+      "grad_norm": 2.6003058315021583,
+      "learning_rate": 1.7220309274897983e-06,
+      "loss": 0.6932,
+      "num_input_tokens_seen": 98796750,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5581674983466602,
+      "grad_norm": 2.288405646894083,
+      "learning_rate": 1.721259537215063e-06,
+      "loss": 0.7386,
+      "num_input_tokens_seen": 98816450,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5582877412372993,
+      "grad_norm": 3.3409610807535572,
+      "learning_rate": 1.720488189225035e-06,
+      "loss": 0.7296,
+      "num_input_tokens_seen": 98833870,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5584079841279385,
+      "grad_norm": 2.7983596583535006,
+      "learning_rate": 1.7197168836367265e-06,
+      "loss": 0.788,
+      "num_input_tokens_seen": 98850400,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5585282270185775,
+      "grad_norm": 3.5772989624295537,
+      "learning_rate": 1.7189456205671437e-06,
+      "loss": 0.8143,
+      "num_input_tokens_seen": 98868965,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5586484699092166,
+      "grad_norm": 2.7775518455902763,
+      "learning_rate": 1.7181744001332866e-06,
+      "loss": 0.8142,
+      "num_input_tokens_seen": 98887295,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5587687127998557,
+      "grad_norm": 2.328807090449223,
+      "learning_rate": 1.7174032224521493e-06,
+      "loss": 0.6336,
+      "num_input_tokens_seen": 98905725,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5588889556904948,
+      "grad_norm": 2.440446316656876,
+      "learning_rate": 1.7166320876407184e-06,
+      "loss": 0.6936,
+      "num_input_tokens_seen": 98924865,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5590091985811338,
+      "grad_norm": 2.553572799425318,
+      "learning_rate": 1.7158609958159746e-06,
+      "loss": 0.674,
+      "num_input_tokens_seen": 98941990,
+      "step": 4649
+    },
+    {
+      "epoch": 0.559129441471773,
+      "grad_norm": 2.433166228039536,
+      "learning_rate": 1.7150899470948907e-06,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 98956975,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5592496843624121,
+      "grad_norm": 0.8130248712780672,
+      "learning_rate": 1.7143189415944365e-06,
+      "loss": 0.5968,
+      "num_input_tokens_seen": 99021155,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5593699272530511,
+      "grad_norm": 2.100087619942274,
+      "learning_rate": 1.7135479794315714e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 99037830,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5594901701436903,
+      "grad_norm": 2.957547300753262,
+      "learning_rate": 1.7127770607232502e-06,
+      "loss": 0.78,
+      "num_input_tokens_seen": 99056095,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5596104130343293,
+      "grad_norm": 3.2637453569110533,
+      "learning_rate": 1.7120061855864204e-06,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 99075825,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5597306559249684,
+      "grad_norm": 2.186395482101678,
+      "learning_rate": 1.7112353541380233e-06,
+      "loss": 0.7154,
+      "num_input_tokens_seen": 99095405,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5598508988156076,
+      "grad_norm": 1.729936473836241,
+      "learning_rate": 1.7104645664949935e-06,
+      "loss": 0.7172,
+      "num_input_tokens_seen": 99117595,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5599711417062466,
+      "grad_norm": 3.5900431691580956,
+      "learning_rate": 1.7096938227742588e-06,
+      "loss": 0.7211,
+      "num_input_tokens_seen": 99138445,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5600913845968857,
+      "grad_norm": 2.0502035553852913,
+      "learning_rate": 1.7089231230927395e-06,
+      "loss": 0.8315,
+      "num_input_tokens_seen": 99156055,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5602116274875248,
+      "grad_norm": 2.7385146753783998,
+      "learning_rate": 1.7081524675673518e-06,
+      "loss": 0.6723,
+      "num_input_tokens_seen": 99171265,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5603318703781639,
+      "grad_norm": 0.893477876897292,
+      "learning_rate": 1.707381856315003e-06,
+      "loss": 0.6414,
+      "num_input_tokens_seen": 99233065,
+      "step": 4660
+    },
+    {
+      "epoch": 0.560452113268803,
+      "grad_norm": 2.9800139101710004,
+      "learning_rate": 1.706611289452594e-06,
+      "loss": 0.8646,
+      "num_input_tokens_seen": 99250865,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5605723561594421,
+      "grad_norm": 2.126245528566991,
+      "learning_rate": 1.7058407670970177e-06,
+      "loss": 0.7235,
+      "num_input_tokens_seen": 99272060,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5606925990500812,
+      "grad_norm": 3.498600359393606,
+      "learning_rate": 1.7050702893651643e-06,
+      "loss": 0.611,
+      "num_input_tokens_seen": 99291360,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5608128419407202,
+      "grad_norm": 2.6478961158511494,
+      "learning_rate": 1.7042998563739134e-06,
+      "loss": 0.7474,
+      "num_input_tokens_seen": 99309430,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5609330848313594,
+      "grad_norm": 2.677276122384816,
+      "learning_rate": 1.7035294682401394e-06,
+      "loss": 0.7097,
+      "num_input_tokens_seen": 99328020,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5610533277219985,
+      "grad_norm": 3.2005083765255287,
+      "learning_rate": 1.7027591250807088e-06,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 99344915,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5611735706126375,
+      "grad_norm": 2.91488459926229,
+      "learning_rate": 1.701988827012483e-06,
+      "loss": 0.841,
+      "num_input_tokens_seen": 99361800,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5612938135032767,
+      "grad_norm": 3.880008334058756,
+      "learning_rate": 1.701218574152315e-06,
+      "loss": 0.8098,
+      "num_input_tokens_seen": 99377845,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5614140563939157,
+      "grad_norm": 2.3294073073337755,
+      "learning_rate": 1.700448366617052e-06,
+      "loss": 0.6208,
+      "num_input_tokens_seen": 99398060,
+      "step": 4669
+    },
+    {
+      "epoch": 0.5615342992845548,
+      "grad_norm": 3.087004830021334,
+      "learning_rate": 1.6996782045235326e-06,
+      "loss": 0.8037,
+      "num_input_tokens_seen": 99417645,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5616545421751938,
+      "grad_norm": 4.244046296640389,
+      "learning_rate": 1.6989080879885918e-06,
+      "loss": 0.6879,
+      "num_input_tokens_seen": 99435225,
+      "step": 4671
+    },
+    {
+      "epoch": 0.561774785065833,
+      "grad_norm": 0.995728679876948,
+      "learning_rate": 1.6981380171290544e-06,
+      "loss": 0.6373,
+      "num_input_tokens_seen": 99495970,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5618950279564721,
+      "grad_norm": 2.6484523022181303,
+      "learning_rate": 1.69736799206174e-06,
+      "loss": 0.7384,
+      "num_input_tokens_seen": 99513225,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5620152708471111,
+      "grad_norm": 10.702345713206823,
+      "learning_rate": 1.6965980129034603e-06,
+      "loss": 0.8471,
+      "num_input_tokens_seen": 99530330,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5621355137377503,
+      "grad_norm": 1.6272927703435291,
+      "learning_rate": 1.6958280797710209e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 99551975,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5622557566283893,
+      "grad_norm": 0.748868144638413,
+      "learning_rate": 1.6950581927812202e-06,
+      "loss": 0.5574,
+      "num_input_tokens_seen": 99611265,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5623759995190284,
+      "grad_norm": 2.8826021047167494,
+      "learning_rate": 1.694288352050849e-06,
+      "loss": 0.7848,
+      "num_input_tokens_seen": 99629720,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5624962424096676,
+      "grad_norm": 2.6061119480622605,
+      "learning_rate": 1.693518557696691e-06,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 99648580,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5626164853003066,
+      "grad_norm": 2.3687967539788364,
+      "learning_rate": 1.6927488098355252e-06,
+      "loss": 0.8802,
+      "num_input_tokens_seen": 99665930,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5627367281909457,
+      "grad_norm": 0.9109420691961535,
+      "learning_rate": 1.6919791085841201e-06,
+      "loss": 0.6759,
+      "num_input_tokens_seen": 99723060,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5628569710815848,
+      "grad_norm": 3.486596242675488,
+      "learning_rate": 1.6912094540592396e-06,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 99738300,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5629772139722239,
+      "grad_norm": 2.474714027044474,
+      "learning_rate": 1.6904398463776393e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 99751820,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5630974568628629,
+      "grad_norm": 2.253690521874694,
+      "learning_rate": 1.6896702856560683e-06,
+      "loss": 0.7214,
+      "num_input_tokens_seen": 99770635,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5632176997535021,
+      "grad_norm": 4.310262499383267,
+      "learning_rate": 1.688900772011268e-06,
+      "loss": 0.6884,
+      "num_input_tokens_seen": 99788100,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5633379426441412,
+      "grad_norm": 1.8696980579151923,
+      "learning_rate": 1.6881313055599734e-06,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 99807750,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5634581855347802,
+      "grad_norm": 3.1088012571700845,
+      "learning_rate": 1.687361886418911e-06,
+      "loss": 0.809,
+      "num_input_tokens_seen": 99823240,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5635784284254194,
+      "grad_norm": 3.5329863794713483,
+      "learning_rate": 1.686592514704803e-06,
+      "loss": 0.7769,
+      "num_input_tokens_seen": 99840355,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5636986713160584,
+      "grad_norm": 3.007595584657204,
+      "learning_rate": 1.685823190534361e-06,
+      "loss": 0.6897,
+      "num_input_tokens_seen": 99858315,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5638189142066975,
+      "grad_norm": 2.2467799706758105,
+      "learning_rate": 1.6850539140242913e-06,
+      "loss": 0.8259,
+      "num_input_tokens_seen": 99877295,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5639391570973367,
+      "grad_norm": 2.1387424014555334,
+      "learning_rate": 1.684284685291292e-06,
+      "loss": 0.8125,
+      "num_input_tokens_seen": 99898660,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5640593999879757,
+      "grad_norm": 3.4426414998710575,
+      "learning_rate": 1.683515504452055e-06,
+      "loss": 0.7992,
+      "num_input_tokens_seen": 99915755,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5641796428786148,
+      "grad_norm": 2.0982403644815433,
+      "learning_rate": 1.6827463716232648e-06,
+      "loss": 0.6572,
+      "num_input_tokens_seen": 99936135,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5642998857692539,
+      "grad_norm": 2.5682011332678822,
+      "learning_rate": 1.6819772869215976e-06,
+      "loss": 0.744,
+      "num_input_tokens_seen": 99954935,
+      "step": 4693
+    },
+    {
+      "epoch": 0.564420128659893,
+      "grad_norm": 2.0625392868648333,
+      "learning_rate": 1.6812082504637223e-06,
+      "loss": 0.8152,
+      "num_input_tokens_seen": 99975975,
+      "step": 4694
+    },
+    {
+      "epoch": 0.564540371550532,
+      "grad_norm": 1.8386542305886315,
+      "learning_rate": 1.6804392623663025e-06,
+      "loss": 0.7414,
+      "num_input_tokens_seen": 99996900,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5646606144411712,
+      "grad_norm": 2.268285779114741,
+      "learning_rate": 1.679670322745993e-06,
+      "loss": 0.7724,
+      "num_input_tokens_seen": 100014575,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5647808573318103,
+      "grad_norm": 2.3209607325838655,
+      "learning_rate": 1.6789014317194407e-06,
+      "loss": 0.752,
+      "num_input_tokens_seen": 100035775,
+      "step": 4697
+    },
+    {
+      "epoch": 0.5649011002224493,
+      "grad_norm": 2.9827854843983523,
+      "learning_rate": 1.6781325894032853e-06,
+      "loss": 0.7236,
+      "num_input_tokens_seen": 100054455,
+      "step": 4698
+    },
+    {
+      "epoch": 0.5650213431130885,
+      "grad_norm": 2.5487859154302153,
+      "learning_rate": 1.6773637959141608e-06,
+      "loss": 0.9062,
+      "num_input_tokens_seen": 100071150,
+      "step": 4699
+    },
+    {
+      "epoch": 0.5651415860037275,
+      "grad_norm": 4.150395540936103,
+      "learning_rate": 1.6765950513686917e-06,
+      "loss": 0.6591,
+      "num_input_tokens_seen": 100088980,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5652618288943666,
+      "grad_norm": 2.012919905937243,
+      "learning_rate": 1.6758263558834963e-06,
+      "loss": 0.7625,
+      "num_input_tokens_seen": 100107915,
+      "step": 4701
+    },
+    {
+      "epoch": 0.5653820717850057,
+      "grad_norm": 3.0719535535141844,
+      "learning_rate": 1.6750577095751844e-06,
+      "loss": 0.7886,
+      "num_input_tokens_seen": 100126745,
+      "step": 4702
+    },
+    {
+      "epoch": 0.5655023146756448,
+      "grad_norm": 2.1765740281078574,
+      "learning_rate": 1.67428911256036e-06,
+      "loss": 0.7247,
+      "num_input_tokens_seen": 100147370,
+      "step": 4703
+    },
+    {
+      "epoch": 0.5656225575662839,
+      "grad_norm": 2.1326518121074014,
+      "learning_rate": 1.673520564955619e-06,
+      "loss": 0.7114,
+      "num_input_tokens_seen": 100166960,
+      "step": 4704
+    },
+    {
+      "epoch": 0.5657428004569229,
+      "grad_norm": 1.9497628301241579,
+      "learning_rate": 1.672752066877548e-06,
+      "loss": 0.8387,
+      "num_input_tokens_seen": 100186965,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5658630433475621,
+      "grad_norm": 1.8369030137304878,
+      "learning_rate": 1.6719836184427275e-06,
+      "loss": 0.7472,
+      "num_input_tokens_seen": 100206990,
+      "step": 4706
+    },
+    {
+      "epoch": 0.5659832862382012,
+      "grad_norm": 2.1361175471219203,
+      "learning_rate": 1.6712152197677325e-06,
+      "loss": 0.6434,
+      "num_input_tokens_seen": 100226170,
+      "step": 4707
+    },
+    {
+      "epoch": 0.5661035291288402,
+      "grad_norm": 2.280137482340126,
+      "learning_rate": 1.670446870969127e-06,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 100243555,
+      "step": 4708
+    },
+    {
+      "epoch": 0.5662237720194794,
+      "grad_norm": 2.8577436853610183,
+      "learning_rate": 1.669678572163469e-06,
+      "loss": 0.8032,
+      "num_input_tokens_seen": 100257760,
+      "step": 4709
+    },
+    {
+      "epoch": 0.5663440149101184,
+      "grad_norm": 2.6087569199749625,
+      "learning_rate": 1.6689103234673086e-06,
+      "loss": 0.7228,
+      "num_input_tokens_seen": 100275800,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5664642578007575,
+      "grad_norm": 3.010724151913435,
+      "learning_rate": 1.6681421249971895e-06,
+      "loss": 0.7666,
+      "num_input_tokens_seen": 100295180,
+      "step": 4711
+    },
+    {
+      "epoch": 0.5665845006913967,
+      "grad_norm": 0.7763559321278201,
+      "learning_rate": 1.6673739768696457e-06,
+      "loss": 0.6276,
+      "num_input_tokens_seen": 100361470,
+      "step": 4712
+    },
+    {
+      "epoch": 0.5667047435820357,
+      "grad_norm": 1.8815636913253833,
+      "learning_rate": 1.6666058792012056e-06,
+      "loss": 0.7681,
+      "num_input_tokens_seen": 100382075,
+      "step": 4713
+    },
+    {
+      "epoch": 0.5668249864726748,
+      "grad_norm": 0.8886862743750386,
+      "learning_rate": 1.6658378321083874e-06,
+      "loss": 0.7013,
+      "num_input_tokens_seen": 100446125,
+      "step": 4714
+    },
+    {
+      "epoch": 0.5669452293633139,
+      "grad_norm": 3.766664952998253,
+      "learning_rate": 1.6650698357077055e-06,
+      "loss": 0.8148,
+      "num_input_tokens_seen": 100462890,
+      "step": 4715
+    },
+    {
+      "epoch": 0.567065472253953,
+      "grad_norm": 2.8099670810823345,
+      "learning_rate": 1.6643018901156632e-06,
+      "loss": 0.8,
+      "num_input_tokens_seen": 100481705,
+      "step": 4716
+    },
+    {
+      "epoch": 0.567185715144592,
+      "grad_norm": 2.94318610588146,
+      "learning_rate": 1.663533995448757e-06,
+      "loss": 0.7817,
+      "num_input_tokens_seen": 100497300,
+      "step": 4717
+    },
+    {
+      "epoch": 0.5673059580352312,
+      "grad_norm": 1.892527479553599,
+      "learning_rate": 1.6627661518234758e-06,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 100516275,
+      "step": 4718
+    },
+    {
+      "epoch": 0.5674262009258703,
+      "grad_norm": 2.044476209301902,
+      "learning_rate": 1.661998359356302e-06,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 100535025,
+      "step": 4719
+    },
+    {
+      "epoch": 0.5675464438165093,
+      "grad_norm": 0.841166206280218,
+      "learning_rate": 1.6612306181637081e-06,
+      "loss": 0.5928,
+      "num_input_tokens_seen": 100594070,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5676666867071485,
+      "grad_norm": 2.6185538357674534,
+      "learning_rate": 1.6604629283621604e-06,
+      "loss": 0.6586,
+      "num_input_tokens_seen": 100611720,
+      "step": 4721
+    },
+    {
+      "epoch": 0.5677869295977875,
+      "grad_norm": 2.3964198008744093,
+      "learning_rate": 1.6596952900681152e-06,
+      "loss": 0.746,
+      "num_input_tokens_seen": 100632200,
+      "step": 4722
+    },
+    {
+      "epoch": 0.5679071724884266,
+      "grad_norm": 2.5021946362775016,
+      "learning_rate": 1.658927703398025e-06,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 100651985,
+      "step": 4723
+    },
+    {
+      "epoch": 0.5680274153790658,
+      "grad_norm": 2.3453428445957787,
+      "learning_rate": 1.6581601684683309e-06,
+      "loss": 0.7738,
+      "num_input_tokens_seen": 100672130,
+      "step": 4724
+    },
+    {
+      "epoch": 0.5681476582697048,
+      "grad_norm": 3.230107739551636,
+      "learning_rate": 1.6573926853954674e-06,
+      "loss": 0.6821,
+      "num_input_tokens_seen": 100689435,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5682679011603439,
+      "grad_norm": 4.2877522687975524,
+      "learning_rate": 1.6566252542958608e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 100708655,
+      "step": 4726
+    },
+    {
+      "epoch": 0.568388144050983,
+      "grad_norm": 3.1070111416941155,
+      "learning_rate": 1.6558578752859305e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 100727335,
+      "step": 4727
+    },
+    {
+      "epoch": 0.5685083869416221,
+      "grad_norm": 2.738361123409445,
+      "learning_rate": 1.6550905484820867e-06,
+      "loss": 0.7788,
+      "num_input_tokens_seen": 100745515,
+      "step": 4728
+    },
+    {
+      "epoch": 0.5686286298322611,
+      "grad_norm": 3.982191595958614,
+      "learning_rate": 1.6543232740007328e-06,
+      "loss": 0.7939,
+      "num_input_tokens_seen": 100762350,
+      "step": 4729
+    },
+    {
+      "epoch": 0.5687488727229003,
+      "grad_norm": 7.7677271441508156,
+      "learning_rate": 1.6535560519582626e-06,
+      "loss": 0.6655,
+      "num_input_tokens_seen": 100781750,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5688691156135394,
+      "grad_norm": 5.060103201206567,
+      "learning_rate": 1.6527888824710642e-06,
+      "loss": 0.7311,
+      "num_input_tokens_seen": 100801070,
+      "step": 4731
+    },
+    {
+      "epoch": 0.5689893585041784,
+      "grad_norm": 3.339791122579335,
+      "learning_rate": 1.6520217656555166e-06,
+      "loss": 0.7617,
+      "num_input_tokens_seen": 100820080,
+      "step": 4732
+    },
+    {
+      "epoch": 0.5691096013948175,
+      "grad_norm": 1.8199434351839006,
+      "learning_rate": 1.6512547016279905e-06,
+      "loss": 0.7063,
+      "num_input_tokens_seen": 100840155,
+      "step": 4733
+    },
+    {
+      "epoch": 0.5692298442854566,
+      "grad_norm": 2.369494032971793,
+      "learning_rate": 1.6504876905048485e-06,
+      "loss": 0.6963,
+      "num_input_tokens_seen": 100856835,
+      "step": 4734
+    },
+    {
+      "epoch": 0.5693500871760957,
+      "grad_norm": 3.842521411276859,
+      "learning_rate": 1.6497207324024464e-06,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 100875455,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5694703300667348,
+      "grad_norm": 1.9911215986332345,
+      "learning_rate": 1.6489538274371305e-06,
+      "loss": 0.8214,
+      "num_input_tokens_seen": 100893780,
+      "step": 4736
+    },
+    {
+      "epoch": 0.5695905729573739,
+      "grad_norm": 2.548021081688492,
+      "learning_rate": 1.64818697572524e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 100911835,
+      "step": 4737
+    },
+    {
+      "epoch": 0.569710815848013,
+      "grad_norm": 1.49836542829536,
+      "learning_rate": 1.6474201773831047e-06,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 100934425,
+      "step": 4738
+    },
+    {
+      "epoch": 0.569831058738652,
+      "grad_norm": 2.133852703510713,
+      "learning_rate": 1.646653432527049e-06,
+      "loss": 0.7307,
+      "num_input_tokens_seen": 100954785,
+      "step": 4739
+    },
+    {
+      "epoch": 0.5699513016292912,
+      "grad_norm": 1.5879879796960523,
+      "learning_rate": 1.6458867412733865e-06,
+      "loss": 0.7418,
+      "num_input_tokens_seen": 100976320,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5700715445199303,
+      "grad_norm": 8.443052490242112,
+      "learning_rate": 1.645120103738424e-06,
+      "loss": 0.7413,
+      "num_input_tokens_seen": 100993550,
+      "step": 4741
+    },
+    {
+      "epoch": 0.5701917874105693,
+      "grad_norm": 2.2819956511790593,
+      "learning_rate": 1.6443535200384591e-06,
+      "loss": 0.8333,
+      "num_input_tokens_seen": 101011445,
+      "step": 4742
+    },
+    {
+      "epoch": 0.5703120303012085,
+      "grad_norm": 1.809403089738324,
+      "learning_rate": 1.6435869902897827e-06,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 101029745,
+      "step": 4743
+    },
+    {
+      "epoch": 0.5704322731918475,
+      "grad_norm": 0.8329593107263714,
+      "learning_rate": 1.6428205146086769e-06,
+      "loss": 0.6472,
+      "num_input_tokens_seen": 101091445,
+      "step": 4744
+    },
+    {
+      "epoch": 0.5705525160824866,
+      "grad_norm": 1.6313255063108463,
+      "learning_rate": 1.6420540931114146e-06,
+      "loss": 0.6978,
+      "num_input_tokens_seen": 101111755,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5706727589731257,
+      "grad_norm": 3.0912751483316834,
+      "learning_rate": 1.6412877259142612e-06,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 101131395,
+      "step": 4746
+    },
+    {
+      "epoch": 0.5707930018637648,
+      "grad_norm": 2.6910238741825587,
+      "learning_rate": 1.6405214131334757e-06,
+      "loss": 0.7438,
+      "num_input_tokens_seen": 101149640,
+      "step": 4747
+    },
+    {
+      "epoch": 0.5709132447544039,
+      "grad_norm": 2.422751478236722,
+      "learning_rate": 1.6397551548853056e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 101167525,
+      "step": 4748
+    },
+    {
+      "epoch": 0.571033487645043,
+      "grad_norm": 2.30627703476244,
+      "learning_rate": 1.6389889512859921e-06,
+      "loss": 0.7003,
+      "num_input_tokens_seen": 101186905,
+      "step": 4749
+    },
+    {
+      "epoch": 0.5711537305356821,
+      "grad_norm": 0.896744373482785,
+      "learning_rate": 1.638222802451767e-06,
+      "loss": 0.652,
+      "num_input_tokens_seen": 101248105,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5712739734263211,
+      "grad_norm": 2.2774020046789674,
+      "learning_rate": 1.6374567084988557e-06,
+      "loss": 0.7479,
+      "num_input_tokens_seen": 101269010,
+      "step": 4751
+    },
+    {
+      "epoch": 0.5713942163169603,
+      "grad_norm": 2.6025776408135477,
+      "learning_rate": 1.6366906695434738e-06,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 101291250,
+      "step": 4752
+    },
+    {
+      "epoch": 0.5715144592075994,
+      "grad_norm": 3.427391158594594,
+      "learning_rate": 1.6359246857018282e-06,
+      "loss": 0.8534,
+      "num_input_tokens_seen": 101308500,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5716347020982384,
+      "grad_norm": 1.9613229460632502,
+      "learning_rate": 1.6351587570901178e-06,
+      "loss": 0.7721,
+      "num_input_tokens_seen": 101328345,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5717549449888776,
+      "grad_norm": 3.22808589114031,
+      "learning_rate": 1.6343928838245344e-06,
+      "loss": 0.7442,
+      "num_input_tokens_seen": 101340065,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5718751878795166,
+      "grad_norm": 2.03266459375331,
+      "learning_rate": 1.63362706602126e-06,
+      "loss": 0.6838,
+      "num_input_tokens_seen": 101361380,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5719954307701557,
+      "grad_norm": 3.5285262606304415,
+      "learning_rate": 1.632861303796468e-06,
+      "loss": 0.6587,
+      "num_input_tokens_seen": 101384165,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5721156736607949,
+      "grad_norm": 2.227232349426142,
+      "learning_rate": 1.6320955972663237e-06,
+      "loss": 0.6779,
+      "num_input_tokens_seen": 101403480,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5722359165514339,
+      "grad_norm": 2.3870076280270935,
+      "learning_rate": 1.6313299465469857e-06,
+      "loss": 0.6491,
+      "num_input_tokens_seen": 101425930,
+      "step": 4759
+    },
+    {
+      "epoch": 0.572356159442073,
+      "grad_norm": 3.939530810206527,
+      "learning_rate": 1.6305643517546014e-06,
+      "loss": 0.7867,
+      "num_input_tokens_seen": 101441030,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5724764023327121,
+      "grad_norm": 2.1623517362724543,
+      "learning_rate": 1.6297988130053116e-06,
+      "loss": 0.8418,
+      "num_input_tokens_seen": 101460470,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5725966452233512,
+      "grad_norm": 2.080168075052579,
+      "learning_rate": 1.6290333304152469e-06,
+      "loss": 0.7037,
+      "num_input_tokens_seen": 101480065,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5727168881139902,
+      "grad_norm": 3.125384086160007,
+      "learning_rate": 1.6282679041005314e-06,
+      "loss": 0.5689,
+      "num_input_tokens_seen": 101505375,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5728371310046293,
+      "grad_norm": 2.244403543248274,
+      "learning_rate": 1.6275025341772793e-06,
+      "loss": 0.8676,
+      "num_input_tokens_seen": 101521400,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5729573738952685,
+      "grad_norm": 5.964647442346836,
+      "learning_rate": 1.6267372207615965e-06,
+      "loss": 0.8224,
+      "num_input_tokens_seen": 101538585,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5730776167859075,
+      "grad_norm": 3.0481800204566665,
+      "learning_rate": 1.62597196396958e-06,
+      "loss": 0.7857,
+      "num_input_tokens_seen": 101556475,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5731978596765466,
+      "grad_norm": 2.743014184261631,
+      "learning_rate": 1.6252067639173197e-06,
+      "loss": 0.8503,
+      "num_input_tokens_seen": 101578105,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5733181025671857,
+      "grad_norm": 2.670306404747414,
+      "learning_rate": 1.6244416207208956e-06,
+      "loss": 0.6942,
+      "num_input_tokens_seen": 101598760,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5734383454578248,
+      "grad_norm": 1.8075204728786771,
+      "learning_rate": 1.6236765344963787e-06,
+      "loss": 0.7358,
+      "num_input_tokens_seen": 101619740,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5735585883484638,
+      "grad_norm": 3.0653207775171425,
+      "learning_rate": 1.6229115053598322e-06,
+      "loss": 0.6939,
+      "num_input_tokens_seen": 101641215,
+      "step": 4770
+    },
+    {
+      "epoch": 0.573678831239103,
+      "grad_norm": 2.9889363357792944,
+      "learning_rate": 1.6221465334273108e-06,
+      "loss": 0.7106,
+      "num_input_tokens_seen": 101660145,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5737990741297421,
+      "grad_norm": 2.5190817645108803,
+      "learning_rate": 1.6213816188148597e-06,
+      "loss": 0.6036,
+      "num_input_tokens_seen": 101678570,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5739193170203811,
+      "grad_norm": 1.9760531538922232,
+      "learning_rate": 1.6206167616385162e-06,
+      "loss": 0.7627,
+      "num_input_tokens_seen": 101699355,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5740395599110203,
+      "grad_norm": 3.404496126595559,
+      "learning_rate": 1.6198519620143074e-06,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 101716230,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5741598028016593,
+      "grad_norm": 2.149950119420719,
+      "learning_rate": 1.6190872200582546e-06,
+      "loss": 0.7725,
+      "num_input_tokens_seen": 101737690,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5742800456922984,
+      "grad_norm": 3.078191430272785,
+      "learning_rate": 1.6183225358863676e-06,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 101754305,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5744002885829376,
+      "grad_norm": 2.764595124398925,
+      "learning_rate": 1.6175579096146485e-06,
+      "loss": 0.7048,
+      "num_input_tokens_seen": 101773460,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5745205314735766,
+      "grad_norm": 1.9478939482039526,
+      "learning_rate": 1.6167933413590899e-06,
+      "loss": 0.8504,
+      "num_input_tokens_seen": 101792085,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5746407743642157,
+      "grad_norm": 3.915496492629683,
+      "learning_rate": 1.6160288312356773e-06,
+      "loss": 0.9071,
+      "num_input_tokens_seen": 101808935,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5747610172548548,
+      "grad_norm": 3.2983233051837852,
+      "learning_rate": 1.6152643793603857e-06,
+      "loss": 0.8188,
+      "num_input_tokens_seen": 101829005,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5748812601454939,
+      "grad_norm": 2.192637542769304,
+      "learning_rate": 1.6144999858491819e-06,
+      "loss": 0.8691,
+      "num_input_tokens_seen": 101847355,
+      "step": 4781
+    },
+    {
+      "epoch": 0.575001503036133,
+      "grad_norm": 1.889045816775095,
+      "learning_rate": 1.6137356508180227e-06,
+      "loss": 0.8513,
+      "num_input_tokens_seen": 101868785,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5751217459267721,
+      "grad_norm": 2.423801981230503,
+      "learning_rate": 1.6129713743828593e-06,
+      "loss": 0.8106,
+      "num_input_tokens_seen": 101887515,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5752419888174112,
+      "grad_norm": 1.5807037643188067,
+      "learning_rate": 1.6122071566596302e-06,
+      "loss": 0.7551,
+      "num_input_tokens_seen": 101907510,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5753622317080502,
+      "grad_norm": 2.4985192169439205,
+      "learning_rate": 1.6114429977642678e-06,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 101921735,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5754824745986894,
+      "grad_norm": 1.9433843077420119,
+      "learning_rate": 1.6106788978126926e-06,
+      "loss": 0.7313,
+      "num_input_tokens_seen": 101940430,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5756027174893285,
+      "grad_norm": 4.223511918574018,
+      "learning_rate": 1.6099148569208196e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 101957370,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5757229603799675,
+      "grad_norm": 1.9386536826596157,
+      "learning_rate": 1.609150875204553e-06,
+      "loss": 0.6265,
+      "num_input_tokens_seen": 101977970,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5758432032706067,
+      "grad_norm": 2.0811855703967903,
+      "learning_rate": 1.6083869527797875e-06,
+      "loss": 0.8564,
+      "num_input_tokens_seen": 101997060,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5759634461612457,
+      "grad_norm": 2.4288178422566076,
+      "learning_rate": 1.6076230897624098e-06,
+      "loss": 0.7425,
+      "num_input_tokens_seen": 102018985,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5760836890518848,
+      "grad_norm": 3.5922269408610146,
+      "learning_rate": 1.6068592862682974e-06,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 102036860,
+      "step": 4791
+    },
+    {
+      "epoch": 0.576203931942524,
+      "grad_norm": 2.5200497048020476,
+      "learning_rate": 1.6060955424133191e-06,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 102057505,
+      "step": 4792
+    },
+    {
+      "epoch": 0.576324174833163,
+      "grad_norm": 2.1441582970490916,
+      "learning_rate": 1.6053318583133336e-06,
+      "loss": 0.8984,
+      "num_input_tokens_seen": 102078095,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5764444177238021,
+      "grad_norm": 3.0081400330823747,
+      "learning_rate": 1.6045682340841907e-06,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 102096740,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5765646606144411,
+      "grad_norm": 0.8042994637477046,
+      "learning_rate": 1.6038046698417332e-06,
+      "loss": 0.6163,
+      "num_input_tokens_seen": 102157355,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5766849035050803,
+      "grad_norm": 2.4414848786422176,
+      "learning_rate": 1.6030411657017919e-06,
+      "loss": 0.6901,
+      "num_input_tokens_seen": 102176730,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5768051463957193,
+      "grad_norm": 2.47665713085787,
+      "learning_rate": 1.6022777217801907e-06,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 102193405,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5769253892863584,
+      "grad_norm": 2.0361900406680213,
+      "learning_rate": 1.601514338192742e-06,
+      "loss": 0.7259,
+      "num_input_tokens_seen": 102213055,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5770456321769976,
+      "grad_norm": 3.437548489150342,
+      "learning_rate": 1.6007510150552518e-06,
+      "loss": 0.7085,
+      "num_input_tokens_seen": 102230835,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5771658750676366,
+      "grad_norm": 2.16951101487682,
+      "learning_rate": 1.5999877524835154e-06,
+      "loss": 0.6228,
+      "num_input_tokens_seen": 102255000,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5772861179582757,
+      "grad_norm": 3.2188623855678995,
+      "learning_rate": 1.5992245505933188e-06,
+      "loss": 0.6706,
+      "num_input_tokens_seen": 102274420,
+      "step": 4801
+    },
+    {
+      "epoch": 0.5774063608489148,
+      "grad_norm": 2.77323124188225,
+      "learning_rate": 1.5984614095004382e-06,
+      "loss": 0.7033,
+      "num_input_tokens_seen": 102295275,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5775266037395539,
+      "grad_norm": 2.4737891455268706,
+      "learning_rate": 1.5976983293206438e-06,
+      "loss": 0.8092,
+      "num_input_tokens_seen": 102310800,
+      "step": 4803
+    },
+    {
+      "epoch": 0.577646846630193,
+      "grad_norm": 1.7917683031291272,
+      "learning_rate": 1.5969353101696928e-06,
+      "loss": 0.7097,
+      "num_input_tokens_seen": 102328960,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5777670895208321,
+      "grad_norm": 1.7597343711219515,
+      "learning_rate": 1.5961723521633346e-06,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 102349920,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5778873324114712,
+      "grad_norm": 4.2148472966845825,
+      "learning_rate": 1.595409455417309e-06,
+      "loss": 0.9018,
+      "num_input_tokens_seen": 102367630,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5780075753021102,
+      "grad_norm": 2.709987246892763,
+      "learning_rate": 1.5946466200473482e-06,
+      "loss": 0.7905,
+      "num_input_tokens_seen": 102385260,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5781278181927494,
+      "grad_norm": 2.1272388902853074,
+      "learning_rate": 1.5938838461691723e-06,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 102401890,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5782480610833884,
+      "grad_norm": 3.6465495443497047,
+      "learning_rate": 1.5931211338984944e-06,
+      "loss": 0.8312,
+      "num_input_tokens_seen": 102418815,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5783683039740275,
+      "grad_norm": 2.311343798011985,
+      "learning_rate": 1.592358483351016e-06,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 102438710,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5784885468646667,
+      "grad_norm": 2.1380002062817223,
+      "learning_rate": 1.5915958946424326e-06,
+      "loss": 0.7167,
+      "num_input_tokens_seen": 102457115,
+      "step": 4811
+    },
+    {
+      "epoch": 0.5786087897553057,
+      "grad_norm": 2.5982033638435826,
+      "learning_rate": 1.5908333678884271e-06,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 102483255,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5787290326459448,
+      "grad_norm": 2.3516452356233244,
+      "learning_rate": 1.5900709032046743e-06,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 102501050,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5788492755365839,
+      "grad_norm": 2.2369910595999243,
+      "learning_rate": 1.5893085007068391e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 102518330,
+      "step": 4814
+    },
+    {
+      "epoch": 0.578969518427223,
+      "grad_norm": 1.8778943552202543,
+      "learning_rate": 1.5885461605105786e-06,
+      "loss": 0.7048,
+      "num_input_tokens_seen": 102539650,
+      "step": 4815
+    },
+    {
+      "epoch": 0.579089761317862,
+      "grad_norm": 2.1204666426980716,
+      "learning_rate": 1.587783882731538e-06,
+      "loss": 0.767,
+      "num_input_tokens_seen": 102557915,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5792100042085012,
+      "grad_norm": 11.031681896377048,
+      "learning_rate": 1.587021667485355e-06,
+      "loss": 0.6919,
+      "num_input_tokens_seen": 102577005,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5793302470991403,
+      "grad_norm": 2.2090396212456915,
+      "learning_rate": 1.5862595148876554e-06,
+      "loss": 0.7774,
+      "num_input_tokens_seen": 102596830,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5794504899897793,
+      "grad_norm": 3.1309532003069016,
+      "learning_rate": 1.5854974250540595e-06,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 102611295,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5795707328804185,
+      "grad_norm": 2.8950045941247535,
+      "learning_rate": 1.5847353981001747e-06,
+      "loss": 0.7547,
+      "num_input_tokens_seen": 102628195,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5796909757710575,
+      "grad_norm": 1.6182434235722363,
+      "learning_rate": 1.5839734341415997e-06,
+      "loss": 0.6862,
+      "num_input_tokens_seen": 102650115,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5798112186616966,
+      "grad_norm": 1.9753134092030766,
+      "learning_rate": 1.5832115332939238e-06,
+      "loss": 0.7618,
+      "num_input_tokens_seen": 102668275,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5799314615523358,
+      "grad_norm": 3.5992823794245634,
+      "learning_rate": 1.5824496956727272e-06,
+      "loss": 0.7521,
+      "num_input_tokens_seen": 102685200,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5800517044429748,
+      "grad_norm": 2.419276183619453,
+      "learning_rate": 1.5816879213935797e-06,
+      "loss": 0.721,
+      "num_input_tokens_seen": 102703730,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5801719473336139,
+      "grad_norm": 2.1693232174276345,
+      "learning_rate": 1.580926210572042e-06,
+      "loss": 0.7846,
+      "num_input_tokens_seen": 102724490,
+      "step": 4825
+    },
+    {
+      "epoch": 0.580292190224253,
+      "grad_norm": 2.0034353019539655,
+      "learning_rate": 1.580164563323664e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 102745195,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5804124331148921,
+      "grad_norm": 2.3195682505585333,
+      "learning_rate": 1.579402979763989e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 102765250,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5805326760055312,
+      "grad_norm": 2.430777259244859,
+      "learning_rate": 1.578641460008548e-06,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 102782705,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5806529188961702,
+      "grad_norm": 2.2615825411291492,
+      "learning_rate": 1.5778800041728617e-06,
+      "loss": 0.6688,
+      "num_input_tokens_seen": 102798715,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5807731617868094,
+      "grad_norm": 1.7188285956099998,
+      "learning_rate": 1.5771186123724426e-06,
+      "loss": 0.6585,
+      "num_input_tokens_seen": 102820275,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5808934046774484,
+      "grad_norm": 1.9729082198831696,
+      "learning_rate": 1.5763572847227943e-06,
+      "loss": 0.706,
+      "num_input_tokens_seen": 102840880,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5810136475680875,
+      "grad_norm": 2.3762129676769703,
+      "learning_rate": 1.5755960213394096e-06,
+      "loss": 0.804,
+      "num_input_tokens_seen": 102857700,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5811338904587267,
+      "grad_norm": 2.3037636435955355,
+      "learning_rate": 1.5748348223377707e-06,
+      "loss": 0.7789,
+      "num_input_tokens_seen": 102874975,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5812541333493657,
+      "grad_norm": 1.829828029407949,
+      "learning_rate": 1.5740736878333507e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 102892535,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5813743762400048,
+      "grad_norm": 2.6787328061838296,
+      "learning_rate": 1.5733126179416143e-06,
+      "loss": 0.7758,
+      "num_input_tokens_seen": 102906740,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5814946191306439,
+      "grad_norm": 2.4523824946009314,
+      "learning_rate": 1.5725516127780144e-06,
+      "loss": 0.7191,
+      "num_input_tokens_seen": 102928595,
+      "step": 4836
+    },
+    {
+      "epoch": 0.581614862021283,
+      "grad_norm": 3.0008383040067512,
+      "learning_rate": 1.5717906724579947e-06,
+      "loss": 0.8835,
+      "num_input_tokens_seen": 102945375,
+      "step": 4837
+    },
+    {
+      "epoch": 0.581735104911922,
+      "grad_norm": 2.4488302291988027,
+      "learning_rate": 1.571029797096989e-06,
+      "loss": 0.6769,
+      "num_input_tokens_seen": 102966200,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5818553478025612,
+      "grad_norm": 1.8453097831632193,
+      "learning_rate": 1.5702689868104227e-06,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 102985815,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5819755906932003,
+      "grad_norm": 2.2388402217495105,
+      "learning_rate": 1.5695082417137096e-06,
+      "loss": 0.7473,
+      "num_input_tokens_seen": 103003410,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5820958335838393,
+      "grad_norm": 1.8119756428290785,
+      "learning_rate": 1.5687475619222539e-06,
+      "loss": 0.7479,
+      "num_input_tokens_seen": 103023085,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5822160764744785,
+      "grad_norm": 3.056226364776706,
+      "learning_rate": 1.5679869475514496e-06,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 103039740,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5823363193651175,
+      "grad_norm": 2.899919038780509,
+      "learning_rate": 1.5672263987166825e-06,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 103059375,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5824565622557566,
+      "grad_norm": 2.370603530474994,
+      "learning_rate": 1.5664659155333263e-06,
+      "loss": 0.6185,
+      "num_input_tokens_seen": 103081125,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5825768051463958,
+      "grad_norm": 2.5721388576486297,
+      "learning_rate": 1.5657054981167463e-06,
+      "loss": 0.8808,
+      "num_input_tokens_seen": 103099740,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5826970480370348,
+      "grad_norm": 2.03662999121406,
+      "learning_rate": 1.564945146582296e-06,
+      "loss": 0.6727,
+      "num_input_tokens_seen": 103120850,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5828172909276739,
+      "grad_norm": 2.1717091557638364,
+      "learning_rate": 1.5641848610453218e-06,
+      "loss": 0.8256,
+      "num_input_tokens_seen": 103139230,
+      "step": 4847
+    },
+    {
+      "epoch": 0.582937533818313,
+      "grad_norm": 2.983228714278711,
+      "learning_rate": 1.563424641621158e-06,
+      "loss": 0.855,
+      "num_input_tokens_seen": 103158130,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5830577767089521,
+      "grad_norm": 2.296346511754696,
+      "learning_rate": 1.5626644884251286e-06,
+      "loss": 0.6972,
+      "num_input_tokens_seen": 103177370,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5831780195995911,
+      "grad_norm": 1.816630278458807,
+      "learning_rate": 1.5619044015725481e-06,
+      "loss": 0.8778,
+      "num_input_tokens_seen": 103196780,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5832982624902303,
+      "grad_norm": 2.5257346420752373,
+      "learning_rate": 1.5611443811787224e-06,
+      "loss": 0.8681,
+      "num_input_tokens_seen": 103210625,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5834185053808694,
+      "grad_norm": 2.423216048826171,
+      "learning_rate": 1.560384427358945e-06,
+      "loss": 0.6901,
+      "num_input_tokens_seen": 103229890,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5835387482715084,
+      "grad_norm": 1.8768853969488881,
+      "learning_rate": 1.5596245402285002e-06,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 103253135,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5836589911621476,
+      "grad_norm": 2.1376558427561334,
+      "learning_rate": 1.5588647199026619e-06,
+      "loss": 0.814,
+      "num_input_tokens_seen": 103270590,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5837792340527866,
+      "grad_norm": 2.2887586113960556,
+      "learning_rate": 1.5581049664966956e-06,
+      "loss": 0.8773,
+      "num_input_tokens_seen": 103288070,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5838994769434257,
+      "grad_norm": 3.6478355470861517,
+      "learning_rate": 1.5573452801258545e-06,
+      "loss": 0.6902,
+      "num_input_tokens_seen": 103334960,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5840197198340649,
+      "grad_norm": 4.4360439224691985,
+      "learning_rate": 1.5565856609053824e-06,
+      "loss": 0.632,
+      "num_input_tokens_seen": 103353475,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5841399627247039,
+      "grad_norm": 1.9122823239521842,
+      "learning_rate": 1.5558261089505127e-06,
+      "loss": 0.7905,
+      "num_input_tokens_seen": 103371925,
+      "step": 4858
+    },
+    {
+      "epoch": 0.584260205615343,
+      "grad_norm": 2.240434628717895,
+      "learning_rate": 1.5550666243764697e-06,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 103389805,
+      "step": 4859
+    },
+    {
+      "epoch": 0.584380448505982,
+      "grad_norm": 2.5336226641015713,
+      "learning_rate": 1.5543072072984655e-06,
+      "loss": 0.7636,
+      "num_input_tokens_seen": 103407785,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5845006913966212,
+      "grad_norm": 1.9300210680071765,
+      "learning_rate": 1.553547857831704e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 103424015,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5846209342872603,
+      "grad_norm": 1.0540954534540454,
+      "learning_rate": 1.5527885760913767e-06,
+      "loss": 0.7196,
+      "num_input_tokens_seen": 103473625,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5847411771778993,
+      "grad_norm": 2.22492688902478,
+      "learning_rate": 1.5520293621926675e-06,
+      "loss": 0.7555,
+      "num_input_tokens_seen": 103492605,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5848614200685385,
+      "grad_norm": 2.336872192051636,
+      "learning_rate": 1.5512702162507478e-06,
+      "loss": 0.7201,
+      "num_input_tokens_seen": 103512640,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5849816629591775,
+      "grad_norm": 1.1429192558315644,
+      "learning_rate": 1.5505111383807796e-06,
+      "loss": 0.5699,
+      "num_input_tokens_seen": 103575030,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5851019058498166,
+      "grad_norm": 1.7594670781130073,
+      "learning_rate": 1.5497521286979138e-06,
+      "loss": 0.7984,
+      "num_input_tokens_seen": 103594990,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5852221487404557,
+      "grad_norm": 2.8975044302600956,
+      "learning_rate": 1.5489931873172927e-06,
+      "loss": 0.7423,
+      "num_input_tokens_seen": 103616030,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5853423916310948,
+      "grad_norm": 3.5836437823057663,
+      "learning_rate": 1.5482343143540467e-06,
+      "loss": 0.7891,
+      "num_input_tokens_seen": 103637015,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5854626345217339,
+      "grad_norm": 2.0902790807598213,
+      "learning_rate": 1.5474755099232956e-06,
+      "loss": 0.8298,
+      "num_input_tokens_seen": 103653775,
+      "step": 4869
+    },
+    {
+      "epoch": 0.585582877412373,
+      "grad_norm": 0.7667064686663994,
+      "learning_rate": 1.546716774140149e-06,
+      "loss": 0.5902,
+      "num_input_tokens_seen": 103714975,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5857031203030121,
+      "grad_norm": 2.947524138475556,
+      "learning_rate": 1.5459581071197083e-06,
+      "loss": 0.7067,
+      "num_input_tokens_seen": 103730355,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5858233631936511,
+      "grad_norm": 5.5961687268706,
+      "learning_rate": 1.5451995089770624e-06,
+      "loss": 0.8211,
+      "num_input_tokens_seen": 103749860,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5859436060842903,
+      "grad_norm": 1.672344560749858,
+      "learning_rate": 1.544440979827289e-06,
+      "loss": 0.7125,
+      "num_input_tokens_seen": 103773670,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5860638489749294,
+      "grad_norm": 2.1147885425560426,
+      "learning_rate": 1.5436825197854555e-06,
+      "loss": 0.8,
+      "num_input_tokens_seen": 103791870,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5861840918655684,
+      "grad_norm": 3.3177634559123317,
+      "learning_rate": 1.5429241289666219e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 103809090,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5863043347562076,
+      "grad_norm": 3.3844530342066674,
+      "learning_rate": 1.5421658074858346e-06,
+      "loss": 0.694,
+      "num_input_tokens_seen": 103826915,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5864245776468466,
+      "grad_norm": 2.739114405757076,
+      "learning_rate": 1.5414075554581302e-06,
+      "loss": 0.6577,
+      "num_input_tokens_seen": 103844680,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5865448205374857,
+      "grad_norm": 2.3510347997538448,
+      "learning_rate": 1.5406493729985348e-06,
+      "loss": 0.7743,
+      "num_input_tokens_seen": 103863595,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5866650634281249,
+      "grad_norm": 5.748427421319987,
+      "learning_rate": 1.5398912602220644e-06,
+      "loss": 0.7175,
+      "num_input_tokens_seen": 103882590,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5867853063187639,
+      "grad_norm": 2.9413356827540103,
+      "learning_rate": 1.5391332172437243e-06,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 103899330,
+      "step": 4880
+    },
+    {
+      "epoch": 0.586905549209403,
+      "grad_norm": 5.00974686559913,
+      "learning_rate": 1.5383752441785085e-06,
+      "loss": 0.7478,
+      "num_input_tokens_seen": 103918275,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5870257921000421,
+      "grad_norm": 3.0010134130858557,
+      "learning_rate": 1.5376173411414003e-06,
+      "loss": 0.854,
+      "num_input_tokens_seen": 103936035,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5871460349906812,
+      "grad_norm": 2.00597889705922,
+      "learning_rate": 1.5368595082473748e-06,
+      "loss": 0.7793,
+      "num_input_tokens_seen": 103954055,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5872662778813202,
+      "grad_norm": 1.8146919527807026,
+      "learning_rate": 1.5361017456113935e-06,
+      "loss": 0.774,
+      "num_input_tokens_seen": 103974125,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5873865207719594,
+      "grad_norm": 2.538178161739637,
+      "learning_rate": 1.5353440533484085e-06,
+      "loss": 0.8532,
+      "num_input_tokens_seen": 103992700,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5875067636625985,
+      "grad_norm": 2.146481903250869,
+      "learning_rate": 1.534586431573361e-06,
+      "loss": 0.653,
+      "num_input_tokens_seen": 104017360,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5876270065532375,
+      "grad_norm": 2.520778897230332,
+      "learning_rate": 1.533828880401182e-06,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 104036580,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5877472494438767,
+      "grad_norm": 2.6093016659075845,
+      "learning_rate": 1.5330713999467915e-06,
+      "loss": 0.7124,
+      "num_input_tokens_seen": 104055045,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5878674923345157,
+      "grad_norm": 2.083236178913225,
+      "learning_rate": 1.532313990325098e-06,
+      "loss": 0.5748,
+      "num_input_tokens_seen": 104075370,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5879877352251548,
+      "grad_norm": 1.928090608522909,
+      "learning_rate": 1.5315566516509997e-06,
+      "loss": 0.7606,
+      "num_input_tokens_seen": 104093260,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5881079781157939,
+      "grad_norm": 2.1035431120775727,
+      "learning_rate": 1.5307993840393857e-06,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 104111060,
+      "step": 4891
+    },
+    {
+      "epoch": 0.588228221006433,
+      "grad_norm": 2.2642621463515162,
+      "learning_rate": 1.530042187605132e-06,
+      "loss": 0.8052,
+      "num_input_tokens_seen": 104130035,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5883484638970721,
+      "grad_norm": 1.7798274793622242,
+      "learning_rate": 1.5292850624631046e-06,
+      "loss": 0.8342,
+      "num_input_tokens_seen": 104151950,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5884687067877111,
+      "grad_norm": 2.5993270618500466,
+      "learning_rate": 1.5285280087281589e-06,
+      "loss": 0.7857,
+      "num_input_tokens_seen": 104172400,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5885889496783503,
+      "grad_norm": 0.7173286231951608,
+      "learning_rate": 1.5277710265151398e-06,
+      "loss": 0.5678,
+      "num_input_tokens_seen": 104241600,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5887091925689893,
+      "grad_norm": 2.5862250478235302,
+      "learning_rate": 1.5270141159388803e-06,
+      "loss": 0.7687,
+      "num_input_tokens_seen": 104258340,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5888294354596284,
+      "grad_norm": 2.2463954899326204,
+      "learning_rate": 1.5262572771142036e-06,
+      "loss": 0.792,
+      "num_input_tokens_seen": 104279135,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5889496783502676,
+      "grad_norm": 1.9731000110447356,
+      "learning_rate": 1.5255005101559201e-06,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 104296465,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5890699212409066,
+      "grad_norm": 2.0015844887810768,
+      "learning_rate": 1.524743815178833e-06,
+      "loss": 0.7649,
+      "num_input_tokens_seen": 104314145,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5891901641315457,
+      "grad_norm": 1.9543064685316924,
+      "learning_rate": 1.5239871922977315e-06,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 104333780,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5893104070221848,
+      "grad_norm": 2.4083839217209677,
+      "learning_rate": 1.523230641627394e-06,
+      "loss": 0.8912,
+      "num_input_tokens_seen": 104352485,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5894306499128239,
+      "grad_norm": 3.824197071551223,
+      "learning_rate": 1.5224741632825888e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 104372395,
+      "step": 4902
+    },
+    {
+      "epoch": 0.589550892803463,
+      "grad_norm": 1.654199761882125,
+      "learning_rate": 1.521717757378074e-06,
+      "loss": 0.6843,
+      "num_input_tokens_seen": 104392660,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5896711356941021,
+      "grad_norm": 2.722285510590189,
+      "learning_rate": 1.520961424028595e-06,
+      "loss": 0.6847,
+      "num_input_tokens_seen": 104410035,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5897913785847412,
+      "grad_norm": 2.2078585100402215,
+      "learning_rate": 1.520205163348887e-06,
+      "loss": 0.8481,
+      "num_input_tokens_seen": 104427690,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5899116214753802,
+      "grad_norm": 0.783418064196027,
+      "learning_rate": 1.5194489754536735e-06,
+      "loss": 0.5761,
+      "num_input_tokens_seen": 104482510,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5900318643660194,
+      "grad_norm": 2.8402461955712437,
+      "learning_rate": 1.5186928604576692e-06,
+      "loss": 0.755,
+      "num_input_tokens_seen": 104499425,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5901521072566585,
+      "grad_norm": 2.5059622064635487,
+      "learning_rate": 1.5179368184755752e-06,
+      "loss": 0.7715,
+      "num_input_tokens_seen": 104517230,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5902723501472975,
+      "grad_norm": 1.7122431194773249,
+      "learning_rate": 1.5171808496220825e-06,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 104535705,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5903925930379367,
+      "grad_norm": 1.72795648900781,
+      "learning_rate": 1.5164249540118708e-06,
+      "loss": 0.8092,
+      "num_input_tokens_seen": 104554550,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5905128359285757,
+      "grad_norm": 1.871686079541397,
+      "learning_rate": 1.5156691317596093e-06,
+      "loss": 0.8253,
+      "num_input_tokens_seen": 104575695,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5906330788192148,
+      "grad_norm": 2.5585508312082736,
+      "learning_rate": 1.5149133829799556e-06,
+      "loss": 0.6701,
+      "num_input_tokens_seen": 104593410,
+      "step": 4912
+    },
+    {
+      "epoch": 0.590753321709854,
+      "grad_norm": 5.096875372963243,
+      "learning_rate": 1.514157707787556e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 104610455,
+      "step": 4913
+    },
+    {
+      "epoch": 0.590873564600493,
+      "grad_norm": 2.296656236793456,
+      "learning_rate": 1.5134021062970447e-06,
+      "loss": 0.7194,
+      "num_input_tokens_seen": 104628555,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5909938074911321,
+      "grad_norm": 1.9009599285599679,
+      "learning_rate": 1.5126465786230488e-06,
+      "loss": 0.8038,
+      "num_input_tokens_seen": 104645050,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5911140503817712,
+      "grad_norm": 3.185264952966568,
+      "learning_rate": 1.5118911248801787e-06,
+      "loss": 0.8085,
+      "num_input_tokens_seen": 104662780,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5912342932724103,
+      "grad_norm": 4.3621336470149465,
+      "learning_rate": 1.5111357451830368e-06,
+      "loss": 0.7928,
+      "num_input_tokens_seen": 104681195,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5913545361630493,
+      "grad_norm": 2.8599092328203297,
+      "learning_rate": 1.5103804396462127e-06,
+      "loss": 0.7122,
+      "num_input_tokens_seen": 104700850,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5914747790536885,
+      "grad_norm": 1.9355973174636198,
+      "learning_rate": 1.5096252083842877e-06,
+      "loss": 0.7974,
+      "num_input_tokens_seen": 104719780,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5915950219443276,
+      "grad_norm": 3.9243221278078355,
+      "learning_rate": 1.508870051511829e-06,
+      "loss": 0.8318,
+      "num_input_tokens_seen": 104738820,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5917152648349666,
+      "grad_norm": 2.0825569427567308,
+      "learning_rate": 1.5081149691433923e-06,
+      "loss": 0.6637,
+      "num_input_tokens_seen": 104758525,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5918355077256057,
+      "grad_norm": 1.564880547265013,
+      "learning_rate": 1.5073599613935238e-06,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 104780365,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5919557506162448,
+      "grad_norm": 1.833275328046626,
+      "learning_rate": 1.5066050283767574e-06,
+      "loss": 0.5751,
+      "num_input_tokens_seen": 104800765,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5920759935068839,
+      "grad_norm": 2.296490476919945,
+      "learning_rate": 1.5058501702076166e-06,
+      "loss": 0.831,
+      "num_input_tokens_seen": 104817350,
+      "step": 4924
+    },
+    {
+      "epoch": 0.592196236397523,
+      "grad_norm": 2.5717329395143396,
+      "learning_rate": 1.5050953870006112e-06,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 104839370,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5923164792881621,
+      "grad_norm": 3.265502379326082,
+      "learning_rate": 1.504340678870242e-06,
+      "loss": 0.7431,
+      "num_input_tokens_seen": 104857305,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5924367221788012,
+      "grad_norm": 2.1611825172239927,
+      "learning_rate": 1.5035860459309985e-06,
+      "loss": 0.8906,
+      "num_input_tokens_seen": 104874740,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5925569650694402,
+      "grad_norm": 2.291722737396256,
+      "learning_rate": 1.5028314882973568e-06,
+      "loss": 0.629,
+      "num_input_tokens_seen": 104894865,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5926772079600794,
+      "grad_norm": 2.0991735159093627,
+      "learning_rate": 1.502077006083783e-06,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 104913245,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5927974508507184,
+      "grad_norm": 2.173472667579059,
+      "learning_rate": 1.5013225994047315e-06,
+      "loss": 0.7719,
+      "num_input_tokens_seen": 104930595,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5929176937413575,
+      "grad_norm": 2.234205711893548,
+      "learning_rate": 1.5005682683746452e-06,
+      "loss": 0.8013,
+      "num_input_tokens_seen": 104948830,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5930379366319967,
+      "grad_norm": 2.1343571324635477,
+      "learning_rate": 1.4998140131079555e-06,
+      "loss": 0.7213,
+      "num_input_tokens_seen": 104964640,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5931581795226357,
+      "grad_norm": 2.995082430092919,
+      "learning_rate": 1.4990598337190825e-06,
+      "loss": 0.7346,
+      "num_input_tokens_seen": 104980715,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5932784224132748,
+      "grad_norm": 2.262631210577442,
+      "learning_rate": 1.4983057303224336e-06,
+      "loss": 0.6737,
+      "num_input_tokens_seen": 105000250,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5933986653039139,
+      "grad_norm": 1.787158088418598,
+      "learning_rate": 1.4975517030324072e-06,
+      "loss": 0.8608,
+      "num_input_tokens_seen": 105017980,
+      "step": 4935
+    },
+    {
+      "epoch": 0.593518908194553,
+      "grad_norm": 0.8413624732657579,
+      "learning_rate": 1.4967977519633882e-06,
+      "loss": 0.6514,
+      "num_input_tokens_seen": 105075160,
+      "step": 4936
+    },
+    {
+      "epoch": 0.593639151085192,
+      "grad_norm": 2.3749357686263144,
+      "learning_rate": 1.4960438772297498e-06,
+      "loss": 0.7866,
+      "num_input_tokens_seen": 105091925,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5937593939758312,
+      "grad_norm": 2.4666011378311086,
+      "learning_rate": 1.4952900789458545e-06,
+      "loss": 0.7338,
+      "num_input_tokens_seen": 105111410,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5938796368664703,
+      "grad_norm": 2.010896699673101,
+      "learning_rate": 1.4945363572260529e-06,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 105132125,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5939998797571093,
+      "grad_norm": 2.5062262267315596,
+      "learning_rate": 1.4937827121846845e-06,
+      "loss": 0.6732,
+      "num_input_tokens_seen": 105152100,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5941201226477485,
+      "grad_norm": 1.9252044893689988,
+      "learning_rate": 1.4930291439360759e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 105174385,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5942403655383875,
+      "grad_norm": 2.104554654000872,
+      "learning_rate": 1.492275652594542e-06,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 105193415,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5943606084290266,
+      "grad_norm": 0.791358672645102,
+      "learning_rate": 1.4915222382743894e-06,
+      "loss": 0.6184,
+      "num_input_tokens_seen": 105251970,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5944808513196658,
+      "grad_norm": 3.019251210229282,
+      "learning_rate": 1.4907689010899085e-06,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 105269270,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5946010942103048,
+      "grad_norm": 2.1853988812922385,
+      "learning_rate": 1.4900156411553804e-06,
+      "loss": 0.6249,
+      "num_input_tokens_seen": 105288820,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5947213371009439,
+      "grad_norm": 3.5142679560549976,
+      "learning_rate": 1.4892624585850739e-06,
+      "loss": 0.8562,
+      "num_input_tokens_seen": 105306895,
+      "step": 4946
+    },
+    {
+      "epoch": 0.594841579991583,
+      "grad_norm": 2.085009922060645,
+      "learning_rate": 1.4885093534932465e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 105324580,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5949618228822221,
+      "grad_norm": 2.596691222226965,
+      "learning_rate": 1.4877563259941438e-06,
+      "loss": 0.7166,
+      "num_input_tokens_seen": 105342155,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5950820657728612,
+      "grad_norm": 2.2077376289137582,
+      "learning_rate": 1.4870033762019988e-06,
+      "loss": 0.6753,
+      "num_input_tokens_seen": 105362040,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5952023086635003,
+      "grad_norm": 1.7313282846120863,
+      "learning_rate": 1.4862505042310332e-06,
+      "loss": 0.7236,
+      "num_input_tokens_seen": 105381045,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5953225515541394,
+      "grad_norm": 1.8760905122987144,
+      "learning_rate": 1.4854977101954585e-06,
+      "loss": 0.6886,
+      "num_input_tokens_seen": 105402985,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5954427944447784,
+      "grad_norm": 2.4088574329530688,
+      "learning_rate": 1.4847449942094716e-06,
+      "loss": 0.8514,
+      "num_input_tokens_seen": 105421585,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5955630373354175,
+      "grad_norm": 2.1860964710129083,
+      "learning_rate": 1.4839923563872602e-06,
+      "loss": 0.8621,
+      "num_input_tokens_seen": 105439845,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5956832802260567,
+      "grad_norm": 2.348196619975476,
+      "learning_rate": 1.483239796842997e-06,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 105457595,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5958035231166957,
+      "grad_norm": 4.222363764859216,
+      "learning_rate": 1.4824873156908462e-06,
+      "loss": 0.8312,
+      "num_input_tokens_seen": 105475240,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5959237660073348,
+      "grad_norm": 1.8225371821980323,
+      "learning_rate": 1.4817349130449584e-06,
+      "loss": 0.7493,
+      "num_input_tokens_seen": 105494680,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5960440088979739,
+      "grad_norm": 2.0119491302410757,
+      "learning_rate": 1.4809825890194722e-06,
+      "loss": 0.8249,
+      "num_input_tokens_seen": 105513070,
+      "step": 4957
+    },
+    {
+      "epoch": 0.596164251788613,
+      "grad_norm": 1.8592741939123307,
+      "learning_rate": 1.4802303437285139e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 105530060,
+      "step": 4958
+    },
+    {
+      "epoch": 0.596284494679252,
+      "grad_norm": 2.5430385199884515,
+      "learning_rate": 1.4794781772861998e-06,
+      "loss": 0.8006,
+      "num_input_tokens_seen": 105546275,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5964047375698912,
+      "grad_norm": 2.1087308391482207,
+      "learning_rate": 1.4787260898066324e-06,
+      "loss": 0.6654,
+      "num_input_tokens_seen": 105565995,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5965249804605303,
+      "grad_norm": 2.6171970854192477,
+      "learning_rate": 1.4779740814039028e-06,
+      "loss": 0.8491,
+      "num_input_tokens_seen": 105585800,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5966452233511693,
+      "grad_norm": 2.308017448727003,
+      "learning_rate": 1.477222152192089e-06,
+      "loss": 0.6778,
+      "num_input_tokens_seen": 105605545,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5967654662418085,
+      "grad_norm": 5.042651786862969,
+      "learning_rate": 1.4764703022852598e-06,
+      "loss": 0.7328,
+      "num_input_tokens_seen": 105625785,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5968857091324475,
+      "grad_norm": 2.0326451157730454,
+      "learning_rate": 1.4757185317974696e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 105643890,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5970059520230866,
+      "grad_norm": 3.803606333410535,
+      "learning_rate": 1.4749668408427614e-06,
+      "loss": 0.7102,
+      "num_input_tokens_seen": 105663190,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5971261949137258,
+      "grad_norm": 1.9901004225711854,
+      "learning_rate": 1.4742152295351655e-06,
+      "loss": 0.8616,
+      "num_input_tokens_seen": 105682065,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5972464378043648,
+      "grad_norm": 4.005787820222646,
+      "learning_rate": 1.4734636979887016e-06,
+      "loss": 0.6366,
+      "num_input_tokens_seen": 105699245,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5973666806950039,
+      "grad_norm": 2.1013218709649744,
+      "learning_rate": 1.472712246317376e-06,
+      "loss": 0.9003,
+      "num_input_tokens_seen": 105717495,
+      "step": 4968
+    },
+    {
+      "epoch": 0.597486923585643,
+      "grad_norm": 2.212790811263203,
+      "learning_rate": 1.4719608746351834e-06,
+      "loss": 0.6415,
+      "num_input_tokens_seen": 105736775,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5976071664762821,
+      "grad_norm": 3.212131886540118,
+      "learning_rate": 1.4712095830561055e-06,
+      "loss": 0.6975,
+      "num_input_tokens_seen": 105754985,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5977274093669211,
+      "grad_norm": 2.429881239909068,
+      "learning_rate": 1.4704583716941143e-06,
+      "loss": 0.8048,
+      "num_input_tokens_seen": 105773570,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5978476522575603,
+      "grad_norm": 2.421290157295964,
+      "learning_rate": 1.4697072406631672e-06,
+      "loss": 0.7152,
+      "num_input_tokens_seen": 105793195,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5979678951481994,
+      "grad_norm": 2.004255516897595,
+      "learning_rate": 1.4689561900772097e-06,
+      "loss": 0.7265,
+      "num_input_tokens_seen": 105812975,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5980881380388384,
+      "grad_norm": 2.8174139107313603,
+      "learning_rate": 1.4682052200501758e-06,
+      "loss": 0.7247,
+      "num_input_tokens_seen": 105829900,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5982083809294776,
+      "grad_norm": 1.998967924695995,
+      "learning_rate": 1.4674543306959876e-06,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 105849090,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5983286238201166,
+      "grad_norm": 3.2864567987040973,
+      "learning_rate": 1.466703522128554e-06,
+      "loss": 0.8465,
+      "num_input_tokens_seen": 105866450,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5984488667107557,
+      "grad_norm": 1.9857091564113523,
+      "learning_rate": 1.465952794461772e-06,
+      "loss": 0.7335,
+      "num_input_tokens_seen": 105886115,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5985691096013949,
+      "grad_norm": 2.405782674242532,
+      "learning_rate": 1.4652021478095255e-06,
+      "loss": 0.7656,
+      "num_input_tokens_seen": 105904330,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5986893524920339,
+      "grad_norm": 2.0556270955512534,
+      "learning_rate": 1.4644515822856888e-06,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 105922485,
+      "step": 4979
+    },
+    {
+      "epoch": 0.598809595382673,
+      "grad_norm": 0.8015344465263697,
+      "learning_rate": 1.4637010980041215e-06,
+      "loss": 0.5871,
+      "num_input_tokens_seen": 105984315,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5989298382733121,
+      "grad_norm": 3.114296651820016,
+      "learning_rate": 1.4629506950786711e-06,
+      "loss": 0.8972,
+      "num_input_tokens_seen": 106000215,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5990500811639512,
+      "grad_norm": 0.8202103382333706,
+      "learning_rate": 1.4622003736231729e-06,
+      "loss": 0.5862,
+      "num_input_tokens_seen": 106058925,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5991703240545903,
+      "grad_norm": 2.05724219160363,
+      "learning_rate": 1.461450133751451e-06,
+      "loss": 0.7968,
+      "num_input_tokens_seen": 106076715,
+      "step": 4983
+    },
+    {
+      "epoch": 0.5992905669452293,
+      "grad_norm": 1.9559227176330563,
+      "learning_rate": 1.4606999755773153e-06,
+      "loss": 0.7619,
+      "num_input_tokens_seen": 106097640,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5994108098358685,
+      "grad_norm": 3.50697128798293,
+      "learning_rate": 1.4599498992145647e-06,
+      "loss": 0.8124,
+      "num_input_tokens_seen": 106117385,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5995310527265075,
+      "grad_norm": 2.030876906961989,
+      "learning_rate": 1.459199904776984e-06,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 106135960,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5996512956171466,
+      "grad_norm": 2.1728562897198325,
+      "learning_rate": 1.4584499923783486e-06,
+      "loss": 0.7499,
+      "num_input_tokens_seen": 106154260,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5997715385077858,
+      "grad_norm": 2.0056817603173416,
+      "learning_rate": 1.457700162132419e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 106170970,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5998917813984248,
+      "grad_norm": 2.1248362469267583,
+      "learning_rate": 1.4569504141529433e-06,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 106188525,
+      "step": 4989
+    },
+    {
+      "epoch": 0.6000120242890639,
+      "grad_norm": 3.6119619649063766,
+      "learning_rate": 1.456200748553658e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 106206240,
+      "step": 4990
+    },
+    {
+      "epoch": 0.600132267179703,
+      "grad_norm": 1.8177384356022526,
+      "learning_rate": 1.455451165448287e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 106228615,
+      "step": 4991
+    },
+    {
+      "epoch": 0.6002525100703421,
+      "grad_norm": 2.5897948939405757,
+      "learning_rate": 1.4547016649505414e-06,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 106246345,
+      "step": 4992
+    },
+    {
+      "epoch": 0.6003727529609811,
+      "grad_norm": 12.424633892144515,
+      "learning_rate": 1.4539522471741193e-06,
+      "loss": 0.846,
+      "num_input_tokens_seen": 106263490,
+      "step": 4993
+    },
+    {
+      "epoch": 0.6004929958516203,
+      "grad_norm": 2.9335995155779537,
+      "learning_rate": 1.4532029122327063e-06,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 106279995,
+      "step": 4994
+    },
+    {
+      "epoch": 0.6006132387422594,
+      "grad_norm": 3.504969285172157,
+      "learning_rate": 1.4524536602399779e-06,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 106298805,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6007334816328984,
+      "grad_norm": 3.7458148002615292,
+      "learning_rate": 1.4517044913095942e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 106318945,
+      "step": 4996
+    },
+    {
+      "epoch": 0.6008537245235376,
+      "grad_norm": 1.9575211756643291,
+      "learning_rate": 1.4509554055552026e-06,
+      "loss": 0.8038,
+      "num_input_tokens_seen": 106338895,
+      "step": 4997
+    },
+    {
+      "epoch": 0.6009739674141766,
+      "grad_norm": 3.1167229266359033,
+      "learning_rate": 1.450206403090439e-06,
+      "loss": 0.839,
+      "num_input_tokens_seen": 106356810,
+      "step": 4998
+    },
+    {
+      "epoch": 0.6010942103048157,
+      "grad_norm": 2.2724159543601865,
+      "learning_rate": 1.4494574840289274e-06,
+      "loss": 0.8564,
+      "num_input_tokens_seen": 106373645,
+      "step": 4999
+    },
+    {
+      "epoch": 0.6012144531954549,
+      "grad_norm": 2.0981074439106813,
+      "learning_rate": 1.4487086484842782e-06,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 106392010,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6013346960860939,
+      "grad_norm": 2.6937774748522996,
+      "learning_rate": 1.4479598965700883e-06,
+      "loss": 0.599,
+      "num_input_tokens_seen": 106408995,
+      "step": 5001
+    },
+    {
+      "epoch": 0.601454938976733,
+      "grad_norm": 2.6482098193718477,
+      "learning_rate": 1.4472112283999427e-06,
+      "loss": 0.6819,
+      "num_input_tokens_seen": 106427370,
+      "step": 5002
+    },
+    {
+      "epoch": 0.6015751818673721,
+      "grad_norm": 3.2228786947589034,
+      "learning_rate": 1.4464626440874147e-06,
+      "loss": 0.6956,
+      "num_input_tokens_seen": 106446205,
+      "step": 5003
+    },
+    {
+      "epoch": 0.6016954247580112,
+      "grad_norm": 6.177074473220978,
+      "learning_rate": 1.4457141437460636e-06,
+      "loss": 0.741,
+      "num_input_tokens_seen": 106463150,
+      "step": 5004
+    },
+    {
+      "epoch": 0.6018156676486502,
+      "grad_norm": 2.3888212067419667,
+      "learning_rate": 1.4449657274894364e-06,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 106482315,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6019359105392894,
+      "grad_norm": 2.070862719620687,
+      "learning_rate": 1.4442173954310656e-06,
+      "loss": 0.6252,
+      "num_input_tokens_seen": 106504575,
+      "step": 5006
+    },
+    {
+      "epoch": 0.6020561534299285,
+      "grad_norm": 0.8439720855225405,
+      "learning_rate": 1.4434691476844755e-06,
+      "loss": 0.5844,
+      "num_input_tokens_seen": 106565270,
+      "step": 5007
+    },
+    {
+      "epoch": 0.6021763963205675,
+      "grad_norm": 2.370974446366025,
+      "learning_rate": 1.4427209843631729e-06,
+      "loss": 0.6641,
+      "num_input_tokens_seen": 106582040,
+      "step": 5008
+    },
+    {
+      "epoch": 0.6022966392112067,
+      "grad_norm": 2.430251315853662,
+      "learning_rate": 1.4419729055806538e-06,
+      "loss": 0.8157,
+      "num_input_tokens_seen": 106601195,
+      "step": 5009
+    },
+    {
+      "epoch": 0.6024168821018457,
+      "grad_norm": 2.2779119577947817,
+      "learning_rate": 1.441224911450401e-06,
+      "loss": 0.8232,
+      "num_input_tokens_seen": 106616870,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6025371249924848,
+      "grad_norm": 1.861604255120992,
+      "learning_rate": 1.4404770020858851e-06,
+      "loss": 0.8248,
+      "num_input_tokens_seen": 106636075,
+      "step": 5011
+    },
+    {
+      "epoch": 0.602657367883124,
+      "grad_norm": 1.7258341523817926,
+      "learning_rate": 1.4397291776005633e-06,
+      "loss": 0.857,
+      "num_input_tokens_seen": 106656290,
+      "step": 5012
+    },
+    {
+      "epoch": 0.602777610773763,
+      "grad_norm": 2.6731758872586284,
+      "learning_rate": 1.4389814381078797e-06,
+      "loss": 0.7226,
+      "num_input_tokens_seen": 106675250,
+      "step": 5013
+    },
+    {
+      "epoch": 0.6028978536644021,
+      "grad_norm": 2.364498912842888,
+      "learning_rate": 1.438233783721265e-06,
+      "loss": 0.791,
+      "num_input_tokens_seen": 106691135,
+      "step": 5014
+    },
+    {
+      "epoch": 0.6030180965550412,
+      "grad_norm": 2.408715271742756,
+      "learning_rate": 1.4374862145541395e-06,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 106707290,
+      "step": 5015
+    },
+    {
+      "epoch": 0.6031383394456803,
+      "grad_norm": 2.627943053028283,
+      "learning_rate": 1.4367387307199082e-06,
+      "loss": 0.7964,
+      "num_input_tokens_seen": 106723860,
+      "step": 5016
+    },
+    {
+      "epoch": 0.6032585823363193,
+      "grad_norm": 2.3517121038524085,
+      "learning_rate": 1.4359913323319632e-06,
+      "loss": 0.8201,
+      "num_input_tokens_seen": 106740750,
+      "step": 5017
+    },
+    {
+      "epoch": 0.6033788252269584,
+      "grad_norm": 2.234584789360631,
+      "learning_rate": 1.4352440195036847e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 106760645,
+      "step": 5018
+    },
+    {
+      "epoch": 0.6034990681175976,
+      "grad_norm": 2.5540391739281776,
+      "learning_rate": 1.4344967923484395e-06,
+      "loss": 0.7974,
+      "num_input_tokens_seen": 106782335,
+      "step": 5019
+    },
+    {
+      "epoch": 0.6036193110082366,
+      "grad_norm": 10.972469959501462,
+      "learning_rate": 1.4337496509795814e-06,
+      "loss": 0.7177,
+      "num_input_tokens_seen": 106802040,
+      "step": 5020
+    },
+    {
+      "epoch": 0.6037395538988757,
+      "grad_norm": 2.6401056483469008,
+      "learning_rate": 1.433002595510451e-06,
+      "loss": 0.6853,
+      "num_input_tokens_seen": 106820540,
+      "step": 5021
+    },
+    {
+      "epoch": 0.6038597967895148,
+      "grad_norm": 1.7705639102595552,
+      "learning_rate": 1.4322556260543753e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 106836835,
+      "step": 5022
+    },
+    {
+      "epoch": 0.6039800396801539,
+      "grad_norm": 0.9678063972551598,
+      "learning_rate": 1.4315087427246703e-06,
+      "loss": 0.657,
+      "num_input_tokens_seen": 106890380,
+      "step": 5023
+    },
+    {
+      "epoch": 0.604100282570793,
+      "grad_norm": 0.9313413350856867,
+      "learning_rate": 1.4307619456346372e-06,
+      "loss": 0.6199,
+      "num_input_tokens_seen": 106934405,
+      "step": 5024
+    },
+    {
+      "epoch": 0.6042205254614321,
+      "grad_norm": 5.4036484591824765,
+      "learning_rate": 1.4300152348975645e-06,
+      "loss": 0.7282,
+      "num_input_tokens_seen": 106957405,
+      "step": 5025
+    },
+    {
+      "epoch": 0.6043407683520712,
+      "grad_norm": 1.9999655933491143,
+      "learning_rate": 1.429268610626727e-06,
+      "loss": 0.6585,
+      "num_input_tokens_seen": 106979975,
+      "step": 5026
+    },
+    {
+      "epoch": 0.6044610112427102,
+      "grad_norm": 1.7010107526145257,
+      "learning_rate": 1.4285220729353876e-06,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 106998235,
+      "step": 5027
+    },
+    {
+      "epoch": 0.6045812541333494,
+      "grad_norm": 2.500426602665383,
+      "learning_rate": 1.4277756219367957e-06,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 107014980,
+      "step": 5028
+    },
+    {
+      "epoch": 0.6047014970239885,
+      "grad_norm": 2.8071285986868566,
+      "learning_rate": 1.4270292577441866e-06,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 107034205,
+      "step": 5029
+    },
+    {
+      "epoch": 0.6048217399146275,
+      "grad_norm": 1.8683207219837716,
+      "learning_rate": 1.4262829804707831e-06,
+      "loss": 0.7124,
+      "num_input_tokens_seen": 107055915,
+      "step": 5030
+    },
+    {
+      "epoch": 0.6049419828052667,
+      "grad_norm": 1.9613628675111048,
+      "learning_rate": 1.4255367902297958e-06,
+      "loss": 0.6903,
+      "num_input_tokens_seen": 107076965,
+      "step": 5031
+    },
+    {
+      "epoch": 0.6050622256959057,
+      "grad_norm": 2.415337000943828,
+      "learning_rate": 1.424790687134421e-06,
+      "loss": 0.7846,
+      "num_input_tokens_seen": 107092080,
+      "step": 5032
+    },
+    {
+      "epoch": 0.6051824685865448,
+      "grad_norm": 2.760119057147909,
+      "learning_rate": 1.4240446712978415e-06,
+      "loss": 0.751,
+      "num_input_tokens_seen": 107110785,
+      "step": 5033
+    },
+    {
+      "epoch": 0.605302711477184,
+      "grad_norm": 2.191495633778257,
+      "learning_rate": 1.423298742833227e-06,
+      "loss": 0.734,
+      "num_input_tokens_seen": 107129165,
+      "step": 5034
+    },
+    {
+      "epoch": 0.605422954367823,
+      "grad_norm": 1.9591306095223673,
+      "learning_rate": 1.4225529018537352e-06,
+      "loss": 0.7163,
+      "num_input_tokens_seen": 107144390,
+      "step": 5035
+    },
+    {
+      "epoch": 0.6055431972584621,
+      "grad_norm": 1.7703099064437682,
+      "learning_rate": 1.4218071484725086e-06,
+      "loss": 0.7765,
+      "num_input_tokens_seen": 107166230,
+      "step": 5036
+    },
+    {
+      "epoch": 0.6056634401491012,
+      "grad_norm": 2.4377035068953288,
+      "learning_rate": 1.4210614828026786e-06,
+      "loss": 0.7551,
+      "num_input_tokens_seen": 107183800,
+      "step": 5037
+    },
+    {
+      "epoch": 0.6057836830397403,
+      "grad_norm": 1.8072833674767195,
+      "learning_rate": 1.4203159049573601e-06,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 107204755,
+      "step": 5038
+    },
+    {
+      "epoch": 0.6059039259303793,
+      "grad_norm": 2.961817696679899,
+      "learning_rate": 1.4195704150496589e-06,
+      "loss": 0.8743,
+      "num_input_tokens_seen": 107222190,
+      "step": 5039
+    },
+    {
+      "epoch": 0.6060241688210185,
+      "grad_norm": 1.9461971441292651,
+      "learning_rate": 1.4188250131926643e-06,
+      "loss": 0.7301,
+      "num_input_tokens_seen": 107240710,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6061444117116576,
+      "grad_norm": 2.2240655078578695,
+      "learning_rate": 1.418079699499453e-06,
+      "loss": 0.8134,
+      "num_input_tokens_seen": 107257845,
+      "step": 5041
+    },
+    {
+      "epoch": 0.6062646546022966,
+      "grad_norm": 3.0376143383356085,
+      "learning_rate": 1.4173344740830877e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 107276695,
+      "step": 5042
+    },
+    {
+      "epoch": 0.6063848974929358,
+      "grad_norm": 1.99583096715096,
+      "learning_rate": 1.4165893370566202e-06,
+      "loss": 0.7001,
+      "num_input_tokens_seen": 107300170,
+      "step": 5043
+    },
+    {
+      "epoch": 0.6065051403835748,
+      "grad_norm": 3.0366590795962103,
+      "learning_rate": 1.4158442885330865e-06,
+      "loss": 0.768,
+      "num_input_tokens_seen": 107318460,
+      "step": 5044
+    },
+    {
+      "epoch": 0.6066253832742139,
+      "grad_norm": 1.9997730250866659,
+      "learning_rate": 1.4150993286255098e-06,
+      "loss": 0.7882,
+      "num_input_tokens_seen": 107337430,
+      "step": 5045
+    },
+    {
+      "epoch": 0.6067456261648531,
+      "grad_norm": 2.2530438950249616,
+      "learning_rate": 1.4143544574468993e-06,
+      "loss": 0.7833,
+      "num_input_tokens_seen": 107355510,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6068658690554921,
+      "grad_norm": 1.9389585403274003,
+      "learning_rate": 1.4136096751102527e-06,
+      "loss": 0.8134,
+      "num_input_tokens_seen": 107373560,
+      "step": 5047
+    },
+    {
+      "epoch": 0.6069861119461312,
+      "grad_norm": 2.5696302055732496,
+      "learning_rate": 1.4128649817285516e-06,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 107391415,
+      "step": 5048
+    },
+    {
+      "epoch": 0.6071063548367702,
+      "grad_norm": 2.1162437236382994,
+      "learning_rate": 1.4121203774147663e-06,
+      "loss": 0.6263,
+      "num_input_tokens_seen": 107411325,
+      "step": 5049
+    },
+    {
+      "epoch": 0.6072265977274094,
+      "grad_norm": 1.8490427762859487,
+      "learning_rate": 1.4113758622818517e-06,
+      "loss": 0.6976,
+      "num_input_tokens_seen": 107431110,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6073468406180484,
+      "grad_norm": 2.505605132691777,
+      "learning_rate": 1.410631436442751e-06,
+      "loss": 0.8274,
+      "num_input_tokens_seen": 107449625,
+      "step": 5051
+    },
+    {
+      "epoch": 0.6074670835086875,
+      "grad_norm": 2.3471302473028164,
+      "learning_rate": 1.4098871000103936e-06,
+      "loss": 0.8537,
+      "num_input_tokens_seen": 107467945,
+      "step": 5052
+    },
+    {
+      "epoch": 0.6075873263993267,
+      "grad_norm": 1.8458454664124124,
+      "learning_rate": 1.4091428530976935e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 107487905,
+      "step": 5053
+    },
+    {
+      "epoch": 0.6077075692899657,
+      "grad_norm": 3.7489760041746147,
+      "learning_rate": 1.4083986958175524e-06,
+      "loss": 0.794,
+      "num_input_tokens_seen": 107504850,
+      "step": 5054
+    },
+    {
+      "epoch": 0.6078278121806048,
+      "grad_norm": 2.7510365566388324,
+      "learning_rate": 1.4076546282828593e-06,
+      "loss": 0.6922,
+      "num_input_tokens_seen": 107527425,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6079480550712439,
+      "grad_norm": 2.326156358615473,
+      "learning_rate": 1.4069106506064878e-06,
+      "loss": 0.6553,
+      "num_input_tokens_seen": 107548570,
+      "step": 5056
+    },
+    {
+      "epoch": 0.608068297961883,
+      "grad_norm": 2.1852317646184027,
+      "learning_rate": 1.4061667629012993e-06,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 107568960,
+      "step": 5057
+    },
+    {
+      "epoch": 0.608188540852522,
+      "grad_norm": 1.705825279562018,
+      "learning_rate": 1.40542296528014e-06,
+      "loss": 0.8298,
+      "num_input_tokens_seen": 107588340,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6083087837431612,
+      "grad_norm": 2.8962502974448907,
+      "learning_rate": 1.4046792578558445e-06,
+      "loss": 0.7527,
+      "num_input_tokens_seen": 107605955,
+      "step": 5059
+    },
+    {
+      "epoch": 0.6084290266338003,
+      "grad_norm": 2.738061258426552,
+      "learning_rate": 1.4039356407412325e-06,
+      "loss": 0.7595,
+      "num_input_tokens_seen": 107618915,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6085492695244393,
+      "grad_norm": 0.878888277036582,
+      "learning_rate": 1.40319211404911e-06,
+      "loss": 0.6189,
+      "num_input_tokens_seen": 107673635,
+      "step": 5061
+    },
+    {
+      "epoch": 0.6086695124150785,
+      "grad_norm": 2.7188654767061893,
+      "learning_rate": 1.4024486778922691e-06,
+      "loss": 0.8939,
+      "num_input_tokens_seen": 107691670,
+      "step": 5062
+    },
+    {
+      "epoch": 0.6087897553057176,
+      "grad_norm": 2.004022488511787,
+      "learning_rate": 1.4017053323834884e-06,
+      "loss": 0.7706,
+      "num_input_tokens_seen": 107711220,
+      "step": 5063
+    },
+    {
+      "epoch": 0.6089099981963566,
+      "grad_norm": 2.44263165344024,
+      "learning_rate": 1.4009620776355337e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 107732540,
+      "step": 5064
+    },
+    {
+      "epoch": 0.6090302410869958,
+      "grad_norm": 1.9338179863543867,
+      "learning_rate": 1.4002189137611553e-06,
+      "loss": 0.7868,
+      "num_input_tokens_seen": 107751600,
+      "step": 5065
+    },
+    {
+      "epoch": 0.6091504839776348,
+      "grad_norm": 2.294318138425732,
+      "learning_rate": 1.3994758408730897e-06,
+      "loss": 0.6898,
+      "num_input_tokens_seen": 107770505,
+      "step": 5066
+    },
+    {
+      "epoch": 0.6092707268682739,
+      "grad_norm": 2.3712981821250163,
+      "learning_rate": 1.3987328590840629e-06,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 107791170,
+      "step": 5067
+    },
+    {
+      "epoch": 0.609390969758913,
+      "grad_norm": 2.464814302351344,
+      "learning_rate": 1.397989968506783e-06,
+      "loss": 0.8586,
+      "num_input_tokens_seen": 107809900,
+      "step": 5068
+    },
+    {
+      "epoch": 0.6095112126495521,
+      "grad_norm": 2.3178203751821562,
+      "learning_rate": 1.3972471692539462e-06,
+      "loss": 0.7236,
+      "num_input_tokens_seen": 107824335,
+      "step": 5069
+    },
+    {
+      "epoch": 0.6096314555401912,
+      "grad_norm": 2.1312076206288593,
+      "learning_rate": 1.3965044614382344e-06,
+      "loss": 0.7493,
+      "num_input_tokens_seen": 107839505,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6097516984308303,
+      "grad_norm": 3.1315784753043903,
+      "learning_rate": 1.3957618451723162e-06,
+      "loss": 0.7433,
+      "num_input_tokens_seen": 107855255,
+      "step": 5071
+    },
+    {
+      "epoch": 0.6098719413214694,
+      "grad_norm": 1.8880179293800088,
+      "learning_rate": 1.3950193205688457e-06,
+      "loss": 0.6955,
+      "num_input_tokens_seen": 107874700,
+      "step": 5072
+    },
+    {
+      "epoch": 0.6099921842121084,
+      "grad_norm": 2.2981766263854224,
+      "learning_rate": 1.3942768877404631e-06,
+      "loss": 0.8294,
+      "num_input_tokens_seen": 107893385,
+      "step": 5073
+    },
+    {
+      "epoch": 0.6101124271027476,
+      "grad_norm": 1.808031152046907,
+      "learning_rate": 1.3935345467997946e-06,
+      "loss": 0.738,
+      "num_input_tokens_seen": 107912805,
+      "step": 5074
+    },
+    {
+      "epoch": 0.6102326699933867,
+      "grad_norm": 1.868917717542659,
+      "learning_rate": 1.3927922978594536e-06,
+      "loss": 0.6716,
+      "num_input_tokens_seen": 107933610,
+      "step": 5075
+    },
+    {
+      "epoch": 0.6103529128840257,
+      "grad_norm": 0.8637580117701023,
+      "learning_rate": 1.3920501410320382e-06,
+      "loss": 0.623,
+      "num_input_tokens_seen": 107989445,
+      "step": 5076
+    },
+    {
+      "epoch": 0.6104731557746649,
+      "grad_norm": 2.4600870170206157,
+      "learning_rate": 1.3913080764301333e-06,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 108006125,
+      "step": 5077
+    },
+    {
+      "epoch": 0.6105933986653039,
+      "grad_norm": 1.9191960858975163,
+      "learning_rate": 1.3905661041663085e-06,
+      "loss": 0.7144,
+      "num_input_tokens_seen": 108027030,
+      "step": 5078
+    },
+    {
+      "epoch": 0.610713641555943,
+      "grad_norm": 2.7507603481408656,
+      "learning_rate": 1.389824224353122e-06,
+      "loss": 0.6434,
+      "num_input_tokens_seen": 108048340,
+      "step": 5079
+    },
+    {
+      "epoch": 0.610833884446582,
+      "grad_norm": 1.8112865265708167,
+      "learning_rate": 1.3890824371031151e-06,
+      "loss": 0.7621,
+      "num_input_tokens_seen": 108067330,
+      "step": 5080
+    },
+    {
+      "epoch": 0.6109541273372212,
+      "grad_norm": 2.0813021630729747,
+      "learning_rate": 1.3883407425288172e-06,
+      "loss": 0.778,
+      "num_input_tokens_seen": 108087385,
+      "step": 5081
+    },
+    {
+      "epoch": 0.6110743702278603,
+      "grad_norm": 2.3960643168273226,
+      "learning_rate": 1.3875991407427413e-06,
+      "loss": 0.7855,
+      "num_input_tokens_seen": 108105330,
+      "step": 5082
+    },
+    {
+      "epoch": 0.6111946131184993,
+      "grad_norm": 0.7976938898291167,
+      "learning_rate": 1.38685763185739e-06,
+      "loss": 0.6082,
+      "num_input_tokens_seen": 108158710,
+      "step": 5083
+    },
+    {
+      "epoch": 0.6113148560091385,
+      "grad_norm": 3.933877773617746,
+      "learning_rate": 1.386116215985248e-06,
+      "loss": 0.6704,
+      "num_input_tokens_seen": 108176565,
+      "step": 5084
+    },
+    {
+      "epoch": 0.6114350988997775,
+      "grad_norm": 1.9176158619545272,
+      "learning_rate": 1.3853748932387879e-06,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 108196925,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6115553417904166,
+      "grad_norm": 2.5881255362235573,
+      "learning_rate": 1.3846336637304671e-06,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 108214915,
+      "step": 5086
+    },
+    {
+      "epoch": 0.6116755846810558,
+      "grad_norm": 2.074502626765727,
+      "learning_rate": 1.3838925275727312e-06,
+      "loss": 0.825,
+      "num_input_tokens_seen": 108235375,
+      "step": 5087
+    },
+    {
+      "epoch": 0.6117958275716948,
+      "grad_norm": 1.9143854331843104,
+      "learning_rate": 1.3831514848780089e-06,
+      "loss": 0.7878,
+      "num_input_tokens_seen": 108254670,
+      "step": 5088
+    },
+    {
+      "epoch": 0.6119160704623339,
+      "grad_norm": 2.5227524593987023,
+      "learning_rate": 1.3824105357587157e-06,
+      "loss": 0.9193,
+      "num_input_tokens_seen": 108271495,
+      "step": 5089
+    },
+    {
+      "epoch": 0.612036313352973,
+      "grad_norm": 1.4661906185120732,
+      "learning_rate": 1.381669680327253e-06,
+      "loss": 0.8132,
+      "num_input_tokens_seen": 108292895,
+      "step": 5090
+    },
+    {
+      "epoch": 0.6121565562436121,
+      "grad_norm": 2.329621967642062,
+      "learning_rate": 1.3809289186960085e-06,
+      "loss": 0.7058,
+      "num_input_tokens_seen": 108311385,
+      "step": 5091
+    },
+    {
+      "epoch": 0.6122767991342511,
+      "grad_norm": 2.37929946891987,
+      "learning_rate": 1.3801882509773548e-06,
+      "loss": 0.7034,
+      "num_input_tokens_seen": 108328965,
+      "step": 5092
+    },
+    {
+      "epoch": 0.6123970420248903,
+      "grad_norm": 2.1880234398938825,
+      "learning_rate": 1.3794476772836507e-06,
+      "loss": 0.8099,
+      "num_input_tokens_seen": 108349785,
+      "step": 5093
+    },
+    {
+      "epoch": 0.6125172849155294,
+      "grad_norm": 2.2261573745486998,
+      "learning_rate": 1.3787071977272398e-06,
+      "loss": 0.8322,
+      "num_input_tokens_seen": 108368765,
+      "step": 5094
+    },
+    {
+      "epoch": 0.6126375278061684,
+      "grad_norm": 4.73253345130025,
+      "learning_rate": 1.3779668124204535e-06,
+      "loss": 0.7178,
+      "num_input_tokens_seen": 108384900,
+      "step": 5095
+    },
+    {
+      "epoch": 0.6127577706968076,
+      "grad_norm": 1.7417625466686495,
+      "learning_rate": 1.3772265214756074e-06,
+      "loss": 0.8041,
+      "num_input_tokens_seen": 108404380,
+      "step": 5096
+    },
+    {
+      "epoch": 0.6128780135874466,
+      "grad_norm": 6.476177335190081,
+      "learning_rate": 1.376486325005003e-06,
+      "loss": 0.747,
+      "num_input_tokens_seen": 108422340,
+      "step": 5097
+    },
+    {
+      "epoch": 0.6129982564780857,
+      "grad_norm": 1.9089059341196482,
+      "learning_rate": 1.3757462231209267e-06,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 108442365,
+      "step": 5098
+    },
+    {
+      "epoch": 0.6131184993687249,
+      "grad_norm": 1.9468412238329391,
+      "learning_rate": 1.3750062159356525e-06,
+      "loss": 0.8847,
+      "num_input_tokens_seen": 108461435,
+      "step": 5099
+    },
+    {
+      "epoch": 0.6132387422593639,
+      "grad_norm": 2.185121448268844,
+      "learning_rate": 1.3742663035614386e-06,
+      "loss": 0.8219,
+      "num_input_tokens_seen": 108478525,
+      "step": 5100
+    },
+    {
+      "epoch": 0.613358985150003,
+      "grad_norm": 1.954276696659958,
+      "learning_rate": 1.3735264861105287e-06,
+      "loss": 0.7963,
+      "num_input_tokens_seen": 108498885,
+      "step": 5101
+    },
+    {
+      "epoch": 0.6134792280406421,
+      "grad_norm": 2.4360455721530463,
+      "learning_rate": 1.372786763695152e-06,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 108517365,
+      "step": 5102
+    },
+    {
+      "epoch": 0.6135994709312812,
+      "grad_norm": 2.890445004239762,
+      "learning_rate": 1.3720471364275253e-06,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 108536730,
+      "step": 5103
+    },
+    {
+      "epoch": 0.6137197138219203,
+      "grad_norm": 4.70729623187718,
+      "learning_rate": 1.3713076044198486e-06,
+      "loss": 0.7614,
+      "num_input_tokens_seen": 108553260,
+      "step": 5104
+    },
+    {
+      "epoch": 0.6138399567125594,
+      "grad_norm": 4.302229399118108,
+      "learning_rate": 1.3705681677843086e-06,
+      "loss": 0.803,
+      "num_input_tokens_seen": 108571575,
+      "step": 5105
+    },
+    {
+      "epoch": 0.6139601996031985,
+      "grad_norm": 1.0486411974678187,
+      "learning_rate": 1.3698288266330768e-06,
+      "loss": 0.6341,
+      "num_input_tokens_seen": 108631920,
+      "step": 5106
+    },
+    {
+      "epoch": 0.6140804424938375,
+      "grad_norm": 2.3031076818195966,
+      "learning_rate": 1.3690895810783113e-06,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 108650435,
+      "step": 5107
+    },
+    {
+      "epoch": 0.6142006853844767,
+      "grad_norm": 2.461920486158964,
+      "learning_rate": 1.3683504312321548e-06,
+      "loss": 0.7219,
+      "num_input_tokens_seen": 108670490,
+      "step": 5108
+    },
+    {
+      "epoch": 0.6143209282751158,
+      "grad_norm": 2.513553099661042,
+      "learning_rate": 1.3676113772067355e-06,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 108687265,
+      "step": 5109
+    },
+    {
+      "epoch": 0.6144411711657548,
+      "grad_norm": 2.8415894165921163,
+      "learning_rate": 1.3668724191141667e-06,
+      "loss": 0.7134,
+      "num_input_tokens_seen": 108706255,
+      "step": 5110
+    },
+    {
+      "epoch": 0.6145614140563939,
+      "grad_norm": 2.398817076850106,
+      "learning_rate": 1.3661335570665493e-06,
+      "loss": 0.6609,
+      "num_input_tokens_seen": 108723885,
+      "step": 5111
+    },
+    {
+      "epoch": 0.614681656947033,
+      "grad_norm": 2.668097805964205,
+      "learning_rate": 1.3653947911759676e-06,
+      "loss": 0.7009,
+      "num_input_tokens_seen": 108741155,
+      "step": 5112
+    },
+    {
+      "epoch": 0.6148018998376721,
+      "grad_norm": 2.9196111607981154,
+      "learning_rate": 1.3646561215544909e-06,
+      "loss": 0.7357,
+      "num_input_tokens_seen": 108765515,
+      "step": 5113
+    },
+    {
+      "epoch": 0.6149221427283111,
+      "grad_norm": 2.4142253734809955,
+      "learning_rate": 1.3639175483141756e-06,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 108784500,
+      "step": 5114
+    },
+    {
+      "epoch": 0.6150423856189503,
+      "grad_norm": 3.0514754111793874,
+      "learning_rate": 1.3631790715670626e-06,
+      "loss": 0.7289,
+      "num_input_tokens_seen": 108802625,
+      "step": 5115
+    },
+    {
+      "epoch": 0.6151626285095894,
+      "grad_norm": 3.470882610971371,
+      "learning_rate": 1.3624406914251783e-06,
+      "loss": 0.854,
+      "num_input_tokens_seen": 108819465,
+      "step": 5116
+    },
+    {
+      "epoch": 0.6152828714002284,
+      "grad_norm": 2.337631280806907,
+      "learning_rate": 1.361702408000534e-06,
+      "loss": 0.872,
+      "num_input_tokens_seen": 108836085,
+      "step": 5117
+    },
+    {
+      "epoch": 0.6154031142908676,
+      "grad_norm": 2.130656489460872,
+      "learning_rate": 1.3609642214051262e-06,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 108860030,
+      "step": 5118
+    },
+    {
+      "epoch": 0.6155233571815066,
+      "grad_norm": 2.674008879175692,
+      "learning_rate": 1.3602261317509385e-06,
+      "loss": 0.6592,
+      "num_input_tokens_seen": 108876410,
+      "step": 5119
+    },
+    {
+      "epoch": 0.6156436000721457,
+      "grad_norm": 3.3278163004699244,
+      "learning_rate": 1.3594881391499383e-06,
+      "loss": 0.8071,
+      "num_input_tokens_seen": 108895045,
+      "step": 5120
+    },
+    {
+      "epoch": 0.6157638429627849,
+      "grad_norm": 2.2754206884747274,
+      "learning_rate": 1.3587502437140783e-06,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 108912930,
+      "step": 5121
+    },
+    {
+      "epoch": 0.6158840858534239,
+      "grad_norm": 2.377261135494553,
+      "learning_rate": 1.3580124455552952e-06,
+      "loss": 0.8492,
+      "num_input_tokens_seen": 108932015,
+      "step": 5122
+    },
+    {
+      "epoch": 0.616004328744063,
+      "grad_norm": 1.8012426057885011,
+      "learning_rate": 1.3572747447855148e-06,
+      "loss": 0.8588,
+      "num_input_tokens_seen": 108952145,
+      "step": 5123
+    },
+    {
+      "epoch": 0.6161245716347021,
+      "grad_norm": 2.197149964904538,
+      "learning_rate": 1.3565371415166444e-06,
+      "loss": 0.6903,
+      "num_input_tokens_seen": 108969285,
+      "step": 5124
+    },
+    {
+      "epoch": 0.6162448145253412,
+      "grad_norm": 2.4192402461858205,
+      "learning_rate": 1.355799635860578e-06,
+      "loss": 0.6178,
+      "num_input_tokens_seen": 108988925,
+      "step": 5125
+    },
+    {
+      "epoch": 0.6163650574159802,
+      "grad_norm": 2.2475032023258694,
+      "learning_rate": 1.3550622279291941e-06,
+      "loss": 0.6939,
+      "num_input_tokens_seen": 109006790,
+      "step": 5126
+    },
+    {
+      "epoch": 0.6164853003066194,
+      "grad_norm": 1.477741756577462,
+      "learning_rate": 1.354324917834358e-06,
+      "loss": 0.8266,
+      "num_input_tokens_seen": 109027755,
+      "step": 5127
+    },
+    {
+      "epoch": 0.6166055431972585,
+      "grad_norm": 1.8326367285881393,
+      "learning_rate": 1.3535877056879183e-06,
+      "loss": 0.7611,
+      "num_input_tokens_seen": 109045650,
+      "step": 5128
+    },
+    {
+      "epoch": 0.6167257860878975,
+      "grad_norm": 3.0332064234729663,
+      "learning_rate": 1.3528505916017102e-06,
+      "loss": 0.7199,
+      "num_input_tokens_seen": 109063070,
+      "step": 5129
+    },
+    {
+      "epoch": 0.6168460289785367,
+      "grad_norm": 2.070687900871443,
+      "learning_rate": 1.3521135756875514e-06,
+      "loss": 0.8738,
+      "num_input_tokens_seen": 109079105,
+      "step": 5130
+    },
+    {
+      "epoch": 0.6169662718691757,
+      "grad_norm": 2.3074418713180274,
+      "learning_rate": 1.3513766580572492e-06,
+      "loss": 0.8549,
+      "num_input_tokens_seen": 109101645,
+      "step": 5131
+    },
+    {
+      "epoch": 0.6170865147598148,
+      "grad_norm": 2.228157943273862,
+      "learning_rate": 1.3506398388225924e-06,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 109118685,
+      "step": 5132
+    },
+    {
+      "epoch": 0.617206757650454,
+      "grad_norm": 2.1776296404202062,
+      "learning_rate": 1.3499031180953554e-06,
+      "loss": 0.7051,
+      "num_input_tokens_seen": 109137540,
+      "step": 5133
+    },
+    {
+      "epoch": 0.617327000541093,
+      "grad_norm": 2.122194730063685,
+      "learning_rate": 1.349166495987298e-06,
+      "loss": 0.7313,
+      "num_input_tokens_seen": 109155825,
+      "step": 5134
+    },
+    {
+      "epoch": 0.6174472434317321,
+      "grad_norm": 0.9460531365172357,
+      "learning_rate": 1.3484299726101665e-06,
+      "loss": 0.6732,
+      "num_input_tokens_seen": 109219850,
+      "step": 5135
+    },
+    {
+      "epoch": 0.6175674863223712,
+      "grad_norm": 0.9078130896997626,
+      "learning_rate": 1.3476935480756901e-06,
+      "loss": 0.6099,
+      "num_input_tokens_seen": 109276320,
+      "step": 5136
+    },
+    {
+      "epoch": 0.6176877292130103,
+      "grad_norm": 3.958298093492208,
+      "learning_rate": 1.3469572224955833e-06,
+      "loss": 0.7471,
+      "num_input_tokens_seen": 109293835,
+      "step": 5137
+    },
+    {
+      "epoch": 0.6178079721036493,
+      "grad_norm": 3.0116191923119016,
+      "learning_rate": 1.3462209959815462e-06,
+      "loss": 0.7117,
+      "num_input_tokens_seen": 109308295,
+      "step": 5138
+    },
+    {
+      "epoch": 0.6179282149942885,
+      "grad_norm": 2.5285091245357907,
+      "learning_rate": 1.345484868645265e-06,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 109326825,
+      "step": 5139
+    },
+    {
+      "epoch": 0.6180484578849276,
+      "grad_norm": 2.4294778288751253,
+      "learning_rate": 1.3447488405984088e-06,
+      "loss": 0.7742,
+      "num_input_tokens_seen": 109344805,
+      "step": 5140
+    },
+    {
+      "epoch": 0.6181687007755666,
+      "grad_norm": 2.8709401510474573,
+      "learning_rate": 1.3440129119526327e-06,
+      "loss": 0.6862,
+      "num_input_tokens_seen": 109366950,
+      "step": 5141
+    },
+    {
+      "epoch": 0.6182889436662057,
+      "grad_norm": 0.9983529920984433,
+      "learning_rate": 1.3432770828195757e-06,
+      "loss": 0.5663,
+      "num_input_tokens_seen": 109427655,
+      "step": 5142
+    },
+    {
+      "epoch": 0.6184091865568448,
+      "grad_norm": 3.9876579218549963,
+      "learning_rate": 1.3425413533108635e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 109445975,
+      "step": 5143
+    },
+    {
+      "epoch": 0.6185294294474839,
+      "grad_norm": 3.3593234662234357,
+      "learning_rate": 1.341805723538105e-06,
+      "loss": 0.7067,
+      "num_input_tokens_seen": 109465800,
+      "step": 5144
+    },
+    {
+      "epoch": 0.618649672338123,
+      "grad_norm": 1.7613245857542201,
+      "learning_rate": 1.3410701936128952e-06,
+      "loss": 0.768,
+      "num_input_tokens_seen": 109488300,
+      "step": 5145
+    },
+    {
+      "epoch": 0.6187699152287621,
+      "grad_norm": 4.476894421143781,
+      "learning_rate": 1.340334763646812e-06,
+      "loss": 0.8495,
+      "num_input_tokens_seen": 109502155,
+      "step": 5146
+    },
+    {
+      "epoch": 0.6188901581194012,
+      "grad_norm": 2.198068838673797,
+      "learning_rate": 1.3395994337514218e-06,
+      "loss": 0.743,
+      "num_input_tokens_seen": 109522045,
+      "step": 5147
+    },
+    {
+      "epoch": 0.6190104010100402,
+      "grad_norm": 1.7720964647502353,
+      "learning_rate": 1.3388642040382725e-06,
+      "loss": 0.7765,
+      "num_input_tokens_seen": 109542190,
+      "step": 5148
+    },
+    {
+      "epoch": 0.6191306439006794,
+      "grad_norm": 1.907353934457241,
+      "learning_rate": 1.3381290746188975e-06,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 109561280,
+      "step": 5149
+    },
+    {
+      "epoch": 0.6192508867913185,
+      "grad_norm": 7.693214080085688,
+      "learning_rate": 1.3373940456048152e-06,
+      "loss": 0.6719,
+      "num_input_tokens_seen": 109581025,
+      "step": 5150
+    },
+    {
+      "epoch": 0.6193711296819575,
+      "grad_norm": 1.8089479306400549,
+      "learning_rate": 1.3366591171075299e-06,
+      "loss": 0.587,
+      "num_input_tokens_seen": 109604250,
+      "step": 5151
+    },
+    {
+      "epoch": 0.6194913725725967,
+      "grad_norm": 1.9219118460603057,
+      "learning_rate": 1.3359242892385293e-06,
+      "loss": 0.8976,
+      "num_input_tokens_seen": 109623180,
+      "step": 5152
+    },
+    {
+      "epoch": 0.6196116154632357,
+      "grad_norm": 2.040736516095926,
+      "learning_rate": 1.3351895621092859e-06,
+      "loss": 0.7609,
+      "num_input_tokens_seen": 109643245,
+      "step": 5153
+    },
+    {
+      "epoch": 0.6197318583538748,
+      "grad_norm": 2.455665928742031,
+      "learning_rate": 1.3344549358312565e-06,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 109661365,
+      "step": 5154
+    },
+    {
+      "epoch": 0.619852101244514,
+      "grad_norm": 2.789083886177181,
+      "learning_rate": 1.3337204105158847e-06,
+      "loss": 0.7794,
+      "num_input_tokens_seen": 109679955,
+      "step": 5155
+    },
+    {
+      "epoch": 0.619972344135153,
+      "grad_norm": 2.61961396288061,
+      "learning_rate": 1.332985986274597e-06,
+      "loss": 0.7213,
+      "num_input_tokens_seen": 109697305,
+      "step": 5156
+    },
+    {
+      "epoch": 0.6200925870257921,
+      "grad_norm": 2.2157051448615426,
+      "learning_rate": 1.332251663218805e-06,
+      "loss": 0.7502,
+      "num_input_tokens_seen": 109713920,
+      "step": 5157
+    },
+    {
+      "epoch": 0.6202128299164312,
+      "grad_norm": 1.8940261801803904,
+      "learning_rate": 1.3315174414599045e-06,
+      "loss": 0.6736,
+      "num_input_tokens_seen": 109734960,
+      "step": 5158
+    },
+    {
+      "epoch": 0.6203330728070703,
+      "grad_norm": 2.0284374734956896,
+      "learning_rate": 1.3307833211092768e-06,
+      "loss": 0.7456,
+      "num_input_tokens_seen": 109753345,
+      "step": 5159
+    },
+    {
+      "epoch": 0.6204533156977093,
+      "grad_norm": 2.283071985077825,
+      "learning_rate": 1.3300493022782873e-06,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 109773635,
+      "step": 5160
+    },
+    {
+      "epoch": 0.6205735585883485,
+      "grad_norm": 2.3228423976327925,
+      "learning_rate": 1.3293153850782859e-06,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 109791675,
+      "step": 5161
+    },
+    {
+      "epoch": 0.6206938014789876,
+      "grad_norm": 2.396554790449731,
+      "learning_rate": 1.3285815696206065e-06,
+      "loss": 0.7034,
+      "num_input_tokens_seen": 109812940,
+      "step": 5162
+    },
+    {
+      "epoch": 0.6208140443696266,
+      "grad_norm": 2.5119891701219306,
+      "learning_rate": 1.32784785601657e-06,
+      "loss": 0.7633,
+      "num_input_tokens_seen": 109832070,
+      "step": 5163
+    },
+    {
+      "epoch": 0.6209342872602658,
+      "grad_norm": 2.087007035465976,
+      "learning_rate": 1.3271142443774794e-06,
+      "loss": 0.7345,
+      "num_input_tokens_seen": 109854025,
+      "step": 5164
+    },
+    {
+      "epoch": 0.6210545301509048,
+      "grad_norm": 4.797031861823209,
+      "learning_rate": 1.3263807348146233e-06,
+      "loss": 0.815,
+      "num_input_tokens_seen": 109873600,
+      "step": 5165
+    },
+    {
+      "epoch": 0.6211747730415439,
+      "grad_norm": 1.9231491035555146,
+      "learning_rate": 1.3256473274392727e-06,
+      "loss": 0.7352,
+      "num_input_tokens_seen": 109894665,
+      "step": 5166
+    },
+    {
+      "epoch": 0.6212950159321831,
+      "grad_norm": 1.9195123469973472,
+      "learning_rate": 1.3249140223626873e-06,
+      "loss": 0.6984,
+      "num_input_tokens_seen": 109916005,
+      "step": 5167
+    },
+    {
+      "epoch": 0.6214152588228221,
+      "grad_norm": 1.9849459265124092,
+      "learning_rate": 1.3241808196961081e-06,
+      "loss": 0.7588,
+      "num_input_tokens_seen": 109936850,
+      "step": 5168
+    },
+    {
+      "epoch": 0.6215355017134612,
+      "grad_norm": 2.2043648781899225,
+      "learning_rate": 1.3234477195507613e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 109955400,
+      "step": 5169
+    },
+    {
+      "epoch": 0.6216557446041003,
+      "grad_norm": 2.2469087902340252,
+      "learning_rate": 1.322714722037857e-06,
+      "loss": 0.6291,
+      "num_input_tokens_seen": 109976565,
+      "step": 5170
+    },
+    {
+      "epoch": 0.6217759874947394,
+      "grad_norm": 2.264959369627215,
+      "learning_rate": 1.321981827268591e-06,
+      "loss": 0.7678,
+      "num_input_tokens_seen": 109996940,
+      "step": 5171
+    },
+    {
+      "epoch": 0.6218962303853784,
+      "grad_norm": 1.7791049045340406,
+      "learning_rate": 1.321249035354143e-06,
+      "loss": 0.8104,
+      "num_input_tokens_seen": 110018920,
+      "step": 5172
+    },
+    {
+      "epoch": 0.6220164732760175,
+      "grad_norm": 1.987540557729061,
+      "learning_rate": 1.3205163464056766e-06,
+      "loss": 0.7973,
+      "num_input_tokens_seen": 110035245,
+      "step": 5173
+    },
+    {
+      "epoch": 0.6221367161666567,
+      "grad_norm": 2.0960684848173954,
+      "learning_rate": 1.319783760534339e-06,
+      "loss": 0.7203,
+      "num_input_tokens_seen": 110054210,
+      "step": 5174
+    },
+    {
+      "epoch": 0.6222569590572957,
+      "grad_norm": 2.2064178653816136,
+      "learning_rate": 1.3190512778512655e-06,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 110070215,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6223772019479348,
+      "grad_norm": 2.107536350264193,
+      "learning_rate": 1.3183188984675716e-06,
+      "loss": 0.8396,
+      "num_input_tokens_seen": 110088300,
+      "step": 5176
+    },
+    {
+      "epoch": 0.6224974448385739,
+      "grad_norm": 3.023436930587385,
+      "learning_rate": 1.3175866224943586e-06,
+      "loss": 0.7115,
+      "num_input_tokens_seen": 110106740,
+      "step": 5177
+    },
+    {
+      "epoch": 0.622617687729213,
+      "grad_norm": 2.431301220875932,
+      "learning_rate": 1.316854450042712e-06,
+      "loss": 0.7282,
+      "num_input_tokens_seen": 110124400,
+      "step": 5178
+    },
+    {
+      "epoch": 0.622737930619852,
+      "grad_norm": 2.310650671842693,
+      "learning_rate": 1.3161223812237028e-06,
+      "loss": 0.7358,
+      "num_input_tokens_seen": 110143475,
+      "step": 5179
+    },
+    {
+      "epoch": 0.6228581735104912,
+      "grad_norm": 2.6336933873647035,
+      "learning_rate": 1.3153904161483846e-06,
+      "loss": 0.8473,
+      "num_input_tokens_seen": 110158495,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6229784164011303,
+      "grad_norm": 3.1637773947991827,
+      "learning_rate": 1.3146585549277957e-06,
+      "loss": 0.8532,
+      "num_input_tokens_seen": 110176855,
+      "step": 5181
+    },
+    {
+      "epoch": 0.6230986592917693,
+      "grad_norm": 2.575921187182607,
+      "learning_rate": 1.3139267976729587e-06,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 110196765,
+      "step": 5182
+    },
+    {
+      "epoch": 0.6232189021824085,
+      "grad_norm": 3.3904206149608704,
+      "learning_rate": 1.3131951444948815e-06,
+      "loss": 0.7052,
+      "num_input_tokens_seen": 110215885,
+      "step": 5183
+    },
+    {
+      "epoch": 0.6233391450730476,
+      "grad_norm": 2.3995504764809126,
+      "learning_rate": 1.3124635955045546e-06,
+      "loss": 0.7474,
+      "num_input_tokens_seen": 110235420,
+      "step": 5184
+    },
+    {
+      "epoch": 0.6234593879636866,
+      "grad_norm": 3.2262578827880524,
+      "learning_rate": 1.311732150812954e-06,
+      "loss": 0.8349,
+      "num_input_tokens_seen": 110253220,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6235796308543258,
+      "grad_norm": 3.606543327874877,
+      "learning_rate": 1.3110008105310384e-06,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 110272760,
+      "step": 5186
+    },
+    {
+      "epoch": 0.6236998737449648,
+      "grad_norm": 2.0556612667600356,
+      "learning_rate": 1.3102695747697526e-06,
+      "loss": 0.7731,
+      "num_input_tokens_seen": 110295350,
+      "step": 5187
+    },
+    {
+      "epoch": 0.6238201166356039,
+      "grad_norm": 5.111913267016819,
+      "learning_rate": 1.3095384436400237e-06,
+      "loss": 0.8986,
+      "num_input_tokens_seen": 110306600,
+      "step": 5188
+    },
+    {
+      "epoch": 0.623940359526243,
+      "grad_norm": 45.930871334175414,
+      "learning_rate": 1.3088074172527637e-06,
+      "loss": 0.8258,
+      "num_input_tokens_seen": 110323450,
+      "step": 5189
+    },
+    {
+      "epoch": 0.6240606024168821,
+      "grad_norm": 2.030760971136488,
+      "learning_rate": 1.3080764957188684e-06,
+      "loss": 0.7158,
+      "num_input_tokens_seen": 110343415,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6241808453075212,
+      "grad_norm": 2.193745687130762,
+      "learning_rate": 1.3073456791492192e-06,
+      "loss": 0.7002,
+      "num_input_tokens_seen": 110362845,
+      "step": 5191
+    },
+    {
+      "epoch": 0.6243010881981603,
+      "grad_norm": 1.8849290094762998,
+      "learning_rate": 1.3066149676546801e-06,
+      "loss": 0.7838,
+      "num_input_tokens_seen": 110380745,
+      "step": 5192
+    },
+    {
+      "epoch": 0.6244213310887994,
+      "grad_norm": 1.748856858841141,
+      "learning_rate": 1.3058843613460985e-06,
+      "loss": 0.6469,
+      "num_input_tokens_seen": 110398405,
+      "step": 5193
+    },
+    {
+      "epoch": 0.6245415739794384,
+      "grad_norm": 2.2374547719770335,
+      "learning_rate": 1.3051538603343075e-06,
+      "loss": 0.7421,
+      "num_input_tokens_seen": 110416055,
+      "step": 5194
+    },
+    {
+      "epoch": 0.6246618168700776,
+      "grad_norm": 1.8904234910984918,
+      "learning_rate": 1.3044234647301235e-06,
+      "loss": 0.6724,
+      "num_input_tokens_seen": 110433800,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6247820597607167,
+      "grad_norm": 2.0445874268778077,
+      "learning_rate": 1.3036931746443474e-06,
+      "loss": 0.7158,
+      "num_input_tokens_seen": 110450995,
+      "step": 5196
+    },
+    {
+      "epoch": 0.6249023026513557,
+      "grad_norm": 4.7571959269373005,
+      "learning_rate": 1.3029629901877625e-06,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 110470090,
+      "step": 5197
+    },
+    {
+      "epoch": 0.6250225455419949,
+      "grad_norm": 2.9820216325560542,
+      "learning_rate": 1.3022329114711371e-06,
+      "loss": 0.7758,
+      "num_input_tokens_seen": 110488520,
+      "step": 5198
+    },
+    {
+      "epoch": 0.6251427884326339,
+      "grad_norm": 2.59725265414932,
+      "learning_rate": 1.3015029386052252e-06,
+      "loss": 0.6942,
+      "num_input_tokens_seen": 110508410,
+      "step": 5199
+    },
+    {
+      "epoch": 0.625263031323273,
+      "grad_norm": 2.011996771573217,
+      "learning_rate": 1.3007730717007622e-06,
+      "loss": 0.7213,
+      "num_input_tokens_seen": 110528945,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6253832742139122,
+      "grad_norm": 2.9973102063433426,
+      "learning_rate": 1.300043310868468e-06,
+      "loss": 0.7536,
+      "num_input_tokens_seen": 110549165,
+      "step": 5201
+    },
+    {
+      "epoch": 0.6255035171045512,
+      "grad_norm": 2.473410629263274,
+      "learning_rate": 1.2993136562190467e-06,
+      "loss": 0.7952,
+      "num_input_tokens_seen": 110568005,
+      "step": 5202
+    },
+    {
+      "epoch": 0.6256237599951903,
+      "grad_norm": 1.564251548673153,
+      "learning_rate": 1.2985841078631871e-06,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 110587045,
+      "step": 5203
+    },
+    {
+      "epoch": 0.6257440028858293,
+      "grad_norm": 2.0462399026820357,
+      "learning_rate": 1.2978546659115608e-06,
+      "loss": 0.7742,
+      "num_input_tokens_seen": 110604845,
+      "step": 5204
+    },
+    {
+      "epoch": 0.6258642457764685,
+      "grad_norm": 2.1250060517098364,
+      "learning_rate": 1.2971253304748234e-06,
+      "loss": 0.8496,
+      "num_input_tokens_seen": 110622280,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6259844886671075,
+      "grad_norm": 1.7850933096364225,
+      "learning_rate": 1.2963961016636136e-06,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 110638560,
+      "step": 5206
+    },
+    {
+      "epoch": 0.6261047315577466,
+      "grad_norm": 2.318819752791727,
+      "learning_rate": 1.2956669795885565e-06,
+      "loss": 0.8369,
+      "num_input_tokens_seen": 110654910,
+      "step": 5207
+    },
+    {
+      "epoch": 0.6262249744483858,
+      "grad_norm": 2.3073701442344055,
+      "learning_rate": 1.294937964360259e-06,
+      "loss": 0.6897,
+      "num_input_tokens_seen": 110674900,
+      "step": 5208
+    },
+    {
+      "epoch": 0.6263452173390248,
+      "grad_norm": 2.4655920038086943,
+      "learning_rate": 1.2942090560893112e-06,
+      "loss": 0.6928,
+      "num_input_tokens_seen": 110694025,
+      "step": 5209
+    },
+    {
+      "epoch": 0.6264654602296639,
+      "grad_norm": 2.2813960820589454,
+      "learning_rate": 1.2934802548862878e-06,
+      "loss": 0.5986,
+      "num_input_tokens_seen": 110716530,
+      "step": 5210
+    },
+    {
+      "epoch": 0.626585703120303,
+      "grad_norm": 3.2720546801358266,
+      "learning_rate": 1.292751560861749e-06,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 110731155,
+      "step": 5211
+    },
+    {
+      "epoch": 0.6267059460109421,
+      "grad_norm": 1.9641602027685314,
+      "learning_rate": 1.2920229741262356e-06,
+      "loss": 0.7927,
+      "num_input_tokens_seen": 110748880,
+      "step": 5212
+    },
+    {
+      "epoch": 0.6268261889015811,
+      "grad_norm": 2.48002937166285,
+      "learning_rate": 1.2912944947902745e-06,
+      "loss": 0.74,
+      "num_input_tokens_seen": 110765085,
+      "step": 5213
+    },
+    {
+      "epoch": 0.6269464317922203,
+      "grad_norm": 3.052319556127872,
+      "learning_rate": 1.2905661229643742e-06,
+      "loss": 0.7162,
+      "num_input_tokens_seen": 110784565,
+      "step": 5214
+    },
+    {
+      "epoch": 0.6270666746828594,
+      "grad_norm": 2.395072844226662,
+      "learning_rate": 1.2898378587590299e-06,
+      "loss": 0.8397,
+      "num_input_tokens_seen": 110800885,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6271869175734984,
+      "grad_norm": 2.356659738438972,
+      "learning_rate": 1.2891097022847177e-06,
+      "loss": 0.8693,
+      "num_input_tokens_seen": 110817950,
+      "step": 5216
+    },
+    {
+      "epoch": 0.6273071604641376,
+      "grad_norm": 2.3337986702990676,
+      "learning_rate": 1.288381653651898e-06,
+      "loss": 0.6687,
+      "num_input_tokens_seen": 110838810,
+      "step": 5217
+    },
+    {
+      "epoch": 0.6274274033547766,
+      "grad_norm": 2.413457694513621,
+      "learning_rate": 1.2876537129710155e-06,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 110856260,
+      "step": 5218
+    },
+    {
+      "epoch": 0.6275476462454157,
+      "grad_norm": 2.358554182318023,
+      "learning_rate": 1.2869258803524987e-06,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 110874840,
+      "step": 5219
+    },
+    {
+      "epoch": 0.6276678891360549,
+      "grad_norm": 1.7908847530506782,
+      "learning_rate": 1.2861981559067592e-06,
+      "loss": 0.6999,
+      "num_input_tokens_seen": 110895165,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6277881320266939,
+      "grad_norm": 2.031556287969863,
+      "learning_rate": 1.2854705397441917e-06,
+      "loss": 0.7995,
+      "num_input_tokens_seen": 110910425,
+      "step": 5221
+    },
+    {
+      "epoch": 0.627908374917333,
+      "grad_norm": 3.001299653552301,
+      "learning_rate": 1.2847430319751747e-06,
+      "loss": 0.7721,
+      "num_input_tokens_seen": 110928240,
+      "step": 5222
+    },
+    {
+      "epoch": 0.6280286178079721,
+      "grad_norm": 2.7303845128361237,
+      "learning_rate": 1.2840156327100716e-06,
+      "loss": 0.6697,
+      "num_input_tokens_seen": 110945085,
+      "step": 5223
+    },
+    {
+      "epoch": 0.6281488606986112,
+      "grad_norm": 4.18336343067612,
+      "learning_rate": 1.2832883420592278e-06,
+      "loss": 0.7175,
+      "num_input_tokens_seen": 110963700,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6282691035892503,
+      "grad_norm": 2.282553383531497,
+      "learning_rate": 1.2825611601329725e-06,
+      "loss": 0.638,
+      "num_input_tokens_seen": 110983940,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6283893464798894,
+      "grad_norm": 2.082559058799252,
+      "learning_rate": 1.2818340870416182e-06,
+      "loss": 0.805,
+      "num_input_tokens_seen": 111004795,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6285095893705285,
+      "grad_norm": 4.089667592479948,
+      "learning_rate": 1.2811071228954626e-06,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 111023150,
+      "step": 5227
+    },
+    {
+      "epoch": 0.6286298322611675,
+      "grad_norm": 2.106316160231089,
+      "learning_rate": 1.2803802678047846e-06,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 111043020,
+      "step": 5228
+    },
+    {
+      "epoch": 0.6287500751518067,
+      "grad_norm": 2.1346415468226847,
+      "learning_rate": 1.2796535218798483e-06,
+      "loss": 0.73,
+      "num_input_tokens_seen": 111062805,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6288703180424458,
+      "grad_norm": 2.250278938404751,
+      "learning_rate": 1.2789268852308992e-06,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 111077735,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6289905609330848,
+      "grad_norm": 2.4570592454405453,
+      "learning_rate": 1.2782003579681688e-06,
+      "loss": 0.6954,
+      "num_input_tokens_seen": 111096985,
+      "step": 5231
+    },
+    {
+      "epoch": 0.629110803823724,
+      "grad_norm": 2.5659629633958296,
+      "learning_rate": 1.2774739402018701e-06,
+      "loss": 0.7351,
+      "num_input_tokens_seen": 111117540,
+      "step": 5232
+    },
+    {
+      "epoch": 0.629231046714363,
+      "grad_norm": 1.8538971436082687,
+      "learning_rate": 1.2767476320422006e-06,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 111137185,
+      "step": 5233
+    },
+    {
+      "epoch": 0.6293512896050021,
+      "grad_norm": 0.7297993800686974,
+      "learning_rate": 1.2760214335993392e-06,
+      "loss": 0.5962,
+      "num_input_tokens_seen": 111203550,
+      "step": 5234
+    },
+    {
+      "epoch": 0.6294715324956413,
+      "grad_norm": 2.06204278575427,
+      "learning_rate": 1.2752953449834514e-06,
+      "loss": 0.586,
+      "num_input_tokens_seen": 111225720,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6295917753862803,
+      "grad_norm": 1.9061635460802566,
+      "learning_rate": 1.2745693663046836e-06,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 111244510,
+      "step": 5236
+    },
+    {
+      "epoch": 0.6297120182769194,
+      "grad_norm": 1.9899008352941812,
+      "learning_rate": 1.2738434976731662e-06,
+      "loss": 0.7991,
+      "num_input_tokens_seen": 111262415,
+      "step": 5237
+    },
+    {
+      "epoch": 0.6298322611675584,
+      "grad_norm": 1.738945131494004,
+      "learning_rate": 1.2731177391990125e-06,
+      "loss": 0.7481,
+      "num_input_tokens_seen": 111282060,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6299525040581976,
+      "grad_norm": 16.045329240179676,
+      "learning_rate": 1.2723920909923203e-06,
+      "loss": 0.8142,
+      "num_input_tokens_seen": 111297525,
+      "step": 5239
+    },
+    {
+      "epoch": 0.6300727469488366,
+      "grad_norm": 0.9340783517036609,
+      "learning_rate": 1.2716665531631692e-06,
+      "loss": 0.6518,
+      "num_input_tokens_seen": 111351530,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6301929898394757,
+      "grad_norm": 2.1276929111388667,
+      "learning_rate": 1.270941125821623e-06,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 111371675,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6303132327301149,
+      "grad_norm": 1.7322007291532513,
+      "learning_rate": 1.2702158090777273e-06,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 111392485,
+      "step": 5242
+    },
+    {
+      "epoch": 0.6304334756207539,
+      "grad_norm": 2.2456092856542367,
+      "learning_rate": 1.2694906030415141e-06,
+      "loss": 0.744,
+      "num_input_tokens_seen": 111409950,
+      "step": 5243
+    },
+    {
+      "epoch": 0.630553718511393,
+      "grad_norm": 2.712737074675974,
+      "learning_rate": 1.2687655078229958e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 111424000,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6306739614020321,
+      "grad_norm": 2.1064104580625087,
+      "learning_rate": 1.2680405235321683e-06,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 111445055,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6307942042926712,
+      "grad_norm": 10.412549822211044,
+      "learning_rate": 1.267315650279011e-06,
+      "loss": 0.7828,
+      "num_input_tokens_seen": 111463245,
+      "step": 5246
+    },
+    {
+      "epoch": 0.6309144471833102,
+      "grad_norm": 2.3720885630859287,
+      "learning_rate": 1.2665908881734874e-06,
+      "loss": 0.74,
+      "num_input_tokens_seen": 111481800,
+      "step": 5247
+    },
+    {
+      "epoch": 0.6310346900739494,
+      "grad_norm": 2.6979891231226505,
+      "learning_rate": 1.2658662373255432e-06,
+      "loss": 0.846,
+      "num_input_tokens_seen": 111499910,
+      "step": 5248
+    },
+    {
+      "epoch": 0.6311549329645885,
+      "grad_norm": 0.8806303462501649,
+      "learning_rate": 1.265141697845107e-06,
+      "loss": 0.5723,
+      "num_input_tokens_seen": 111565015,
+      "step": 5249
+    },
+    {
+      "epoch": 0.6312751758552275,
+      "grad_norm": 2.2954593267314936,
+      "learning_rate": 1.2644172698420899e-06,
+      "loss": 0.6465,
+      "num_input_tokens_seen": 111586840,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6313954187458667,
+      "grad_norm": 1.8471206967788458,
+      "learning_rate": 1.2636929534263894e-06,
+      "loss": 0.8412,
+      "num_input_tokens_seen": 111605545,
+      "step": 5251
+    },
+    {
+      "epoch": 0.6315156616365057,
+      "grad_norm": 2.3994975481123513,
+      "learning_rate": 1.2629687487078821e-06,
+      "loss": 0.7681,
+      "num_input_tokens_seen": 111624075,
+      "step": 5252
+    },
+    {
+      "epoch": 0.6316359045271448,
+      "grad_norm": 4.065423323777952,
+      "learning_rate": 1.2622446557964298e-06,
+      "loss": 0.7555,
+      "num_input_tokens_seen": 111641800,
+      "step": 5253
+    },
+    {
+      "epoch": 0.631756147417784,
+      "grad_norm": 1.6867561317674031,
+      "learning_rate": 1.2615206748018757e-06,
+      "loss": 0.7012,
+      "num_input_tokens_seen": 111662115,
+      "step": 5254
+    },
+    {
+      "epoch": 0.631876390308423,
+      "grad_norm": 2.980653683989346,
+      "learning_rate": 1.2607968058340488e-06,
+      "loss": 0.7213,
+      "num_input_tokens_seen": 111681530,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6319966331990621,
+      "grad_norm": 2.410032744526805,
+      "learning_rate": 1.2600730490027586e-06,
+      "loss": 0.7286,
+      "num_input_tokens_seen": 111701490,
+      "step": 5256
+    },
+    {
+      "epoch": 0.6321168760897012,
+      "grad_norm": 1.8658546078770732,
+      "learning_rate": 1.2593494044177986e-06,
+      "loss": 0.799,
+      "num_input_tokens_seen": 111719515,
+      "step": 5257
+    },
+    {
+      "epoch": 0.6322371189803403,
+      "grad_norm": 2.3220802559303566,
+      "learning_rate": 1.2586258721889448e-06,
+      "loss": 0.7953,
+      "num_input_tokens_seen": 111736585,
+      "step": 5258
+    },
+    {
+      "epoch": 0.6323573618709794,
+      "grad_norm": 2.2001562839333415,
+      "learning_rate": 1.2579024524259573e-06,
+      "loss": 0.8064,
+      "num_input_tokens_seen": 111752565,
+      "step": 5259
+    },
+    {
+      "epoch": 0.6324776047616185,
+      "grad_norm": 1.8832792685980189,
+      "learning_rate": 1.2571791452385775e-06,
+      "loss": 0.9057,
+      "num_input_tokens_seen": 111769550,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6325978476522576,
+      "grad_norm": 1.6084239787433037,
+      "learning_rate": 1.2564559507365306e-06,
+      "loss": 0.7635,
+      "num_input_tokens_seen": 111791675,
+      "step": 5261
+    },
+    {
+      "epoch": 0.6327180905428966,
+      "grad_norm": 2.06494265719428,
+      "learning_rate": 1.2557328690295244e-06,
+      "loss": 0.7877,
+      "num_input_tokens_seen": 111809585,
+      "step": 5262
+    },
+    {
+      "epoch": 0.6328383334335358,
+      "grad_norm": 6.204705968815963,
+      "learning_rate": 1.2550099002272506e-06,
+      "loss": 0.7563,
+      "num_input_tokens_seen": 111828330,
+      "step": 5263
+    },
+    {
+      "epoch": 0.6329585763241748,
+      "grad_norm": 2.4256622602510536,
+      "learning_rate": 1.254287044439383e-06,
+      "loss": 0.7926,
+      "num_input_tokens_seen": 111847655,
+      "step": 5264
+    },
+    {
+      "epoch": 0.6330788192148139,
+      "grad_norm": 0.8313187081462069,
+      "learning_rate": 1.2535643017755776e-06,
+      "loss": 0.5632,
+      "num_input_tokens_seen": 111909565,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6331990621054531,
+      "grad_norm": 2.443150304967501,
+      "learning_rate": 1.2528416723454737e-06,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 111925955,
+      "step": 5266
+    },
+    {
+      "epoch": 0.6333193049960921,
+      "grad_norm": 7.38401961084352,
+      "learning_rate": 1.2521191562586945e-06,
+      "loss": 0.7059,
+      "num_input_tokens_seen": 111949325,
+      "step": 5267
+    },
+    {
+      "epoch": 0.6334395478867312,
+      "grad_norm": 2.217107078064154,
+      "learning_rate": 1.251396753624845e-06,
+      "loss": 0.7686,
+      "num_input_tokens_seen": 111965365,
+      "step": 5268
+    },
+    {
+      "epoch": 0.6335597907773702,
+      "grad_norm": 2.739152607110941,
+      "learning_rate": 1.2506744645535122e-06,
+      "loss": 0.8047,
+      "num_input_tokens_seen": 111985515,
+      "step": 5269
+    },
+    {
+      "epoch": 0.6336800336680094,
+      "grad_norm": 4.005794505068641,
+      "learning_rate": 1.2499522891542667e-06,
+      "loss": 0.5896,
+      "num_input_tokens_seen": 112005275,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6338002765586485,
+      "grad_norm": 1.8565679058295572,
+      "learning_rate": 1.2492302275366635e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 112024670,
+      "step": 5271
+    },
+    {
+      "epoch": 0.6339205194492875,
+      "grad_norm": 3.329445595786218,
+      "learning_rate": 1.2485082798102377e-06,
+      "loss": 0.6509,
+      "num_input_tokens_seen": 112044805,
+      "step": 5272
+    },
+    {
+      "epoch": 0.6340407623399267,
+      "grad_norm": 5.293613099723363,
+      "learning_rate": 1.2477864460845086e-06,
+      "loss": 0.687,
+      "num_input_tokens_seen": 112060925,
+      "step": 5273
+    },
+    {
+      "epoch": 0.6341610052305657,
+      "grad_norm": 5.0474857942866045,
+      "learning_rate": 1.247064726468977e-06,
+      "loss": 0.7385,
+      "num_input_tokens_seen": 112079125,
+      "step": 5274
+    },
+    {
+      "epoch": 0.6342812481212048,
+      "grad_norm": 3.908715415333567,
+      "learning_rate": 1.2463431210731282e-06,
+      "loss": 0.7127,
+      "num_input_tokens_seen": 112098430,
+      "step": 5275
+    },
+    {
+      "epoch": 0.634401491011844,
+      "grad_norm": 2.3496489092615525,
+      "learning_rate": 1.2456216300064289e-06,
+      "loss": 0.7557,
+      "num_input_tokens_seen": 112115700,
+      "step": 5276
+    },
+    {
+      "epoch": 0.634521733902483,
+      "grad_norm": 2.366313895944972,
+      "learning_rate": 1.2449002533783284e-06,
+      "loss": 0.775,
+      "num_input_tokens_seen": 112135475,
+      "step": 5277
+    },
+    {
+      "epoch": 0.6346419767931221,
+      "grad_norm": 2.324993340810823,
+      "learning_rate": 1.2441789912982579e-06,
+      "loss": 0.6925,
+      "num_input_tokens_seen": 112152280,
+      "step": 5278
+    },
+    {
+      "epoch": 0.6347622196837612,
+      "grad_norm": 2.3095453146778793,
+      "learning_rate": 1.2434578438756346e-06,
+      "loss": 0.6462,
+      "num_input_tokens_seen": 112172430,
+      "step": 5279
+    },
+    {
+      "epoch": 0.6348824625744003,
+      "grad_norm": 2.5763976446510513,
+      "learning_rate": 1.242736811219855e-06,
+      "loss": 0.7824,
+      "num_input_tokens_seen": 112198110,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6350027054650393,
+      "grad_norm": 1.8010468456990276,
+      "learning_rate": 1.2420158934402988e-06,
+      "loss": 0.8148,
+      "num_input_tokens_seen": 112218445,
+      "step": 5281
+    },
+    {
+      "epoch": 0.6351229483556785,
+      "grad_norm": 1.9541332765371024,
+      "learning_rate": 1.2412950906463286e-06,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 112235470,
+      "step": 5282
+    },
+    {
+      "epoch": 0.6352431912463176,
+      "grad_norm": 1.932823796223964,
+      "learning_rate": 1.2405744029472902e-06,
+      "loss": 0.8892,
+      "num_input_tokens_seen": 112254675,
+      "step": 5283
+    },
+    {
+      "epoch": 0.6353634341369566,
+      "grad_norm": 2.3826998521820535,
+      "learning_rate": 1.2398538304525113e-06,
+      "loss": 0.7591,
+      "num_input_tokens_seen": 112273020,
+      "step": 5284
+    },
+    {
+      "epoch": 0.6354836770275958,
+      "grad_norm": 2.3542005417520295,
+      "learning_rate": 1.2391333732713016e-06,
+      "loss": 0.7527,
+      "num_input_tokens_seen": 112290545,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6356039199182348,
+      "grad_norm": 3.218954584587067,
+      "learning_rate": 1.2384130315129536e-06,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 112308590,
+      "step": 5286
+    },
+    {
+      "epoch": 0.6357241628088739,
+      "grad_norm": 2.6630769999364166,
+      "learning_rate": 1.2376928052867442e-06,
+      "loss": 0.7259,
+      "num_input_tokens_seen": 112327430,
+      "step": 5287
+    },
+    {
+      "epoch": 0.6358444056995131,
+      "grad_norm": 2.591191587501824,
+      "learning_rate": 1.2369726947019299e-06,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 112347625,
+      "step": 5288
+    },
+    {
+      "epoch": 0.6359646485901521,
+      "grad_norm": 2.454139769225393,
+      "learning_rate": 1.2362526998677516e-06,
+      "loss": 0.6645,
+      "num_input_tokens_seen": 112363710,
+      "step": 5289
+    },
+    {
+      "epoch": 0.6360848914807912,
+      "grad_norm": 1.9795957021037067,
+      "learning_rate": 1.2355328208934301e-06,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 112382305,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6362051343714303,
+      "grad_norm": 1.849343974248388,
+      "learning_rate": 1.2348130578881728e-06,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 112400245,
+      "step": 5291
+    },
+    {
+      "epoch": 0.6363253772620694,
+      "grad_norm": 4.805070229333753,
+      "learning_rate": 1.2340934109611664e-06,
+      "loss": 0.7536,
+      "num_input_tokens_seen": 112420725,
+      "step": 5292
+    },
+    {
+      "epoch": 0.6364456201527084,
+      "grad_norm": 3.299814878926741,
+      "learning_rate": 1.2333738802215803e-06,
+      "loss": 0.6807,
+      "num_input_tokens_seen": 112440665,
+      "step": 5293
+    },
+    {
+      "epoch": 0.6365658630433476,
+      "grad_norm": 2.514701341713591,
+      "learning_rate": 1.2326544657785668e-06,
+      "loss": 0.8081,
+      "num_input_tokens_seen": 112460075,
+      "step": 5294
+    },
+    {
+      "epoch": 0.6366861059339867,
+      "grad_norm": 2.5878788315120347,
+      "learning_rate": 1.2319351677412612e-06,
+      "loss": 0.7374,
+      "num_input_tokens_seen": 112476840,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6368063488246257,
+      "grad_norm": 1.9853188416144236,
+      "learning_rate": 1.2312159862187796e-06,
+      "loss": 0.7359,
+      "num_input_tokens_seen": 112494970,
+      "step": 5296
+    },
+    {
+      "epoch": 0.6369265917152649,
+      "grad_norm": 1.7032026923010202,
+      "learning_rate": 1.2304969213202221e-06,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 112515950,
+      "step": 5297
+    },
+    {
+      "epoch": 0.6370468346059039,
+      "grad_norm": 2.9977384166703778,
+      "learning_rate": 1.2297779731546687e-06,
+      "loss": 0.7892,
+      "num_input_tokens_seen": 112534765,
+      "step": 5298
+    },
+    {
+      "epoch": 0.637167077496543,
+      "grad_norm": 2.1589928030699017,
+      "learning_rate": 1.2290591418311853e-06,
+      "loss": 0.7768,
+      "num_input_tokens_seen": 112551880,
+      "step": 5299
+    },
+    {
+      "epoch": 0.637287320387182,
+      "grad_norm": 1.9339938587171805,
+      "learning_rate": 1.2283404274588172e-06,
+      "loss": 0.704,
+      "num_input_tokens_seen": 112570545,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6374075632778212,
+      "grad_norm": 0.7951509615224558,
+      "learning_rate": 1.2276218301465925e-06,
+      "loss": 0.5556,
+      "num_input_tokens_seen": 112625630,
+      "step": 5301
+    },
+    {
+      "epoch": 0.6375278061684603,
+      "grad_norm": 1.8959663808941072,
+      "learning_rate": 1.2269033500035217e-06,
+      "loss": 0.7835,
+      "num_input_tokens_seen": 112645485,
+      "step": 5302
+    },
+    {
+      "epoch": 0.6376480490590993,
+      "grad_norm": 1.946591753994788,
+      "learning_rate": 1.2261849871385988e-06,
+      "loss": 0.7383,
+      "num_input_tokens_seen": 112666310,
+      "step": 5303
+    },
+    {
+      "epoch": 0.6377682919497385,
+      "grad_norm": 2.4535601168950643,
+      "learning_rate": 1.2254667416607976e-06,
+      "loss": 0.6121,
+      "num_input_tokens_seen": 112687630,
+      "step": 5304
+    },
+    {
+      "epoch": 0.6378885348403776,
+      "grad_norm": 2.0176746018851457,
+      "learning_rate": 1.2247486136790762e-06,
+      "loss": 0.8319,
+      "num_input_tokens_seen": 112706830,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6380087777310166,
+      "grad_norm": 2.2291941657011063,
+      "learning_rate": 1.2240306033023726e-06,
+      "loss": 0.7976,
+      "num_input_tokens_seen": 112724375,
+      "step": 5306
+    },
+    {
+      "epoch": 0.6381290206216558,
+      "grad_norm": 2.0293592734779153,
+      "learning_rate": 1.2233127106396106e-06,
+      "loss": 0.7174,
+      "num_input_tokens_seen": 112742815,
+      "step": 5307
+    },
+    {
+      "epoch": 0.6382492635122948,
+      "grad_norm": 3.711336783216749,
+      "learning_rate": 1.2225949357996928e-06,
+      "loss": 0.8587,
+      "num_input_tokens_seen": 112760660,
+      "step": 5308
+    },
+    {
+      "epoch": 0.6383695064029339,
+      "grad_norm": 9.492569741622521,
+      "learning_rate": 1.221877278891505e-06,
+      "loss": 0.8002,
+      "num_input_tokens_seen": 112779635,
+      "step": 5309
+    },
+    {
+      "epoch": 0.638489749293573,
+      "grad_norm": 4.011802883703717,
+      "learning_rate": 1.221159740023915e-06,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 112799185,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6386099921842121,
+      "grad_norm": 2.446543728621817,
+      "learning_rate": 1.2204423193057735e-06,
+      "loss": 0.7302,
+      "num_input_tokens_seen": 112817735,
+      "step": 5311
+    },
+    {
+      "epoch": 0.6387302350748512,
+      "grad_norm": 0.9916017969666181,
+      "learning_rate": 1.2197250168459126e-06,
+      "loss": 0.6863,
+      "num_input_tokens_seen": 112873855,
+      "step": 5312
+    },
+    {
+      "epoch": 0.6388504779654903,
+      "grad_norm": 2.4507939987983143,
+      "learning_rate": 1.2190078327531458e-06,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 112889820,
+      "step": 5313
+    },
+    {
+      "epoch": 0.6389707208561294,
+      "grad_norm": 2.067654429337552,
+      "learning_rate": 1.2182907671362693e-06,
+      "loss": 0.7231,
+      "num_input_tokens_seen": 112910235,
+      "step": 5314
+    },
+    {
+      "epoch": 0.6390909637467684,
+      "grad_norm": 1.9994166148385504,
+      "learning_rate": 1.2175738201040626e-06,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 112926995,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6392112066374076,
+      "grad_norm": 2.0051756709731685,
+      "learning_rate": 1.2168569917652855e-06,
+      "loss": 0.7893,
+      "num_input_tokens_seen": 112946570,
+      "step": 5316
+    },
+    {
+      "epoch": 0.6393314495280467,
+      "grad_norm": 1.9420777912835931,
+      "learning_rate": 1.2161402822286802e-06,
+      "loss": 0.6367,
+      "num_input_tokens_seen": 112966975,
+      "step": 5317
+    },
+    {
+      "epoch": 0.6394516924186857,
+      "grad_norm": 2.2461965804328914,
+      "learning_rate": 1.2154236916029698e-06,
+      "loss": 0.7903,
+      "num_input_tokens_seen": 112984670,
+      "step": 5318
+    },
+    {
+      "epoch": 0.6395719353093249,
+      "grad_norm": 2.9939395699251117,
+      "learning_rate": 1.2147072199968627e-06,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 113003025,
+      "step": 5319
+    },
+    {
+      "epoch": 0.6396921781999639,
+      "grad_norm": 2.50220961228838,
+      "learning_rate": 1.2139908675190454e-06,
+      "loss": 0.7146,
+      "num_input_tokens_seen": 113021955,
+      "step": 5320
+    },
+    {
+      "epoch": 0.639812421090603,
+      "grad_norm": 2.4236534102059886,
+      "learning_rate": 1.2132746342781887e-06,
+      "loss": 0.7459,
+      "num_input_tokens_seen": 113042835,
+      "step": 5321
+    },
+    {
+      "epoch": 0.6399326639812422,
+      "grad_norm": 3.673772215391414,
+      "learning_rate": 1.2125585203829437e-06,
+      "loss": 0.785,
+      "num_input_tokens_seen": 113058195,
+      "step": 5322
+    },
+    {
+      "epoch": 0.6400529068718812,
+      "grad_norm": 2.3936903601375445,
+      "learning_rate": 1.211842525941946e-06,
+      "loss": 0.7313,
+      "num_input_tokens_seen": 113077710,
+      "step": 5323
+    },
+    {
+      "epoch": 0.6401731497625203,
+      "grad_norm": 2.3263710461321,
+      "learning_rate": 1.2111266510638105e-06,
+      "loss": 0.7936,
+      "num_input_tokens_seen": 113100355,
+      "step": 5324
+    },
+    {
+      "epoch": 0.6402933926531594,
+      "grad_norm": 1.7408512655587947,
+      "learning_rate": 1.2104108958571346e-06,
+      "loss": 0.7958,
+      "num_input_tokens_seen": 113118345,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6404136355437985,
+      "grad_norm": 2.354395345675336,
+      "learning_rate": 1.2096952604304975e-06,
+      "loss": 0.751,
+      "num_input_tokens_seen": 113138495,
+      "step": 5326
+    },
+    {
+      "epoch": 0.6405338784344375,
+      "grad_norm": 2.6098000250523166,
+      "learning_rate": 1.2089797448924616e-06,
+      "loss": 0.7016,
+      "num_input_tokens_seen": 113162090,
+      "step": 5327
+    },
+    {
+      "epoch": 0.6406541213250767,
+      "grad_norm": 2.3030690169309738,
+      "learning_rate": 1.2082643493515696e-06,
+      "loss": 0.651,
+      "num_input_tokens_seen": 113180130,
+      "step": 5328
+    },
+    {
+      "epoch": 0.6407743642157158,
+      "grad_norm": 2.149240351075889,
+      "learning_rate": 1.207549073916346e-06,
+      "loss": 0.814,
+      "num_input_tokens_seen": 113200785,
+      "step": 5329
+    },
+    {
+      "epoch": 0.6408946071063548,
+      "grad_norm": 2.540532925962855,
+      "learning_rate": 1.2068339186952974e-06,
+      "loss": 0.7836,
+      "num_input_tokens_seen": 113218045,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6410148499969939,
+      "grad_norm": 2.700466189649394,
+      "learning_rate": 1.2061188837969133e-06,
+      "loss": 0.7238,
+      "num_input_tokens_seen": 113237375,
+      "step": 5331
+    },
+    {
+      "epoch": 0.641135092887633,
+      "grad_norm": 4.889911437451184,
+      "learning_rate": 1.2054039693296631e-06,
+      "loss": 0.8389,
+      "num_input_tokens_seen": 113255090,
+      "step": 5332
+    },
+    {
+      "epoch": 0.6412553357782721,
+      "grad_norm": 1.8542335704279096,
+      "learning_rate": 1.2046891754019996e-06,
+      "loss": 0.8063,
+      "num_input_tokens_seen": 113275420,
+      "step": 5333
+    },
+    {
+      "epoch": 0.6413755786689112,
+      "grad_norm": 3.3089048526070197,
+      "learning_rate": 1.2039745021223548e-06,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 113292560,
+      "step": 5334
+    },
+    {
+      "epoch": 0.6414958215595503,
+      "grad_norm": 0.912853591214701,
+      "learning_rate": 1.2032599495991456e-06,
+      "loss": 0.6196,
+      "num_input_tokens_seen": 113357020,
+      "step": 5335
+    },
+    {
+      "epoch": 0.6416160644501894,
+      "grad_norm": 2.3202516007485223,
+      "learning_rate": 1.2025455179407685e-06,
+      "loss": 0.694,
+      "num_input_tokens_seen": 113377900,
+      "step": 5336
+    },
+    {
+      "epoch": 0.6417363073408284,
+      "grad_norm": 2.407591367601122,
+      "learning_rate": 1.2018312072556027e-06,
+      "loss": 0.7331,
+      "num_input_tokens_seen": 113396120,
+      "step": 5337
+    },
+    {
+      "epoch": 0.6418565502314676,
+      "grad_norm": 2.8326555720773103,
+      "learning_rate": 1.2011170176520077e-06,
+      "loss": 0.7399,
+      "num_input_tokens_seen": 113416755,
+      "step": 5338
+    },
+    {
+      "epoch": 0.6419767931221066,
+      "grad_norm": 9.54917696404613,
+      "learning_rate": 1.200402949238326e-06,
+      "loss": 0.8065,
+      "num_input_tokens_seen": 113437815,
+      "step": 5339
+    },
+    {
+      "epoch": 0.6420970360127457,
+      "grad_norm": 2.056981520561676,
+      "learning_rate": 1.1996890021228814e-06,
+      "loss": 0.7382,
+      "num_input_tokens_seen": 113454310,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6422172789033849,
+      "grad_norm": 2.585042162478798,
+      "learning_rate": 1.198975176413979e-06,
+      "loss": 0.6928,
+      "num_input_tokens_seen": 113477680,
+      "step": 5341
+    },
+    {
+      "epoch": 0.6423375217940239,
+      "grad_norm": 1.8762190567855141,
+      "learning_rate": 1.198261472219904e-06,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 113498575,
+      "step": 5342
+    },
+    {
+      "epoch": 0.642457764684663,
+      "grad_norm": 2.175959659945122,
+      "learning_rate": 1.1975478896489276e-06,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 113516130,
+      "step": 5343
+    },
+    {
+      "epoch": 0.6425780075753021,
+      "grad_norm": 2.472418910018756,
+      "learning_rate": 1.1968344288092981e-06,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 113532430,
+      "step": 5344
+    },
+    {
+      "epoch": 0.6426982504659412,
+      "grad_norm": 2.07724398552733,
+      "learning_rate": 1.1961210898092473e-06,
+      "loss": 0.6407,
+      "num_input_tokens_seen": 113551100,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6428184933565803,
+      "grad_norm": 6.145346676051869,
+      "learning_rate": 1.1954078727569874e-06,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 113568120,
+      "step": 5346
+    },
+    {
+      "epoch": 0.6429387362472194,
+      "grad_norm": 3.6960047360292507,
+      "learning_rate": 1.1946947777607141e-06,
+      "loss": 0.777,
+      "num_input_tokens_seen": 113588975,
+      "step": 5347
+    },
+    {
+      "epoch": 0.6430589791378585,
+      "grad_norm": 1.9334417258068015,
+      "learning_rate": 1.1939818049286028e-06,
+      "loss": 0.7936,
+      "num_input_tokens_seen": 113606855,
+      "step": 5348
+    },
+    {
+      "epoch": 0.6431792220284975,
+      "grad_norm": 1.8516167601704034,
+      "learning_rate": 1.1932689543688103e-06,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 113627680,
+      "step": 5349
+    },
+    {
+      "epoch": 0.6432994649191367,
+      "grad_norm": 2.465884311197901,
+      "learning_rate": 1.1925562261894756e-06,
+      "loss": 0.7201,
+      "num_input_tokens_seen": 113646480,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6434197078097758,
+      "grad_norm": 1.8920295170593262,
+      "learning_rate": 1.1918436204987203e-06,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 113668060,
+      "step": 5351
+    },
+    {
+      "epoch": 0.6435399507004148,
+      "grad_norm": 2.3995250093597424,
+      "learning_rate": 1.191131137404645e-06,
+      "loss": 0.8122,
+      "num_input_tokens_seen": 113684520,
+      "step": 5352
+    },
+    {
+      "epoch": 0.643660193591054,
+      "grad_norm": 2.3709927214297157,
+      "learning_rate": 1.190418777015333e-06,
+      "loss": 0.7698,
+      "num_input_tokens_seen": 113703150,
+      "step": 5353
+    },
+    {
+      "epoch": 0.643780436481693,
+      "grad_norm": 3.104706877201076,
+      "learning_rate": 1.1897065394388487e-06,
+      "loss": 0.7301,
+      "num_input_tokens_seen": 113723310,
+      "step": 5354
+    },
+    {
+      "epoch": 0.6439006793723321,
+      "grad_norm": 1.933048973726292,
+      "learning_rate": 1.1889944247832385e-06,
+      "loss": 0.7603,
+      "num_input_tokens_seen": 113743270,
+      "step": 5355
+    },
+    {
+      "epoch": 0.6440209222629713,
+      "grad_norm": 2.620356832022428,
+      "learning_rate": 1.188282433156529e-06,
+      "loss": 0.7003,
+      "num_input_tokens_seen": 113762450,
+      "step": 5356
+    },
+    {
+      "epoch": 0.6441411651536103,
+      "grad_norm": 3.012886796589671,
+      "learning_rate": 1.187570564666729e-06,
+      "loss": 0.8862,
+      "num_input_tokens_seen": 113780060,
+      "step": 5357
+    },
+    {
+      "epoch": 0.6442614080442494,
+      "grad_norm": 2.7104177992956417,
+      "learning_rate": 1.1868588194218277e-06,
+      "loss": 0.747,
+      "num_input_tokens_seen": 113800160,
+      "step": 5358
+    },
+    {
+      "epoch": 0.6443816509348885,
+      "grad_norm": 1.7538525126651472,
+      "learning_rate": 1.1861471975297979e-06,
+      "loss": 0.7333,
+      "num_input_tokens_seen": 113821575,
+      "step": 5359
+    },
+    {
+      "epoch": 0.6445018938255276,
+      "grad_norm": 1.829430267584359,
+      "learning_rate": 1.185435699098591e-06,
+      "loss": 0.7015,
+      "num_input_tokens_seen": 113847490,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6446221367161666,
+      "grad_norm": 9.654782899790336,
+      "learning_rate": 1.1847243242361407e-06,
+      "loss": 0.767,
+      "num_input_tokens_seen": 113865800,
+      "step": 5361
+    },
+    {
+      "epoch": 0.6447423796068057,
+      "grad_norm": 1.8164042710850974,
+      "learning_rate": 1.184013073050362e-06,
+      "loss": 0.782,
+      "num_input_tokens_seen": 113886800,
+      "step": 5362
+    },
+    {
+      "epoch": 0.6448626224974449,
+      "grad_norm": 1.9237541469885715,
+      "learning_rate": 1.1833019456491518e-06,
+      "loss": 0.7482,
+      "num_input_tokens_seen": 113908050,
+      "step": 5363
+    },
+    {
+      "epoch": 0.6449828653880839,
+      "grad_norm": 2.2789041149074185,
+      "learning_rate": 1.1825909421403871e-06,
+      "loss": 0.7729,
+      "num_input_tokens_seen": 113926865,
+      "step": 5364
+    },
+    {
+      "epoch": 0.645103108278723,
+      "grad_norm": 2.044262204553878,
+      "learning_rate": 1.1818800626319263e-06,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 113945920,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6452233511693621,
+      "grad_norm": 2.7351556874246166,
+      "learning_rate": 1.181169307231609e-06,
+      "loss": 0.8573,
+      "num_input_tokens_seen": 113963320,
+      "step": 5366
+    },
+    {
+      "epoch": 0.6453435940600012,
+      "grad_norm": 3.6908690635817285,
+      "learning_rate": 1.1804586760472574e-06,
+      "loss": 0.8317,
+      "num_input_tokens_seen": 113979505,
+      "step": 5367
+    },
+    {
+      "epoch": 0.6454638369506402,
+      "grad_norm": 2.076550799308343,
+      "learning_rate": 1.1797481691866732e-06,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 113996450,
+      "step": 5368
+    },
+    {
+      "epoch": 0.6455840798412794,
+      "grad_norm": 2.736922152546507,
+      "learning_rate": 1.1790377867576393e-06,
+      "loss": 0.8174,
+      "num_input_tokens_seen": 114013920,
+      "step": 5369
+    },
+    {
+      "epoch": 0.6457043227319185,
+      "grad_norm": 2.1888097470898153,
+      "learning_rate": 1.1783275288679203e-06,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 114030805,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6458245656225575,
+      "grad_norm": 0.9497446653973504,
+      "learning_rate": 1.177617395625262e-06,
+      "loss": 0.6493,
+      "num_input_tokens_seen": 114088500,
+      "step": 5371
+    },
+    {
+      "epoch": 0.6459448085131967,
+      "grad_norm": 1.9630646231636049,
+      "learning_rate": 1.176907387137391e-06,
+      "loss": 0.7597,
+      "num_input_tokens_seen": 114108425,
+      "step": 5372
+    },
+    {
+      "epoch": 0.6460650514038357,
+      "grad_norm": 1.7904798573865792,
+      "learning_rate": 1.176197503512015e-06,
+      "loss": 0.8381,
+      "num_input_tokens_seen": 114127860,
+      "step": 5373
+    },
+    {
+      "epoch": 0.6461852942944748,
+      "grad_norm": 2.593414773562563,
+      "learning_rate": 1.175487744856822e-06,
+      "loss": 0.8212,
+      "num_input_tokens_seen": 114147035,
+      "step": 5374
+    },
+    {
+      "epoch": 0.646305537185114,
+      "grad_norm": 4.997265095623739,
+      "learning_rate": 1.1747781112794833e-06,
+      "loss": 0.897,
+      "num_input_tokens_seen": 114163250,
+      "step": 5375
+    },
+    {
+      "epoch": 0.646425780075753,
+      "grad_norm": 1.8633750597736216,
+      "learning_rate": 1.1740686028876492e-06,
+      "loss": 0.8165,
+      "num_input_tokens_seen": 114181835,
+      "step": 5376
+    },
+    {
+      "epoch": 0.6465460229663921,
+      "grad_norm": 2.7343963932742574,
+      "learning_rate": 1.173359219788951e-06,
+      "loss": 0.7448,
+      "num_input_tokens_seen": 114198465,
+      "step": 5377
+    },
+    {
+      "epoch": 0.6466662658570312,
+      "grad_norm": 5.201720885059953,
+      "learning_rate": 1.1726499620910014e-06,
+      "loss": 0.7143,
+      "num_input_tokens_seen": 114218465,
+      "step": 5378
+    },
+    {
+      "epoch": 0.6467865087476703,
+      "grad_norm": 2.225503499233112,
+      "learning_rate": 1.171940829901395e-06,
+      "loss": 0.759,
+      "num_input_tokens_seen": 114236910,
+      "step": 5379
+    },
+    {
+      "epoch": 0.6469067516383094,
+      "grad_norm": 22.63692718935013,
+      "learning_rate": 1.1712318233277067e-06,
+      "loss": 0.7567,
+      "num_input_tokens_seen": 114255650,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6470269945289485,
+      "grad_norm": 0.7747348307908427,
+      "learning_rate": 1.1705229424774918e-06,
+      "loss": 0.5908,
+      "num_input_tokens_seen": 114309640,
+      "step": 5381
+    },
+    {
+      "epoch": 0.6471472374195876,
+      "grad_norm": 1.7503124099654213,
+      "learning_rate": 1.1698141874582867e-06,
+      "loss": 0.6379,
+      "num_input_tokens_seen": 114330405,
+      "step": 5382
+    },
+    {
+      "epoch": 0.6472674803102266,
+      "grad_norm": 2.482338896064831,
+      "learning_rate": 1.1691055583776094e-06,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 114350215,
+      "step": 5383
+    },
+    {
+      "epoch": 0.6473877232008658,
+      "grad_norm": 2.9346558193922503,
+      "learning_rate": 1.1683970553429587e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 114371390,
+      "step": 5384
+    },
+    {
+      "epoch": 0.6475079660915048,
+      "grad_norm": 2.3210669462163787,
+      "learning_rate": 1.1676886784618128e-06,
+      "loss": 0.8106,
+      "num_input_tokens_seen": 114387775,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6476282089821439,
+      "grad_norm": 2.3177982054961106,
+      "learning_rate": 1.1669804278416332e-06,
+      "loss": 0.8234,
+      "num_input_tokens_seen": 114402220,
+      "step": 5386
+    },
+    {
+      "epoch": 0.6477484518727831,
+      "grad_norm": 1.9091213905273532,
+      "learning_rate": 1.1662723035898602e-06,
+      "loss": 0.709,
+      "num_input_tokens_seen": 114421700,
+      "step": 5387
+    },
+    {
+      "epoch": 0.6478686947634221,
+      "grad_norm": 2.026477861519642,
+      "learning_rate": 1.1655643058139158e-06,
+      "loss": 0.8125,
+      "num_input_tokens_seen": 114440420,
+      "step": 5388
+    },
+    {
+      "epoch": 0.6479889376540612,
+      "grad_norm": 1.737474323674123,
+      "learning_rate": 1.164856434621203e-06,
+      "loss": 0.8099,
+      "num_input_tokens_seen": 114459260,
+      "step": 5389
+    },
+    {
+      "epoch": 0.6481091805447003,
+      "grad_norm": 2.137194915622085,
+      "learning_rate": 1.164148690119104e-06,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 114480260,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6482294234353394,
+      "grad_norm": 2.113936030754424,
+      "learning_rate": 1.163441072414985e-06,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 114500185,
+      "step": 5391
+    },
+    {
+      "epoch": 0.6483496663259785,
+      "grad_norm": 2.181219042106376,
+      "learning_rate": 1.16273358161619e-06,
+      "loss": 0.6891,
+      "num_input_tokens_seen": 114520240,
+      "step": 5392
+    },
+    {
+      "epoch": 0.6484699092166175,
+      "grad_norm": 2.0493167137015176,
+      "learning_rate": 1.1620262178300455e-06,
+      "loss": 0.8415,
+      "num_input_tokens_seen": 114538575,
+      "step": 5393
+    },
+    {
+      "epoch": 0.6485901521072567,
+      "grad_norm": 1.8379586571441586,
+      "learning_rate": 1.1613189811638563e-06,
+      "loss": 0.7531,
+      "num_input_tokens_seen": 114560020,
+      "step": 5394
+    },
+    {
+      "epoch": 0.6487103949978957,
+      "grad_norm": 3.8303575091283943,
+      "learning_rate": 1.1606118717249117e-06,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 114579840,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6488306378885348,
+      "grad_norm": 2.138371651536261,
+      "learning_rate": 1.1599048896204787e-06,
+      "loss": 0.679,
+      "num_input_tokens_seen": 114599440,
+      "step": 5396
+    },
+    {
+      "epoch": 0.648950880779174,
+      "grad_norm": 1.834884138733889,
+      "learning_rate": 1.1591980349578061e-06,
+      "loss": 0.803,
+      "num_input_tokens_seen": 114617830,
+      "step": 5397
+    },
+    {
+      "epoch": 0.649071123669813,
+      "grad_norm": 0.8080736411275956,
+      "learning_rate": 1.1584913078441222e-06,
+      "loss": 0.5745,
+      "num_input_tokens_seen": 114677470,
+      "step": 5398
+    },
+    {
+      "epoch": 0.6491913665604521,
+      "grad_norm": 2.7017641710080182,
+      "learning_rate": 1.1577847083866387e-06,
+      "loss": 0.8382,
+      "num_input_tokens_seen": 114696225,
+      "step": 5399
+    },
+    {
+      "epoch": 0.6493116094510912,
+      "grad_norm": 2.833949755276881,
+      "learning_rate": 1.1570782366925453e-06,
+      "loss": 0.725,
+      "num_input_tokens_seen": 114714460,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6494318523417303,
+      "grad_norm": 6.20291758446645,
+      "learning_rate": 1.1563718928690132e-06,
+      "loss": 0.7527,
+      "num_input_tokens_seen": 114731615,
+      "step": 5401
+    },
+    {
+      "epoch": 0.6495520952323693,
+      "grad_norm": 2.5196673123363316,
+      "learning_rate": 1.1556656770231942e-06,
+      "loss": 0.7039,
+      "num_input_tokens_seen": 114747530,
+      "step": 5402
+    },
+    {
+      "epoch": 0.6496723381230085,
+      "grad_norm": 2.299612062614275,
+      "learning_rate": 1.1549595892622207e-06,
+      "loss": 0.7569,
+      "num_input_tokens_seen": 114766020,
+      "step": 5403
+    },
+    {
+      "epoch": 0.6497925810136476,
+      "grad_norm": 0.8593892284736993,
+      "learning_rate": 1.1542536296932053e-06,
+      "loss": 0.6337,
+      "num_input_tokens_seen": 114829275,
+      "step": 5404
+    },
+    {
+      "epoch": 0.6499128239042866,
+      "grad_norm": 2.1367610561197155,
+      "learning_rate": 1.1535477984232423e-06,
+      "loss": 0.6917,
+      "num_input_tokens_seen": 114848870,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6500330667949258,
+      "grad_norm": 2.240752533806768,
+      "learning_rate": 1.152842095559404e-06,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 114869250,
+      "step": 5406
+    },
+    {
+      "epoch": 0.6501533096855648,
+      "grad_norm": 1.857743792423614,
+      "learning_rate": 1.1521365212087474e-06,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 114888955,
+      "step": 5407
+    },
+    {
+      "epoch": 0.6502735525762039,
+      "grad_norm": 1.8777365174935885,
+      "learning_rate": 1.1514310754783062e-06,
+      "loss": 0.7005,
+      "num_input_tokens_seen": 114911625,
+      "step": 5408
+    },
+    {
+      "epoch": 0.6503937954668431,
+      "grad_norm": 1.8850690756719408,
+      "learning_rate": 1.1507257584750964e-06,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 114931525,
+      "step": 5409
+    },
+    {
+      "epoch": 0.6505140383574821,
+      "grad_norm": 1.9504362481009716,
+      "learning_rate": 1.150020570306113e-06,
+      "loss": 0.7632,
+      "num_input_tokens_seen": 114950385,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6506342812481212,
+      "grad_norm": 2.786111719453489,
+      "learning_rate": 1.1493155110783338e-06,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 114968630,
+      "step": 5411
+    },
+    {
+      "epoch": 0.6507545241387603,
+      "grad_norm": 2.743658247547822,
+      "learning_rate": 1.1486105808987155e-06,
+      "loss": 0.7047,
+      "num_input_tokens_seen": 114989840,
+      "step": 5412
+    },
+    {
+      "epoch": 0.6508747670293994,
+      "grad_norm": 1.9627647652122042,
+      "learning_rate": 1.1479057798741947e-06,
+      "loss": 0.803,
+      "num_input_tokens_seen": 115007615,
+      "step": 5413
+    },
+    {
+      "epoch": 0.6509950099200384,
+      "grad_norm": 0.8519248943625141,
+      "learning_rate": 1.1472011081116893e-06,
+      "loss": 0.5777,
+      "num_input_tokens_seen": 115064565,
+      "step": 5414
+    },
+    {
+      "epoch": 0.6511152528106776,
+      "grad_norm": 2.5983168388011895,
+      "learning_rate": 1.146496565718098e-06,
+      "loss": 0.7624,
+      "num_input_tokens_seen": 115084855,
+      "step": 5415
+    },
+    {
+      "epoch": 0.6512354957013167,
+      "grad_norm": 4.170478578604415,
+      "learning_rate": 1.1457921528002996e-06,
+      "loss": 0.7485,
+      "num_input_tokens_seen": 115103010,
+      "step": 5416
+    },
+    {
+      "epoch": 0.6513557385919557,
+      "grad_norm": 3.1160852940465325,
+      "learning_rate": 1.145087869465153e-06,
+      "loss": 0.7185,
+      "num_input_tokens_seen": 115123295,
+      "step": 5417
+    },
+    {
+      "epoch": 0.6514759814825949,
+      "grad_norm": 5.550907757235389,
+      "learning_rate": 1.1443837158194954e-06,
+      "loss": 0.6221,
+      "num_input_tokens_seen": 115138160,
+      "step": 5418
+    },
+    {
+      "epoch": 0.651596224373234,
+      "grad_norm": 2.3418374344765738,
+      "learning_rate": 1.1436796919701484e-06,
+      "loss": 0.7352,
+      "num_input_tokens_seen": 115156595,
+      "step": 5419
+    },
+    {
+      "epoch": 0.651716467263873,
+      "grad_norm": 2.1147793267346837,
+      "learning_rate": 1.1429757980239115e-06,
+      "loss": 0.6148,
+      "num_input_tokens_seen": 115176740,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6518367101545122,
+      "grad_norm": 2.750431851853045,
+      "learning_rate": 1.1422720340875644e-06,
+      "loss": 0.8059,
+      "num_input_tokens_seen": 115195210,
+      "step": 5421
+    },
+    {
+      "epoch": 0.6519569530451512,
+      "grad_norm": 2.7279162412048805,
+      "learning_rate": 1.1415684002678671e-06,
+      "loss": 0.7871,
+      "num_input_tokens_seen": 115213690,
+      "step": 5422
+    },
+    {
+      "epoch": 0.6520771959357903,
+      "grad_norm": 3.6482507766970644,
+      "learning_rate": 1.1408648966715617e-06,
+      "loss": 0.7803,
+      "num_input_tokens_seen": 115230930,
+      "step": 5423
+    },
+    {
+      "epoch": 0.6521974388264293,
+      "grad_norm": 3.831495756765936,
+      "learning_rate": 1.1401615234053683e-06,
+      "loss": 0.7197,
+      "num_input_tokens_seen": 115249470,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6523176817170685,
+      "grad_norm": 1.842908548710883,
+      "learning_rate": 1.1394582805759885e-06,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 115268470,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6524379246077076,
+      "grad_norm": 1.8582930045716146,
+      "learning_rate": 1.1387551682901022e-06,
+      "loss": 0.7572,
+      "num_input_tokens_seen": 115288795,
+      "step": 5426
+    },
+    {
+      "epoch": 0.6525581674983466,
+      "grad_norm": 2.7718112736944365,
+      "learning_rate": 1.138052186654373e-06,
+      "loss": 0.7008,
+      "num_input_tokens_seen": 115305985,
+      "step": 5427
+    },
+    {
+      "epoch": 0.6526784103889858,
+      "grad_norm": 2.951645136073294,
+      "learning_rate": 1.1373493357754417e-06,
+      "loss": 0.8755,
+      "num_input_tokens_seen": 115324610,
+      "step": 5428
+    },
+    {
+      "epoch": 0.6527986532796248,
+      "grad_norm": 2.0342297387896298,
+      "learning_rate": 1.1366466157599303e-06,
+      "loss": 0.7677,
+      "num_input_tokens_seen": 115343605,
+      "step": 5429
+    },
+    {
+      "epoch": 0.6529188961702639,
+      "grad_norm": 3.0844687272804356,
+      "learning_rate": 1.1359440267144412e-06,
+      "loss": 0.7539,
+      "num_input_tokens_seen": 115360780,
+      "step": 5430
+    },
+    {
+      "epoch": 0.653039139060903,
+      "grad_norm": 1.8415268492795913,
+      "learning_rate": 1.1352415687455556e-06,
+      "loss": 0.7384,
+      "num_input_tokens_seen": 115381760,
+      "step": 5431
+    },
+    {
+      "epoch": 0.6531593819515421,
+      "grad_norm": 3.175285154046051,
+      "learning_rate": 1.1345392419598368e-06,
+      "loss": 0.6365,
+      "num_input_tokens_seen": 115400360,
+      "step": 5432
+    },
+    {
+      "epoch": 0.6532796248421812,
+      "grad_norm": 1.9449097702459905,
+      "learning_rate": 1.133837046463827e-06,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 115419480,
+      "step": 5433
+    },
+    {
+      "epoch": 0.6533998677328203,
+      "grad_norm": 2.860088198945128,
+      "learning_rate": 1.1331349823640474e-06,
+      "loss": 0.6358,
+      "num_input_tokens_seen": 115436630,
+      "step": 5434
+    },
+    {
+      "epoch": 0.6535201106234594,
+      "grad_norm": 2.6007283199817963,
+      "learning_rate": 1.132433049767003e-06,
+      "loss": 0.7828,
+      "num_input_tokens_seen": 115454265,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6536403535140984,
+      "grad_norm": 1.628706954462291,
+      "learning_rate": 1.1317312487791748e-06,
+      "loss": 0.7998,
+      "num_input_tokens_seen": 115475635,
+      "step": 5436
+    },
+    {
+      "epoch": 0.6537605964047376,
+      "grad_norm": 2.8244763712729517,
+      "learning_rate": 1.131029579507026e-06,
+      "loss": 0.7192,
+      "num_input_tokens_seen": 115495295,
+      "step": 5437
+    },
+    {
+      "epoch": 0.6538808392953767,
+      "grad_norm": 1.974859420721518,
+      "learning_rate": 1.1303280420569982e-06,
+      "loss": 0.7994,
+      "num_input_tokens_seen": 115516900,
+      "step": 5438
+    },
+    {
+      "epoch": 0.6540010821860157,
+      "grad_norm": 2.1207385191914523,
+      "learning_rate": 1.1296266365355158e-06,
+      "loss": 0.7622,
+      "num_input_tokens_seen": 115540005,
+      "step": 5439
+    },
+    {
+      "epoch": 0.6541213250766549,
+      "grad_norm": 2.2726856058740488,
+      "learning_rate": 1.1289253630489806e-06,
+      "loss": 0.7257,
+      "num_input_tokens_seen": 115560775,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6542415679672939,
+      "grad_norm": 2.4532842367711343,
+      "learning_rate": 1.1282242217037753e-06,
+      "loss": 0.7382,
+      "num_input_tokens_seen": 115577995,
+      "step": 5441
+    },
+    {
+      "epoch": 0.654361810857933,
+      "grad_norm": 2.756443333504138,
+      "learning_rate": 1.1275232126062614e-06,
+      "loss": 0.6213,
+      "num_input_tokens_seen": 115600540,
+      "step": 5442
+    },
+    {
+      "epoch": 0.6544820537485722,
+      "grad_norm": 1.9539630289725078,
+      "learning_rate": 1.1268223358627835e-06,
+      "loss": 0.7241,
+      "num_input_tokens_seen": 115622750,
+      "step": 5443
+    },
+    {
+      "epoch": 0.6546022966392112,
+      "grad_norm": 2.1462075815416015,
+      "learning_rate": 1.126121591579663e-06,
+      "loss": 0.7056,
+      "num_input_tokens_seen": 115641675,
+      "step": 5444
+    },
+    {
+      "epoch": 0.6547225395298503,
+      "grad_norm": 2.7809840030810147,
+      "learning_rate": 1.1254209798632018e-06,
+      "loss": 0.6862,
+      "num_input_tokens_seen": 115662415,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6548427824204894,
+      "grad_norm": 1.8071513463625968,
+      "learning_rate": 1.124720500819683e-06,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 115680290,
+      "step": 5446
+    },
+    {
+      "epoch": 0.6549630253111285,
+      "grad_norm": 2.4942841805542235,
+      "learning_rate": 1.1240201545553682e-06,
+      "loss": 0.8185,
+      "num_input_tokens_seen": 115697810,
+      "step": 5447
+    },
+    {
+      "epoch": 0.6550832682017675,
+      "grad_norm": 7.42106113027378,
+      "learning_rate": 1.1233199411764996e-06,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 115716965,
+      "step": 5448
+    },
+    {
+      "epoch": 0.6552035110924067,
+      "grad_norm": 2.9914338132911285,
+      "learning_rate": 1.1226198607892987e-06,
+      "loss": 0.6882,
+      "num_input_tokens_seen": 115737245,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6553237539830458,
+      "grad_norm": 1.9920378915034498,
+      "learning_rate": 1.1219199134999664e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 115755465,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6554439968736848,
+      "grad_norm": 2.1506121379245746,
+      "learning_rate": 1.1212200994146863e-06,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 115772940,
+      "step": 5451
+    },
+    {
+      "epoch": 0.655564239764324,
+      "grad_norm": 2.6884195725406212,
+      "learning_rate": 1.120520418639618e-06,
+      "loss": 0.7485,
+      "num_input_tokens_seen": 115791195,
+      "step": 5452
+    },
+    {
+      "epoch": 0.655684482654963,
+      "grad_norm": 2.026326990547198,
+      "learning_rate": 1.119820871280903e-06,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 115811990,
+      "step": 5453
+    },
+    {
+      "epoch": 0.6558047255456021,
+      "grad_norm": 6.090404650070804,
+      "learning_rate": 1.1191214574446614e-06,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 115831955,
+      "step": 5454
+    },
+    {
+      "epoch": 0.6559249684362413,
+      "grad_norm": 1.6746053391414106,
+      "learning_rate": 1.118422177236995e-06,
+      "loss": 0.7937,
+      "num_input_tokens_seen": 115853500,
+      "step": 5455
+    },
+    {
+      "epoch": 0.6560452113268803,
+      "grad_norm": 2.4658539440548206,
+      "learning_rate": 1.1177230307639835e-06,
+      "loss": 0.8467,
+      "num_input_tokens_seen": 115870760,
+      "step": 5456
+    },
+    {
+      "epoch": 0.6561654542175194,
+      "grad_norm": 1.993763917924796,
+      "learning_rate": 1.1170240181316865e-06,
+      "loss": 0.7862,
+      "num_input_tokens_seen": 115891925,
+      "step": 5457
+    },
+    {
+      "epoch": 0.6562856971081584,
+      "grad_norm": 2.5433038801142924,
+      "learning_rate": 1.1163251394461433e-06,
+      "loss": 0.795,
+      "num_input_tokens_seen": 115910125,
+      "step": 5458
+    },
+    {
+      "epoch": 0.6564059399987976,
+      "grad_norm": 2.931023486959919,
+      "learning_rate": 1.1156263948133746e-06,
+      "loss": 0.8196,
+      "num_input_tokens_seen": 115926500,
+      "step": 5459
+    },
+    {
+      "epoch": 0.6565261828894366,
+      "grad_norm": 1.9998085982412905,
+      "learning_rate": 1.1149277843393793e-06,
+      "loss": 0.7746,
+      "num_input_tokens_seen": 115947380,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6566464257800757,
+      "grad_norm": 2.454902955488945,
+      "learning_rate": 1.114229308130135e-06,
+      "loss": 0.6329,
+      "num_input_tokens_seen": 115964980,
+      "step": 5461
+    },
+    {
+      "epoch": 0.6567666686707149,
+      "grad_norm": 2.0762901053238747,
+      "learning_rate": 1.1135309662915995e-06,
+      "loss": 0.6685,
+      "num_input_tokens_seen": 115984865,
+      "step": 5462
+    },
+    {
+      "epoch": 0.6568869115613539,
+      "grad_norm": 2.792002177719222,
+      "learning_rate": 1.112832758929712e-06,
+      "loss": 0.5957,
+      "num_input_tokens_seen": 116007195,
+      "step": 5463
+    },
+    {
+      "epoch": 0.657007154451993,
+      "grad_norm": 2.917205676280278,
+      "learning_rate": 1.11213468615039e-06,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 116026345,
+      "step": 5464
+    },
+    {
+      "epoch": 0.6571273973426321,
+      "grad_norm": 7.936242447778208,
+      "learning_rate": 1.1114367480595299e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 116047145,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6572476402332712,
+      "grad_norm": 2.71814201176872,
+      "learning_rate": 1.1107389447630077e-06,
+      "loss": 0.8098,
+      "num_input_tokens_seen": 116065565,
+      "step": 5466
+    },
+    {
+      "epoch": 0.6573678831239103,
+      "grad_norm": 2.2333780829947396,
+      "learning_rate": 1.1100412763666818e-06,
+      "loss": 0.7749,
+      "num_input_tokens_seen": 116080545,
+      "step": 5467
+    },
+    {
+      "epoch": 0.6574881260145494,
+      "grad_norm": 3.4091599597747435,
+      "learning_rate": 1.1093437429763865e-06,
+      "loss": 0.7922,
+      "num_input_tokens_seen": 116100530,
+      "step": 5468
+    },
+    {
+      "epoch": 0.6576083689051885,
+      "grad_norm": 2.651966925931211,
+      "learning_rate": 1.108646344697937e-06,
+      "loss": 0.7401,
+      "num_input_tokens_seen": 116118600,
+      "step": 5469
+    },
+    {
+      "epoch": 0.6577286117958275,
+      "grad_norm": 2.4425894714104186,
+      "learning_rate": 1.1079490816371277e-06,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 116138085,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6578488546864667,
+      "grad_norm": 3.157342247876814,
+      "learning_rate": 1.1072519538997346e-06,
+      "loss": 0.7314,
+      "num_input_tokens_seen": 116156945,
+      "step": 5471
+    },
+    {
+      "epoch": 0.6579690975771058,
+      "grad_norm": 1.6831217598836876,
+      "learning_rate": 1.1065549615915095e-06,
+      "loss": 0.8115,
+      "num_input_tokens_seen": 116176495,
+      "step": 5472
+    },
+    {
+      "epoch": 0.6580893404677448,
+      "grad_norm": 2.9241770060946326,
+      "learning_rate": 1.105858104818187e-06,
+      "loss": 0.7717,
+      "num_input_tokens_seen": 116197370,
+      "step": 5473
+    },
+    {
+      "epoch": 0.658209583358384,
+      "grad_norm": 3.39827239466364,
+      "learning_rate": 1.1051613836854788e-06,
+      "loss": 0.7431,
+      "num_input_tokens_seen": 116213475,
+      "step": 5474
+    },
+    {
+      "epoch": 0.658329826249023,
+      "grad_norm": 0.7825908881542003,
+      "learning_rate": 1.1044647982990771e-06,
+      "loss": 0.5992,
+      "num_input_tokens_seen": 116275080,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6584500691396621,
+      "grad_norm": 2.688796430741427,
+      "learning_rate": 1.1037683487646536e-06,
+      "loss": 0.6326,
+      "num_input_tokens_seen": 116295085,
+      "step": 5476
+    },
+    {
+      "epoch": 0.6585703120303013,
+      "grad_norm": 1.9504613669587676,
+      "learning_rate": 1.1030720351878594e-06,
+      "loss": 0.7665,
+      "num_input_tokens_seen": 116312925,
+      "step": 5477
+    },
+    {
+      "epoch": 0.6586905549209403,
+      "grad_norm": 0.8279952753306795,
+      "learning_rate": 1.102375857674323e-06,
+      "loss": 0.6164,
+      "num_input_tokens_seen": 116374560,
+      "step": 5478
+    },
+    {
+      "epoch": 0.6588107978115794,
+      "grad_norm": 2.096631672149992,
+      "learning_rate": 1.1016798163296561e-06,
+      "loss": 0.8975,
+      "num_input_tokens_seen": 116393480,
+      "step": 5479
+    },
+    {
+      "epoch": 0.6589310407022185,
+      "grad_norm": 2.2720620266549147,
+      "learning_rate": 1.1009839112594471e-06,
+      "loss": 0.6553,
+      "num_input_tokens_seen": 116411225,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6590512835928576,
+      "grad_norm": 2.874935116488654,
+      "learning_rate": 1.1002881425692638e-06,
+      "loss": 0.7156,
+      "num_input_tokens_seen": 116431375,
+      "step": 5481
+    },
+    {
+      "epoch": 0.6591715264834966,
+      "grad_norm": 1.8068882832227608,
+      "learning_rate": 1.0995925103646532e-06,
+      "loss": 0.7487,
+      "num_input_tokens_seen": 116449695,
+      "step": 5482
+    },
+    {
+      "epoch": 0.6592917693741358,
+      "grad_norm": 2.7713059094046475,
+      "learning_rate": 1.0988970147511437e-06,
+      "loss": 0.6666,
+      "num_input_tokens_seen": 116471295,
+      "step": 5483
+    },
+    {
+      "epoch": 0.6594120122647749,
+      "grad_norm": 2.6744358497419354,
+      "learning_rate": 1.0982016558342405e-06,
+      "loss": 0.8009,
+      "num_input_tokens_seen": 116489985,
+      "step": 5484
+    },
+    {
+      "epoch": 0.6595322551554139,
+      "grad_norm": 2.321274491254443,
+      "learning_rate": 1.0975064337194291e-06,
+      "loss": 0.7104,
+      "num_input_tokens_seen": 116507750,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6596524980460531,
+      "grad_norm": 2.062603096622019,
+      "learning_rate": 1.0968113485121734e-06,
+      "loss": 0.6973,
+      "num_input_tokens_seen": 116527060,
+      "step": 5486
+    },
+    {
+      "epoch": 0.6597727409366921,
+      "grad_norm": 2.0816052496221196,
+      "learning_rate": 1.0961164003179185e-06,
+      "loss": 0.797,
+      "num_input_tokens_seen": 116545290,
+      "step": 5487
+    },
+    {
+      "epoch": 0.6598929838273312,
+      "grad_norm": 2.302974991219862,
+      "learning_rate": 1.0954215892420875e-06,
+      "loss": 0.8307,
+      "num_input_tokens_seen": 116565710,
+      "step": 5488
+    },
+    {
+      "epoch": 0.6600132267179702,
+      "grad_norm": 2.353437844330326,
+      "learning_rate": 1.094726915390082e-06,
+      "loss": 0.7012,
+      "num_input_tokens_seen": 116583765,
+      "step": 5489
+    },
+    {
+      "epoch": 0.6601334696086094,
+      "grad_norm": 1.8788638675061986,
+      "learning_rate": 1.0940323788672836e-06,
+      "loss": 0.694,
+      "num_input_tokens_seen": 116602660,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6602537124992485,
+      "grad_norm": 2.212091276382403,
+      "learning_rate": 1.093337979779053e-06,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 116621795,
+      "step": 5491
+    },
+    {
+      "epoch": 0.6603739553898875,
+      "grad_norm": 2.5279204872176817,
+      "learning_rate": 1.0926437182307302e-06,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 116640325,
+      "step": 5492
+    },
+    {
+      "epoch": 0.6604941982805267,
+      "grad_norm": 2.161517934802307,
+      "learning_rate": 1.0919495943276338e-06,
+      "loss": 0.778,
+      "num_input_tokens_seen": 116661065,
+      "step": 5493
+    },
+    {
+      "epoch": 0.6606144411711657,
+      "grad_norm": 3.0094488106314845,
+      "learning_rate": 1.0912556081750611e-06,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 116678715,
+      "step": 5494
+    },
+    {
+      "epoch": 0.6607346840618048,
+      "grad_norm": 2.008382762873296,
+      "learning_rate": 1.0905617598782909e-06,
+      "loss": 0.7611,
+      "num_input_tokens_seen": 116698640,
+      "step": 5495
+    },
+    {
+      "epoch": 0.660854926952444,
+      "grad_norm": 2.7774383930739583,
+      "learning_rate": 1.0898680495425786e-06,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 116716650,
+      "step": 5496
+    },
+    {
+      "epoch": 0.660975169843083,
+      "grad_norm": 1.879071591427866,
+      "learning_rate": 1.0891744772731594e-06,
+      "loss": 0.7989,
+      "num_input_tokens_seen": 116734185,
+      "step": 5497
+    },
+    {
+      "epoch": 0.6610954127337221,
+      "grad_norm": 2.1500111413072323,
+      "learning_rate": 1.0884810431752473e-06,
+      "loss": 0.6432,
+      "num_input_tokens_seen": 116754475,
+      "step": 5498
+    },
+    {
+      "epoch": 0.6612156556243612,
+      "grad_norm": 2.0618194358256523,
+      "learning_rate": 1.0877877473540368e-06,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 116774780,
+      "step": 5499
+    },
+    {
+      "epoch": 0.6613358985150003,
+      "grad_norm": 1.8766934680817506,
+      "learning_rate": 1.0870945899147002e-06,
+      "loss": 0.7217,
+      "num_input_tokens_seen": 116791145,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6614561414056394,
+      "grad_norm": 1.8937098029064863,
+      "learning_rate": 1.0864015709623879e-06,
+      "loss": 0.7456,
+      "num_input_tokens_seen": 116811735,
+      "step": 5501
+    },
+    {
+      "epoch": 0.6615763842962785,
+      "grad_norm": 2.669832105247532,
+      "learning_rate": 1.0857086906022303e-06,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 116829790,
+      "step": 5502
+    },
+    {
+      "epoch": 0.6616966271869176,
+      "grad_norm": 2.367858427789993,
+      "learning_rate": 1.0850159489393388e-06,
+      "loss": 0.7238,
+      "num_input_tokens_seen": 116848770,
+      "step": 5503
+    },
+    {
+      "epoch": 0.6618168700775566,
+      "grad_norm": 2.0560595715073786,
+      "learning_rate": 1.0843233460788e-06,
+      "loss": 0.815,
+      "num_input_tokens_seen": 116865705,
+      "step": 5504
+    },
+    {
+      "epoch": 0.6619371129681958,
+      "grad_norm": 1.9499404117032744,
+      "learning_rate": 1.0836308821256812e-06,
+      "loss": 0.7744,
+      "num_input_tokens_seen": 116886225,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6620573558588349,
+      "grad_norm": 2.0850437278002825,
+      "learning_rate": 1.0829385571850282e-06,
+      "loss": 0.7796,
+      "num_input_tokens_seen": 116902925,
+      "step": 5506
+    },
+    {
+      "epoch": 0.6621775987494739,
+      "grad_norm": 2.945327206187434,
+      "learning_rate": 1.0822463713618679e-06,
+      "loss": 0.8314,
+      "num_input_tokens_seen": 116919500,
+      "step": 5507
+    },
+    {
+      "epoch": 0.6622978416401131,
+      "grad_norm": 4.135214924993876,
+      "learning_rate": 1.0815543247612034e-06,
+      "loss": 0.8365,
+      "num_input_tokens_seen": 116936290,
+      "step": 5508
+    },
+    {
+      "epoch": 0.6624180845307521,
+      "grad_norm": 2.57782346587537,
+      "learning_rate": 1.0808624174880174e-06,
+      "loss": 0.8222,
+      "num_input_tokens_seen": 116956660,
+      "step": 5509
+    },
+    {
+      "epoch": 0.6625383274213912,
+      "grad_norm": 2.5597878091921125,
+      "learning_rate": 1.0801706496472714e-06,
+      "loss": 0.7892,
+      "num_input_tokens_seen": 116976185,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6626585703120303,
+      "grad_norm": 1.7783624732037169,
+      "learning_rate": 1.0794790213439068e-06,
+      "loss": 0.6667,
+      "num_input_tokens_seen": 117002805,
+      "step": 5511
+    },
+    {
+      "epoch": 0.6627788132026694,
+      "grad_norm": 2.2096679364253484,
+      "learning_rate": 1.078787532682843e-06,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 117020000,
+      "step": 5512
+    },
+    {
+      "epoch": 0.6628990560933085,
+      "grad_norm": 2.9892144218478927,
+      "learning_rate": 1.0780961837689781e-06,
+      "loss": 0.7551,
+      "num_input_tokens_seen": 117039230,
+      "step": 5513
+    },
+    {
+      "epoch": 0.6630192989839476,
+      "grad_norm": 1.6151765442348993,
+      "learning_rate": 1.0774049747071883e-06,
+      "loss": 0.6941,
+      "num_input_tokens_seen": 117056830,
+      "step": 5514
+    },
+    {
+      "epoch": 0.6631395418745867,
+      "grad_norm": 2.484489142523293,
+      "learning_rate": 1.0767139056023312e-06,
+      "loss": 0.6822,
+      "num_input_tokens_seen": 117077125,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6632597847652257,
+      "grad_norm": 1.9147662923813924,
+      "learning_rate": 1.07602297655924e-06,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 117095165,
+      "step": 5516
+    },
+    {
+      "epoch": 0.6633800276558649,
+      "grad_norm": 2.1325139450852117,
+      "learning_rate": 1.0753321876827292e-06,
+      "loss": 0.7988,
+      "num_input_tokens_seen": 117114170,
+      "step": 5517
+    },
+    {
+      "epoch": 0.663500270546504,
+      "grad_norm": 2.5324050566788823,
+      "learning_rate": 1.0746415390775902e-06,
+      "loss": 0.7323,
+      "num_input_tokens_seen": 117132020,
+      "step": 5518
+    },
+    {
+      "epoch": 0.663620513437143,
+      "grad_norm": 2.0002547362609895,
+      "learning_rate": 1.0739510308485939e-06,
+      "loss": 0.7746,
+      "num_input_tokens_seen": 117148955,
+      "step": 5519
+    },
+    {
+      "epoch": 0.6637407563277821,
+      "grad_norm": 0.8236460222148144,
+      "learning_rate": 1.07326066310049e-06,
+      "loss": 0.6436,
+      "num_input_tokens_seen": 117212800,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6638609992184212,
+      "grad_norm": 2.749800989513713,
+      "learning_rate": 1.0725704359380065e-06,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 117232375,
+      "step": 5521
+    },
+    {
+      "epoch": 0.6639812421090603,
+      "grad_norm": 2.3778164296012148,
+      "learning_rate": 1.0718803494658497e-06,
+      "loss": 0.7137,
+      "num_input_tokens_seen": 117250985,
+      "step": 5522
+    },
+    {
+      "epoch": 0.6641014849996993,
+      "grad_norm": 2.2148159103921485,
+      "learning_rate": 1.071190403788707e-06,
+      "loss": 0.8342,
+      "num_input_tokens_seen": 117266010,
+      "step": 5523
+    },
+    {
+      "epoch": 0.6642217278903385,
+      "grad_norm": 3.858551778709007,
+      "learning_rate": 1.0705005990112415e-06,
+      "loss": 0.7454,
+      "num_input_tokens_seen": 117285510,
+      "step": 5524
+    },
+    {
+      "epoch": 0.6643419707809776,
+      "grad_norm": 3.325488403273317,
+      "learning_rate": 1.0698109352380957e-06,
+      "loss": 0.7419,
+      "num_input_tokens_seen": 117302830,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6644622136716166,
+      "grad_norm": 2.587646484770462,
+      "learning_rate": 1.0691214125738909e-06,
+      "loss": 0.7732,
+      "num_input_tokens_seen": 117322755,
+      "step": 5526
+    },
+    {
+      "epoch": 0.6645824565622558,
+      "grad_norm": 2.2338633944812245,
+      "learning_rate": 1.0684320311232287e-06,
+      "loss": 0.6143,
+      "num_input_tokens_seen": 117385380,
+      "step": 5527
+    },
+    {
+      "epoch": 0.6647026994528948,
+      "grad_norm": 1.9672865995361937,
+      "learning_rate": 1.0677427909906865e-06,
+      "loss": 0.8057,
+      "num_input_tokens_seen": 117405550,
+      "step": 5528
+    },
+    {
+      "epoch": 0.6648229423435339,
+      "grad_norm": 2.807091670308767,
+      "learning_rate": 1.0670536922808216e-06,
+      "loss": 0.7129,
+      "num_input_tokens_seen": 117425395,
+      "step": 5529
+    },
+    {
+      "epoch": 0.6649431852341731,
+      "grad_norm": 2.5961994790189977,
+      "learning_rate": 1.066364735098169e-06,
+      "loss": 0.7127,
+      "num_input_tokens_seen": 117441495,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6650634281248121,
+      "grad_norm": 2.32410710757479,
+      "learning_rate": 1.0656759195472447e-06,
+      "loss": 0.8046,
+      "num_input_tokens_seen": 117458505,
+      "step": 5531
+    },
+    {
+      "epoch": 0.6651836710154512,
+      "grad_norm": 0.8313125604807928,
+      "learning_rate": 1.0649872457325403e-06,
+      "loss": 0.6416,
+      "num_input_tokens_seen": 117519510,
+      "step": 5532
+    },
+    {
+      "epoch": 0.6653039139060903,
+      "grad_norm": 0.9143975828236155,
+      "learning_rate": 1.0642987137585278e-06,
+      "loss": 0.6133,
+      "num_input_tokens_seen": 117578755,
+      "step": 5533
+    },
+    {
+      "epoch": 0.6654241567967294,
+      "grad_norm": 1.822386305081586,
+      "learning_rate": 1.0636103237296561e-06,
+      "loss": 0.8164,
+      "num_input_tokens_seen": 117597400,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6655443996873684,
+      "grad_norm": 2.059920949425601,
+      "learning_rate": 1.0629220757503538e-06,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 117617135,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6656646425780076,
+      "grad_norm": 2.7126568025008493,
+      "learning_rate": 1.0622339699250274e-06,
+      "loss": 0.7142,
+      "num_input_tokens_seen": 117634775,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6657848854686467,
+      "grad_norm": 1.8709695712420376,
+      "learning_rate": 1.0615460063580624e-06,
+      "loss": 0.7928,
+      "num_input_tokens_seen": 117652970,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6659051283592857,
+      "grad_norm": 2.4054106377231697,
+      "learning_rate": 1.060858185153821e-06,
+      "loss": 0.7311,
+      "num_input_tokens_seen": 117670790,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6660253712499249,
+      "grad_norm": 2.6512031915456378,
+      "learning_rate": 1.0601705064166474e-06,
+      "loss": 0.7622,
+      "num_input_tokens_seen": 117688905,
+      "step": 5539
+    },
+    {
+      "epoch": 0.666145614140564,
+      "grad_norm": 2.304515260912785,
+      "learning_rate": 1.0594829702508605e-06,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 117706340,
+      "step": 5540
+    },
+    {
+      "epoch": 0.666265857031203,
+      "grad_norm": 1.977406769513987,
+      "learning_rate": 1.0587955767607592e-06,
+      "loss": 0.547,
+      "num_input_tokens_seen": 117727920,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6663860999218422,
+      "grad_norm": 3.67318750815592,
+      "learning_rate": 1.0581083260506198e-06,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 117744425,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6665063428124812,
+      "grad_norm": 2.5154150695839004,
+      "learning_rate": 1.0574212182246993e-06,
+      "loss": 0.7631,
+      "num_input_tokens_seen": 117762840,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6666265857031203,
+      "grad_norm": 2.588424178863202,
+      "learning_rate": 1.0567342533872303e-06,
+      "loss": 0.7466,
+      "num_input_tokens_seen": 117782590,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6667468285937594,
+      "grad_norm": 1.9228600443687214,
+      "learning_rate": 1.0560474316424255e-06,
+      "loss": 0.8096,
+      "num_input_tokens_seen": 117802070,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6668670714843985,
+      "grad_norm": 4.824440139837768,
+      "learning_rate": 1.0553607530944746e-06,
+      "loss": 0.7376,
+      "num_input_tokens_seen": 117819845,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6669873143750376,
+      "grad_norm": 2.1490731892797927,
+      "learning_rate": 1.0546742178475463e-06,
+      "loss": 0.8884,
+      "num_input_tokens_seen": 117838560,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6671075572656767,
+      "grad_norm": 1.97939604325987,
+      "learning_rate": 1.0539878260057874e-06,
+      "loss": 0.8603,
+      "num_input_tokens_seen": 117857320,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6672278001563158,
+      "grad_norm": 2.865090115922973,
+      "learning_rate": 1.0533015776733237e-06,
+      "loss": 0.6731,
+      "num_input_tokens_seen": 117873190,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6673480430469548,
+      "grad_norm": 2.942809741309313,
+      "learning_rate": 1.0526154729542566e-06,
+      "loss": 0.7784,
+      "num_input_tokens_seen": 117892970,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6674682859375939,
+      "grad_norm": 3.428035061954582,
+      "learning_rate": 1.0519295119526699e-06,
+      "loss": 0.7931,
+      "num_input_tokens_seen": 117908995,
+      "step": 5551
+    },
+    {
+      "epoch": 0.667588528828233,
+      "grad_norm": 1.9148807291909227,
+      "learning_rate": 1.0512436947726227e-06,
+      "loss": 0.8235,
+      "num_input_tokens_seen": 117930130,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6677087717188721,
+      "grad_norm": 2.528009267599202,
+      "learning_rate": 1.0505580215181525e-06,
+      "loss": 0.6555,
+      "num_input_tokens_seen": 117948090,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6678290146095112,
+      "grad_norm": 0.8537304903581836,
+      "learning_rate": 1.0498724922932746e-06,
+      "loss": 0.5915,
+      "num_input_tokens_seen": 118005925,
+      "step": 5554
+    },
+    {
+      "epoch": 0.6679492575001503,
+      "grad_norm": 2.4357237326526895,
+      "learning_rate": 1.0491871072019851e-06,
+      "loss": 0.8545,
+      "num_input_tokens_seen": 118023535,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6680695003907894,
+      "grad_norm": 2.0398163692095124,
+      "learning_rate": 1.0485018663482555e-06,
+      "loss": 0.6301,
+      "num_input_tokens_seen": 118043275,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6681897432814284,
+      "grad_norm": 3.05609966831787,
+      "learning_rate": 1.0478167698360362e-06,
+      "loss": 0.7031,
+      "num_input_tokens_seen": 118062295,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6683099861720676,
+      "grad_norm": 2.602022834880324,
+      "learning_rate": 1.047131817769255e-06,
+      "loss": 0.697,
+      "num_input_tokens_seen": 118082315,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6684302290627067,
+      "grad_norm": 2.7578551688172204,
+      "learning_rate": 1.0464470102518203e-06,
+      "loss": 0.7472,
+      "num_input_tokens_seen": 118099365,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6685504719533457,
+      "grad_norm": 2.163731678669961,
+      "learning_rate": 1.0457623473876157e-06,
+      "loss": 0.757,
+      "num_input_tokens_seen": 118118590,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6686707148439849,
+      "grad_norm": 2.337917324343052,
+      "learning_rate": 1.0450778292805046e-06,
+      "loss": 0.6988,
+      "num_input_tokens_seen": 118138295,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6687909577346239,
+      "grad_norm": 2.437667662726632,
+      "learning_rate": 1.0443934560343276e-06,
+      "loss": 0.7861,
+      "num_input_tokens_seen": 118159425,
+      "step": 5562
+    },
+    {
+      "epoch": 0.668911200625263,
+      "grad_norm": 2.1442833880309693,
+      "learning_rate": 1.0437092277529034e-06,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 118178400,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6690314435159022,
+      "grad_norm": 2.27078078330895,
+      "learning_rate": 1.0430251445400292e-06,
+      "loss": 0.7406,
+      "num_input_tokens_seen": 118196165,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6691516864065412,
+      "grad_norm": 3.3023129754038925,
+      "learning_rate": 1.0423412064994794e-06,
+      "loss": 0.6225,
+      "num_input_tokens_seen": 118216655,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6692719292971803,
+      "grad_norm": 2.96278896963116,
+      "learning_rate": 1.0416574137350064e-06,
+      "loss": 0.7394,
+      "num_input_tokens_seen": 118237080,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6693921721878194,
+      "grad_norm": 2.939401981641236,
+      "learning_rate": 1.0409737663503428e-06,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 118255180,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6695124150784585,
+      "grad_norm": 2.8569625755783408,
+      "learning_rate": 1.040290264449196e-06,
+      "loss": 0.836,
+      "num_input_tokens_seen": 118273005,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6696326579690975,
+      "grad_norm": 2.2391590844751557,
+      "learning_rate": 1.0396069081352532e-06,
+      "loss": 0.629,
+      "num_input_tokens_seen": 118291880,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6697529008597367,
+      "grad_norm": 0.8826317822324508,
+      "learning_rate": 1.0389236975121782e-06,
+      "loss": 0.596,
+      "num_input_tokens_seen": 118346450,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6698731437503758,
+      "grad_norm": 3.313814000089316,
+      "learning_rate": 1.0382406326836147e-06,
+      "loss": 0.7161,
+      "num_input_tokens_seen": 118365315,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6699933866410148,
+      "grad_norm": 2.4235946257177177,
+      "learning_rate": 1.0375577137531828e-06,
+      "loss": 0.7435,
+      "num_input_tokens_seen": 118383595,
+      "step": 5572
+    },
+    {
+      "epoch": 0.670113629531654,
+      "grad_norm": 1.813027822895774,
+      "learning_rate": 1.0368749408244802e-06,
+      "loss": 0.7149,
+      "num_input_tokens_seen": 118406235,
+      "step": 5573
+    },
+    {
+      "epoch": 0.670233872422293,
+      "grad_norm": 2.259117521308527,
+      "learning_rate": 1.0361923140010827e-06,
+      "loss": 0.781,
+      "num_input_tokens_seen": 118424440,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6703541153129321,
+      "grad_norm": 3.0086824393775347,
+      "learning_rate": 1.0355098333865455e-06,
+      "loss": 0.6374,
+      "num_input_tokens_seen": 118443390,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6704743582035713,
+      "grad_norm": 2.1281985344684466,
+      "learning_rate": 1.0348274990844e-06,
+      "loss": 0.6864,
+      "num_input_tokens_seen": 118465870,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6705946010942103,
+      "grad_norm": 2.1619763019913876,
+      "learning_rate": 1.034145311198155e-06,
+      "loss": 0.721,
+      "num_input_tokens_seen": 118485605,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6707148439848494,
+      "grad_norm": 2.019244425422779,
+      "learning_rate": 1.0334632698312989e-06,
+      "loss": 0.636,
+      "num_input_tokens_seen": 118506120,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6708350868754885,
+      "grad_norm": 2.208751621630716,
+      "learning_rate": 1.0327813750872958e-06,
+      "loss": 0.7459,
+      "num_input_tokens_seen": 118525740,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6709553297661276,
+      "grad_norm": 1.6084535602024201,
+      "learning_rate": 1.0320996270695891e-06,
+      "loss": 0.6659,
+      "num_input_tokens_seen": 118546530,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6710755726567667,
+      "grad_norm": 3.7151223899624943,
+      "learning_rate": 1.0314180258815998e-06,
+      "loss": 0.7297,
+      "num_input_tokens_seen": 118564890,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6711958155474057,
+      "grad_norm": 2.110729006405776,
+      "learning_rate": 1.0307365716267247e-06,
+      "loss": 0.7411,
+      "num_input_tokens_seen": 118585055,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6713160584380449,
+      "grad_norm": 3.1181906940330326,
+      "learning_rate": 1.0300552644083423e-06,
+      "loss": 0.7748,
+      "num_input_tokens_seen": 118603700,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6714363013286839,
+      "grad_norm": 3.2785691103822248,
+      "learning_rate": 1.0293741043298045e-06,
+      "loss": 0.7331,
+      "num_input_tokens_seen": 118621770,
+      "step": 5584
+    },
+    {
+      "epoch": 0.671556544219323,
+      "grad_norm": 3.567506504420239,
+      "learning_rate": 1.0286930914944436e-06,
+      "loss": 0.7147,
+      "num_input_tokens_seen": 118641305,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6716767871099621,
+      "grad_norm": 2.571966391192147,
+      "learning_rate": 1.0280122260055678e-06,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 118656735,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6717970300006012,
+      "grad_norm": 2.1708265279817085,
+      "learning_rate": 1.0273315079664652e-06,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 118674410,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6719172728912403,
+      "grad_norm": 2.196258295026403,
+      "learning_rate": 1.0266509374803992e-06,
+      "loss": 0.7523,
+      "num_input_tokens_seen": 118695290,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6720375157818794,
+      "grad_norm": 7.56596757654837,
+      "learning_rate": 1.0259705146506123e-06,
+      "loss": 0.8359,
+      "num_input_tokens_seen": 118709905,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6721577586725185,
+      "grad_norm": 2.066362511497974,
+      "learning_rate": 1.025290239580324e-06,
+      "loss": 0.7665,
+      "num_input_tokens_seen": 118730295,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6722780015631575,
+      "grad_norm": 2.116341385128596,
+      "learning_rate": 1.0246101123727313e-06,
+      "loss": 0.7529,
+      "num_input_tokens_seen": 118748995,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6723982444537967,
+      "grad_norm": 2.3846078621139855,
+      "learning_rate": 1.023930133131009e-06,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 118766335,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6725184873444358,
+      "grad_norm": 1.6127821865552134,
+      "learning_rate": 1.0232503019583094e-06,
+      "loss": 0.8776,
+      "num_input_tokens_seen": 118785665,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6726387302350748,
+      "grad_norm": 2.17079772582495,
+      "learning_rate": 1.0225706189577619e-06,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 118803910,
+      "step": 5594
+    },
+    {
+      "epoch": 0.672758973125714,
+      "grad_norm": 2.879632471789694,
+      "learning_rate": 1.021891084232475e-06,
+      "loss": 0.7378,
+      "num_input_tokens_seen": 118821565,
+      "step": 5595
+    },
+    {
+      "epoch": 0.672879216016353,
+      "grad_norm": 4.997962772332968,
+      "learning_rate": 1.0212116978855325e-06,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 118839300,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6729994589069921,
+      "grad_norm": 1.7452230524892114,
+      "learning_rate": 1.0205324600199976e-06,
+      "loss": 0.7824,
+      "num_input_tokens_seen": 118858270,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6731197017976313,
+      "grad_norm": 2.0223611672689428,
+      "learning_rate": 1.0198533707389088e-06,
+      "loss": 0.6926,
+      "num_input_tokens_seen": 118878865,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6732399446882703,
+      "grad_norm": 2.0880100859394335,
+      "learning_rate": 1.0191744301452853e-06,
+      "loss": 0.731,
+      "num_input_tokens_seen": 118897885,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6733601875789094,
+      "grad_norm": 3.3222466948532134,
+      "learning_rate": 1.0184956383421208e-06,
+      "loss": 0.6989,
+      "num_input_tokens_seen": 118916255,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6734804304695485,
+      "grad_norm": 3.004318834775927,
+      "learning_rate": 1.0178169954323876e-06,
+      "loss": 0.6486,
+      "num_input_tokens_seen": 118935075,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6736006733601876,
+      "grad_norm": 1.9209239396004463,
+      "learning_rate": 1.0171385015190347e-06,
+      "loss": 0.7374,
+      "num_input_tokens_seen": 118954655,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6737209162508266,
+      "grad_norm": 2.292331758203086,
+      "learning_rate": 1.0164601567049902e-06,
+      "loss": 0.7219,
+      "num_input_tokens_seen": 118972905,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6738411591414658,
+      "grad_norm": 2.223441180967998,
+      "learning_rate": 1.015781961093158e-06,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 118991945,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6739614020321049,
+      "grad_norm": 1.8999873790586812,
+      "learning_rate": 1.0151039147864197e-06,
+      "loss": 0.764,
+      "num_input_tokens_seen": 119011640,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6740816449227439,
+      "grad_norm": 2.341380340720333,
+      "learning_rate": 1.0144260178876342e-06,
+      "loss": 0.6659,
+      "num_input_tokens_seen": 119030705,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6742018878133831,
+      "grad_norm": 2.7409618712656276,
+      "learning_rate": 1.0137482704996388e-06,
+      "loss": 0.6684,
+      "num_input_tokens_seen": 119044775,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6743221307040221,
+      "grad_norm": 6.963618248197937,
+      "learning_rate": 1.0130706727252461e-06,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 119061550,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6744423735946612,
+      "grad_norm": 3.1627852451343337,
+      "learning_rate": 1.0123932246672477e-06,
+      "loss": 0.673,
+      "num_input_tokens_seen": 119075415,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6745626164853004,
+      "grad_norm": 0.814674407097078,
+      "learning_rate": 1.0117159264284114e-06,
+      "loss": 0.5849,
+      "num_input_tokens_seen": 119138305,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6746828593759394,
+      "grad_norm": 2.018503852648043,
+      "learning_rate": 1.0110387781114837e-06,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 119156640,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6748031022665785,
+      "grad_norm": 3.414565340989417,
+      "learning_rate": 1.0103617798191872e-06,
+      "loss": 0.761,
+      "num_input_tokens_seen": 119175835,
+      "step": 5612
+    },
+    {
+      "epoch": 0.6749233451572175,
+      "grad_norm": 2.4903377724153843,
+      "learning_rate": 1.0096849316542217e-06,
+      "loss": 0.82,
+      "num_input_tokens_seen": 119192105,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6750435880478567,
+      "grad_norm": 2.798613139488539,
+      "learning_rate": 1.0090082337192643e-06,
+      "loss": 0.7407,
+      "num_input_tokens_seen": 119211470,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6751638309384957,
+      "grad_norm": 5.579463810680022,
+      "learning_rate": 1.0083316861169705e-06,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 119229925,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6752840738291348,
+      "grad_norm": 3.2200899173212356,
+      "learning_rate": 1.0076552889499713e-06,
+      "loss": 0.7205,
+      "num_input_tokens_seen": 119250410,
+      "step": 5616
+    },
+    {
+      "epoch": 0.675404316719774,
+      "grad_norm": 2.230974019375234,
+      "learning_rate": 1.006979042320876e-06,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 119270345,
+      "step": 5617
+    },
+    {
+      "epoch": 0.675524559610413,
+      "grad_norm": 3.4822170858000105,
+      "learning_rate": 1.0063029463322693e-06,
+      "loss": 0.6313,
+      "num_input_tokens_seen": 119290340,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6756448025010521,
+      "grad_norm": 3.7485509339466563,
+      "learning_rate": 1.0056270010867164e-06,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 119307630,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6757650453916912,
+      "grad_norm": 3.752538983738534,
+      "learning_rate": 1.004951206686757e-06,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 119325625,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6758852882823303,
+      "grad_norm": 4.203164371883197,
+      "learning_rate": 1.0042755632349087e-06,
+      "loss": 0.7127,
+      "num_input_tokens_seen": 119342235,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6760055311729694,
+      "grad_norm": 2.1761723454268687,
+      "learning_rate": 1.0036000708336653e-06,
+      "loss": 0.6199,
+      "num_input_tokens_seen": 119361085,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6761257740636085,
+      "grad_norm": 5.062662890212761,
+      "learning_rate": 1.0029247295854992e-06,
+      "loss": 0.7917,
+      "num_input_tokens_seen": 119377425,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6762460169542476,
+      "grad_norm": 2.517789843598912,
+      "learning_rate": 1.0022495395928588e-06,
+      "loss": 0.7104,
+      "num_input_tokens_seen": 119395625,
+      "step": 5624
+    },
+    {
+      "epoch": 0.6763662598448866,
+      "grad_norm": 0.8184953966886332,
+      "learning_rate": 1.0015745009581697e-06,
+      "loss": 0.639,
+      "num_input_tokens_seen": 119456950,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6764865027355258,
+      "grad_norm": 2.308923968912805,
+      "learning_rate": 1.0008996137838343e-06,
+      "loss": 0.6617,
+      "num_input_tokens_seen": 119475645,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6766067456261649,
+      "grad_norm": 2.3129380116268927,
+      "learning_rate": 1.000224878172234e-06,
+      "loss": 0.7946,
+      "num_input_tokens_seen": 119494490,
+      "step": 5627
+    },
+    {
+      "epoch": 0.6767269885168039,
+      "grad_norm": 2.2867345690262533,
+      "learning_rate": 9.995502942257248e-07,
+      "loss": 0.7289,
+      "num_input_tokens_seen": 119513365,
+      "step": 5628
+    },
+    {
+      "epoch": 0.6768472314074431,
+      "grad_norm": 2.850174201235272,
+      "learning_rate": 9.988758620466402e-07,
+      "loss": 0.7105,
+      "num_input_tokens_seen": 119531955,
+      "step": 5629
+    },
+    {
+      "epoch": 0.6769674742980821,
+      "grad_norm": 2.145077494151295,
+      "learning_rate": 9.982015817372909e-07,
+      "loss": 0.7579,
+      "num_input_tokens_seen": 119552115,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6770877171887212,
+      "grad_norm": 2.3290565314353384,
+      "learning_rate": 9.975274533999657e-07,
+      "loss": 0.8142,
+      "num_input_tokens_seen": 119571365,
+      "step": 5631
+    },
+    {
+      "epoch": 0.6772079600793603,
+      "grad_norm": 3.41008237538907,
+      "learning_rate": 9.96853477136929e-07,
+      "loss": 0.8395,
+      "num_input_tokens_seen": 119585830,
+      "step": 5632
+    },
+    {
+      "epoch": 0.6773282029699994,
+      "grad_norm": 2.4093251328529974,
+      "learning_rate": 9.96179653050422e-07,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 119605710,
+      "step": 5633
+    },
+    {
+      "epoch": 0.6774484458606385,
+      "grad_norm": 2.3429182986714996,
+      "learning_rate": 9.955059812426635e-07,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 119622960,
+      "step": 5634
+    },
+    {
+      "epoch": 0.6775686887512776,
+      "grad_norm": 2.4273938589357074,
+      "learning_rate": 9.948324618158493e-07,
+      "loss": 0.818,
+      "num_input_tokens_seen": 119643020,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6776889316419167,
+      "grad_norm": 3.4230700719679716,
+      "learning_rate": 9.941590948721513e-07,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 119659940,
+      "step": 5636
+    },
+    {
+      "epoch": 0.6778091745325557,
+      "grad_norm": 2.0615651878489665,
+      "learning_rate": 9.934858805137188e-07,
+      "loss": 0.7551,
+      "num_input_tokens_seen": 119680310,
+      "step": 5637
+    },
+    {
+      "epoch": 0.6779294174231949,
+      "grad_norm": 2.0823722460691925,
+      "learning_rate": 9.92812818842677e-07,
+      "loss": 0.8002,
+      "num_input_tokens_seen": 119699205,
+      "step": 5638
+    },
+    {
+      "epoch": 0.678049660313834,
+      "grad_norm": 2.173764720267259,
+      "learning_rate": 9.921399099611306e-07,
+      "loss": 0.6341,
+      "num_input_tokens_seen": 119720090,
+      "step": 5639
+    },
+    {
+      "epoch": 0.678169903204473,
+      "grad_norm": 1.7570937163770215,
+      "learning_rate": 9.914671539711588e-07,
+      "loss": 0.6918,
+      "num_input_tokens_seen": 119739330,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6782901460951122,
+      "grad_norm": 2.4106060797124567,
+      "learning_rate": 9.907945509748176e-07,
+      "loss": 0.7802,
+      "num_input_tokens_seen": 119759445,
+      "step": 5641
+    },
+    {
+      "epoch": 0.6784103889857512,
+      "grad_norm": 2.945091421478644,
+      "learning_rate": 9.9012210107414e-07,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 119778485,
+      "step": 5642
+    },
+    {
+      "epoch": 0.6785306318763903,
+      "grad_norm": 2.31250801659447,
+      "learning_rate": 9.894498043711375e-07,
+      "loss": 0.7438,
+      "num_input_tokens_seen": 119799950,
+      "step": 5643
+    },
+    {
+      "epoch": 0.6786508747670293,
+      "grad_norm": 2.447018036923878,
+      "learning_rate": 9.887776609677962e-07,
+      "loss": 0.6914,
+      "num_input_tokens_seen": 119821040,
+      "step": 5644
+    },
+    {
+      "epoch": 0.6787711176576685,
+      "grad_norm": 5.23400249343789,
+      "learning_rate": 9.881056709660796e-07,
+      "loss": 0.7185,
+      "num_input_tokens_seen": 119839220,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6788913605483076,
+      "grad_norm": 1.9128059623042541,
+      "learning_rate": 9.874338344679274e-07,
+      "loss": 0.7797,
+      "num_input_tokens_seen": 119854785,
+      "step": 5646
+    },
+    {
+      "epoch": 0.6790116034389466,
+      "grad_norm": 2.44979856250486,
+      "learning_rate": 9.867621515752582e-07,
+      "loss": 0.7396,
+      "num_input_tokens_seen": 119874500,
+      "step": 5647
+    },
+    {
+      "epoch": 0.6791318463295858,
+      "grad_norm": 1.5706120592930803,
+      "learning_rate": 9.860906223899651e-07,
+      "loss": 0.7901,
+      "num_input_tokens_seen": 119893615,
+      "step": 5648
+    },
+    {
+      "epoch": 0.6792520892202248,
+      "grad_norm": 2.088331952034577,
+      "learning_rate": 9.854192470139184e-07,
+      "loss": 0.749,
+      "num_input_tokens_seen": 119914815,
+      "step": 5649
+    },
+    {
+      "epoch": 0.6793723321108639,
+      "grad_norm": 2.6482426846029066,
+      "learning_rate": 9.847480255489653e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 119933560,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6794925750015031,
+      "grad_norm": 1.8001434363990196,
+      "learning_rate": 9.840769580969295e-07,
+      "loss": 0.69,
+      "num_input_tokens_seen": 119953720,
+      "step": 5651
+    },
+    {
+      "epoch": 0.6796128178921421,
+      "grad_norm": 2.223310195039357,
+      "learning_rate": 9.834060447596114e-07,
+      "loss": 0.7929,
+      "num_input_tokens_seen": 119972710,
+      "step": 5652
+    },
+    {
+      "epoch": 0.6797330607827812,
+      "grad_norm": 3.0762282727713717,
+      "learning_rate": 9.82735285638788e-07,
+      "loss": 0.7713,
+      "num_input_tokens_seen": 119992140,
+      "step": 5653
+    },
+    {
+      "epoch": 0.6798533036734203,
+      "grad_norm": 0.8182982149600501,
+      "learning_rate": 9.820646808362118e-07,
+      "loss": 0.6533,
+      "num_input_tokens_seen": 120058115,
+      "step": 5654
+    },
+    {
+      "epoch": 0.6799735465640594,
+      "grad_norm": 2.0630662587600384,
+      "learning_rate": 9.813942304536154e-07,
+      "loss": 0.7171,
+      "num_input_tokens_seen": 120075805,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6800937894546984,
+      "grad_norm": 1.9501571398177167,
+      "learning_rate": 9.807239345927043e-07,
+      "loss": 0.6368,
+      "num_input_tokens_seen": 120095535,
+      "step": 5656
+    },
+    {
+      "epoch": 0.6802140323453376,
+      "grad_norm": 3.621713051897509,
+      "learning_rate": 9.80053793355162e-07,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 120113950,
+      "step": 5657
+    },
+    {
+      "epoch": 0.6803342752359767,
+      "grad_norm": 2.695622203290369,
+      "learning_rate": 9.793838068426472e-07,
+      "loss": 0.7407,
+      "num_input_tokens_seen": 120131365,
+      "step": 5658
+    },
+    {
+      "epoch": 0.6804545181266157,
+      "grad_norm": 2.546395038835068,
+      "learning_rate": 9.787139751567983e-07,
+      "loss": 0.6013,
+      "num_input_tokens_seen": 120146950,
+      "step": 5659
+    },
+    {
+      "epoch": 0.6805747610172549,
+      "grad_norm": 1.91082587396936,
+      "learning_rate": 9.780442983992273e-07,
+      "loss": 0.7166,
+      "num_input_tokens_seen": 120165185,
+      "step": 5660
+    },
+    {
+      "epoch": 0.680695003907894,
+      "grad_norm": 1.907277149391533,
+      "learning_rate": 9.773747766715238e-07,
+      "loss": 0.71,
+      "num_input_tokens_seen": 120185725,
+      "step": 5661
+    },
+    {
+      "epoch": 0.680815246798533,
+      "grad_norm": 5.563003309588458,
+      "learning_rate": 9.76705410075253e-07,
+      "loss": 0.8025,
+      "num_input_tokens_seen": 120205395,
+      "step": 5662
+    },
+    {
+      "epoch": 0.6809354896891722,
+      "grad_norm": 2.1069139067868754,
+      "learning_rate": 9.760361987119584e-07,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 120222850,
+      "step": 5663
+    },
+    {
+      "epoch": 0.6810557325798112,
+      "grad_norm": 2.5796659771674517,
+      "learning_rate": 9.753671426831584e-07,
+      "loss": 0.6777,
+      "num_input_tokens_seen": 120238585,
+      "step": 5664
+    },
+    {
+      "epoch": 0.6811759754704503,
+      "grad_norm": 5.37964584866951,
+      "learning_rate": 9.746982420903483e-07,
+      "loss": 0.7882,
+      "num_input_tokens_seen": 120256500,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6812962183610894,
+      "grad_norm": 1.8747188413475615,
+      "learning_rate": 9.740294970349993e-07,
+      "loss": 0.7399,
+      "num_input_tokens_seen": 120272635,
+      "step": 5666
+    },
+    {
+      "epoch": 0.6814164612517285,
+      "grad_norm": 0.9969339598575847,
+      "learning_rate": 9.733609076185602e-07,
+      "loss": 0.68,
+      "num_input_tokens_seen": 120328760,
+      "step": 5667
+    },
+    {
+      "epoch": 0.6815367041423676,
+      "grad_norm": 2.338583704568072,
+      "learning_rate": 9.72692473942455e-07,
+      "loss": 0.8331,
+      "num_input_tokens_seen": 120345705,
+      "step": 5668
+    },
+    {
+      "epoch": 0.6816569470330067,
+      "grad_norm": 1.825224673788321,
+      "learning_rate": 9.720241961080849e-07,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 120364740,
+      "step": 5669
+    },
+    {
+      "epoch": 0.6817771899236458,
+      "grad_norm": 2.5813101895172883,
+      "learning_rate": 9.713560742168259e-07,
+      "loss": 0.7149,
+      "num_input_tokens_seen": 120387085,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6818974328142848,
+      "grad_norm": 2.17815696213839,
+      "learning_rate": 9.706881083700333e-07,
+      "loss": 0.7101,
+      "num_input_tokens_seen": 120406490,
+      "step": 5671
+    },
+    {
+      "epoch": 0.682017675704924,
+      "grad_norm": 2.2964241827977125,
+      "learning_rate": 9.700202986690364e-07,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 120424510,
+      "step": 5672
+    },
+    {
+      "epoch": 0.682137918595563,
+      "grad_norm": 3.135653361618271,
+      "learning_rate": 9.693526452151413e-07,
+      "loss": 0.6511,
+      "num_input_tokens_seen": 120443280,
+      "step": 5673
+    },
+    {
+      "epoch": 0.6822581614862021,
+      "grad_norm": 2.5123060235516554,
+      "learning_rate": 9.686851481096296e-07,
+      "loss": 0.755,
+      "num_input_tokens_seen": 120464310,
+      "step": 5674
+    },
+    {
+      "epoch": 0.6823784043768413,
+      "grad_norm": 2.7598060704314764,
+      "learning_rate": 9.68017807453762e-07,
+      "loss": 0.726,
+      "num_input_tokens_seen": 120482775,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6824986472674803,
+      "grad_norm": 1.8674442909494517,
+      "learning_rate": 9.673506233487721e-07,
+      "loss": 0.7258,
+      "num_input_tokens_seen": 120500460,
+      "step": 5676
+    },
+    {
+      "epoch": 0.6826188901581194,
+      "grad_norm": 2.118331678142326,
+      "learning_rate": 9.666835958958717e-07,
+      "loss": 0.8595,
+      "num_input_tokens_seen": 120519500,
+      "step": 5677
+    },
+    {
+      "epoch": 0.6827391330487584,
+      "grad_norm": 2.084150152897892,
+      "learning_rate": 9.660167251962484e-07,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 120537580,
+      "step": 5678
+    },
+    {
+      "epoch": 0.6828593759393976,
+      "grad_norm": 1.7788692722215675,
+      "learning_rate": 9.653500113510654e-07,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 120556415,
+      "step": 5679
+    },
+    {
+      "epoch": 0.6829796188300367,
+      "grad_norm": 4.2533523454507245,
+      "learning_rate": 9.646834544614635e-07,
+      "loss": 0.6635,
+      "num_input_tokens_seen": 120576635,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6830998617206757,
+      "grad_norm": 2.4656761102258433,
+      "learning_rate": 9.64017054628558e-07,
+      "loss": 0.7544,
+      "num_input_tokens_seen": 120595180,
+      "step": 5681
+    },
+    {
+      "epoch": 0.6832201046113149,
+      "grad_norm": 2.135077869251625,
+      "learning_rate": 9.63350811953441e-07,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 120615275,
+      "step": 5682
+    },
+    {
+      "epoch": 0.6833403475019539,
+      "grad_norm": 2.644328645466541,
+      "learning_rate": 9.626847265371826e-07,
+      "loss": 0.6986,
+      "num_input_tokens_seen": 120634315,
+      "step": 5683
+    },
+    {
+      "epoch": 0.683460590392593,
+      "grad_norm": 3.2994741178463634,
+      "learning_rate": 9.620187984808262e-07,
+      "loss": 0.7801,
+      "num_input_tokens_seen": 120652835,
+      "step": 5684
+    },
+    {
+      "epoch": 0.6835808332832322,
+      "grad_norm": 2.6794205210352104,
+      "learning_rate": 9.613530278853927e-07,
+      "loss": 0.8511,
+      "num_input_tokens_seen": 120672530,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6837010761738712,
+      "grad_norm": 2.5252554782011836,
+      "learning_rate": 9.606874148518782e-07,
+      "loss": 0.739,
+      "num_input_tokens_seen": 120693255,
+      "step": 5686
+    },
+    {
+      "epoch": 0.6838213190645103,
+      "grad_norm": 2.349412548068551,
+      "learning_rate": 9.600219594812575e-07,
+      "loss": 0.7674,
+      "num_input_tokens_seen": 120710915,
+      "step": 5687
+    },
+    {
+      "epoch": 0.6839415619551494,
+      "grad_norm": 2.126578202663835,
+      "learning_rate": 9.593566618744786e-07,
+      "loss": 0.7199,
+      "num_input_tokens_seen": 120730785,
+      "step": 5688
+    },
+    {
+      "epoch": 0.6840618048457885,
+      "grad_norm": 1.9560459136567503,
+      "learning_rate": 9.586915221324668e-07,
+      "loss": 0.729,
+      "num_input_tokens_seen": 120749315,
+      "step": 5689
+    },
+    {
+      "epoch": 0.6841820477364275,
+      "grad_norm": 2.3849725891692786,
+      "learning_rate": 9.580265403561222e-07,
+      "loss": 0.8371,
+      "num_input_tokens_seen": 120767300,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6843022906270667,
+      "grad_norm": 1.8902344500481807,
+      "learning_rate": 9.57361716646324e-07,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 120788235,
+      "step": 5691
+    },
+    {
+      "epoch": 0.6844225335177058,
+      "grad_norm": 2.0329187542770875,
+      "learning_rate": 9.56697051103924e-07,
+      "loss": 0.5945,
+      "num_input_tokens_seen": 120805395,
+      "step": 5692
+    },
+    {
+      "epoch": 0.6845427764083448,
+      "grad_norm": 2.419546725692858,
+      "learning_rate": 9.560325438297522e-07,
+      "loss": 0.8009,
+      "num_input_tokens_seen": 120823425,
+      "step": 5693
+    },
+    {
+      "epoch": 0.684663019298984,
+      "grad_norm": 3.189173872260607,
+      "learning_rate": 9.553681949246134e-07,
+      "loss": 0.8701,
+      "num_input_tokens_seen": 120840770,
+      "step": 5694
+    },
+    {
+      "epoch": 0.684783262189623,
+      "grad_norm": 2.485866181990579,
+      "learning_rate": 9.547040044892886e-07,
+      "loss": 0.7519,
+      "num_input_tokens_seen": 120868005,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6849035050802621,
+      "grad_norm": 0.8976487129122176,
+      "learning_rate": 9.540399726245354e-07,
+      "loss": 0.6448,
+      "num_input_tokens_seen": 120924430,
+      "step": 5696
+    },
+    {
+      "epoch": 0.6850237479709013,
+      "grad_norm": 2.4773478305203485,
+      "learning_rate": 9.533760994310867e-07,
+      "loss": 0.6865,
+      "num_input_tokens_seen": 120944550,
+      "step": 5697
+    },
+    {
+      "epoch": 0.6851439908615403,
+      "grad_norm": 3.7819043827613443,
+      "learning_rate": 9.527123850096508e-07,
+      "loss": 0.742,
+      "num_input_tokens_seen": 120962630,
+      "step": 5698
+    },
+    {
+      "epoch": 0.6852642337521794,
+      "grad_norm": 2.004535920611483,
+      "learning_rate": 9.520488294609142e-07,
+      "loss": 0.7115,
+      "num_input_tokens_seen": 120981130,
+      "step": 5699
+    },
+    {
+      "epoch": 0.6853844766428185,
+      "grad_norm": 0.9791123132195761,
+      "learning_rate": 9.513854328855368e-07,
+      "loss": 0.6,
+      "num_input_tokens_seen": 121038725,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6855047195334576,
+      "grad_norm": 2.2853095570772792,
+      "learning_rate": 9.507221953841558e-07,
+      "loss": 0.8114,
+      "num_input_tokens_seen": 121056075,
+      "step": 5701
+    },
+    {
+      "epoch": 0.6856249624240967,
+      "grad_norm": 2.6403895812060156,
+      "learning_rate": 9.500591170573824e-07,
+      "loss": 0.7716,
+      "num_input_tokens_seen": 121075815,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6857452053147358,
+      "grad_norm": 2.527892227599656,
+      "learning_rate": 9.49396198005807e-07,
+      "loss": 0.7381,
+      "num_input_tokens_seen": 121093130,
+      "step": 5703
+    },
+    {
+      "epoch": 0.6858654482053749,
+      "grad_norm": 3.9979097183038603,
+      "learning_rate": 9.48733438329993e-07,
+      "loss": 0.6802,
+      "num_input_tokens_seen": 121113115,
+      "step": 5704
+    },
+    {
+      "epoch": 0.6859856910960139,
+      "grad_norm": 2.2221661064097593,
+      "learning_rate": 9.480708381304807e-07,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 121134130,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6861059339866531,
+      "grad_norm": 2.164600640702911,
+      "learning_rate": 9.474083975077851e-07,
+      "loss": 0.8336,
+      "num_input_tokens_seen": 121150975,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6862261768772921,
+      "grad_norm": 3.5821268171544793,
+      "learning_rate": 9.467461165623994e-07,
+      "loss": 0.7939,
+      "num_input_tokens_seen": 121169745,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6863464197679312,
+      "grad_norm": 3.203362841298789,
+      "learning_rate": 9.460839953947903e-07,
+      "loss": 0.7871,
+      "num_input_tokens_seen": 121187275,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6864666626585703,
+      "grad_norm": 2.835659700586604,
+      "learning_rate": 9.45422034105402e-07,
+      "loss": 0.6274,
+      "num_input_tokens_seen": 121211780,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6865869055492094,
+      "grad_norm": 2.0545390856391763,
+      "learning_rate": 9.447602327946512e-07,
+      "loss": 0.8033,
+      "num_input_tokens_seen": 121230140,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6867071484398485,
+      "grad_norm": 3.951160177290665,
+      "learning_rate": 9.440985915629345e-07,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 121247190,
+      "step": 5711
+    },
+    {
+      "epoch": 0.6868273913304875,
+      "grad_norm": 2.154865349124956,
+      "learning_rate": 9.434371105106223e-07,
+      "loss": 0.7187,
+      "num_input_tokens_seen": 121264510,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6869476342211267,
+      "grad_norm": 3.12363571666529,
+      "learning_rate": 9.427757897380602e-07,
+      "loss": 0.7088,
+      "num_input_tokens_seen": 121283630,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6870678771117658,
+      "grad_norm": 10.19452029343035,
+      "learning_rate": 9.421146293455695e-07,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 121299090,
+      "step": 5714
+    },
+    {
+      "epoch": 0.6871881200024048,
+      "grad_norm": 1.9349666369499914,
+      "learning_rate": 9.414536294334489e-07,
+      "loss": 0.6757,
+      "num_input_tokens_seen": 121318830,
+      "step": 5715
+    },
+    {
+      "epoch": 0.687308362893044,
+      "grad_norm": 2.1640124836854375,
+      "learning_rate": 9.407927901019714e-07,
+      "loss": 0.6974,
+      "num_input_tokens_seen": 121337680,
+      "step": 5716
+    },
+    {
+      "epoch": 0.687428605783683,
+      "grad_norm": 4.211496220563102,
+      "learning_rate": 9.401321114513854e-07,
+      "loss": 0.7645,
+      "num_input_tokens_seen": 121356295,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6875488486743221,
+      "grad_norm": 1.758534795181603,
+      "learning_rate": 9.394715935819146e-07,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 121376405,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6876690915649613,
+      "grad_norm": 2.501188733728939,
+      "learning_rate": 9.388112365937608e-07,
+      "loss": 0.6215,
+      "num_input_tokens_seen": 121395590,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6877893344556003,
+      "grad_norm": 32.62870847194971,
+      "learning_rate": 9.381510405870985e-07,
+      "loss": 0.8191,
+      "num_input_tokens_seen": 121414325,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6879095773462394,
+      "grad_norm": 2.728350377519243,
+      "learning_rate": 9.374910056620791e-07,
+      "loss": 0.7683,
+      "num_input_tokens_seen": 121433110,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6880298202368785,
+      "grad_norm": 2.4241332673249163,
+      "learning_rate": 9.368311319188293e-07,
+      "loss": 0.811,
+      "num_input_tokens_seen": 121450645,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6881500631275176,
+      "grad_norm": 2.0276284322463924,
+      "learning_rate": 9.361714194574515e-07,
+      "loss": 0.7926,
+      "num_input_tokens_seen": 121472700,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6882703060181566,
+      "grad_norm": 0.765729410977009,
+      "learning_rate": 9.355118683780234e-07,
+      "loss": 0.5957,
+      "num_input_tokens_seen": 121542490,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6883905489087958,
+      "grad_norm": 2.4658602553767643,
+      "learning_rate": 9.348524787805987e-07,
+      "loss": 0.7897,
+      "num_input_tokens_seen": 121557400,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6885107917994349,
+      "grad_norm": 3.8896327875127072,
+      "learning_rate": 9.341932507652053e-07,
+      "loss": 0.8489,
+      "num_input_tokens_seen": 121571610,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6886310346900739,
+      "grad_norm": 2.1344892834610545,
+      "learning_rate": 9.335341844318489e-07,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 121591470,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6887512775807131,
+      "grad_norm": 2.132128088241014,
+      "learning_rate": 9.328752798805091e-07,
+      "loss": 0.7328,
+      "num_input_tokens_seen": 121609660,
+      "step": 5728
+    },
+    {
+      "epoch": 0.6888715204713521,
+      "grad_norm": 3.1093780408853133,
+      "learning_rate": 9.322165372111405e-07,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 121627525,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6889917633619912,
+      "grad_norm": 2.1333715568588474,
+      "learning_rate": 9.315579565236737e-07,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 121646350,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6891120062526304,
+      "grad_norm": 8.139317368291874,
+      "learning_rate": 9.308995379180162e-07,
+      "loss": 0.743,
+      "num_input_tokens_seen": 121665625,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6892322491432694,
+      "grad_norm": 0.8707265602570968,
+      "learning_rate": 9.302412814940488e-07,
+      "loss": 0.6356,
+      "num_input_tokens_seen": 121728120,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6893524920339085,
+      "grad_norm": 2.6750356340592796,
+      "learning_rate": 9.295831873516282e-07,
+      "loss": 0.6991,
+      "num_input_tokens_seen": 121747115,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6894727349245476,
+      "grad_norm": 1.538774631478507,
+      "learning_rate": 9.289252555905865e-07,
+      "loss": 0.758,
+      "num_input_tokens_seen": 121766915,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6895929778151867,
+      "grad_norm": 2.377370084709072,
+      "learning_rate": 9.282674863107325e-07,
+      "loss": 0.7549,
+      "num_input_tokens_seen": 121784450,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6897132207058257,
+      "grad_norm": 2.8749584527365637,
+      "learning_rate": 9.276098796118488e-07,
+      "loss": 0.7581,
+      "num_input_tokens_seen": 121800655,
+      "step": 5736
+    },
+    {
+      "epoch": 0.6898334635964649,
+      "grad_norm": 2.1607664353012934,
+      "learning_rate": 9.269524355936938e-07,
+      "loss": 0.6632,
+      "num_input_tokens_seen": 121823555,
+      "step": 5737
+    },
+    {
+      "epoch": 0.689953706487104,
+      "grad_norm": 1.768322661526969,
+      "learning_rate": 9.262951543560009e-07,
+      "loss": 0.8447,
+      "num_input_tokens_seen": 121842500,
+      "step": 5738
+    },
+    {
+      "epoch": 0.690073949377743,
+      "grad_norm": 2.691020657196348,
+      "learning_rate": 9.256380359984795e-07,
+      "loss": 0.8529,
+      "num_input_tokens_seen": 121859330,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6901941922683821,
+      "grad_norm": 2.1714322533165076,
+      "learning_rate": 9.249810806208139e-07,
+      "loss": 0.7415,
+      "num_input_tokens_seen": 121878315,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6903144351590212,
+      "grad_norm": 2.3190317057241585,
+      "learning_rate": 9.243242883226636e-07,
+      "loss": 0.8114,
+      "num_input_tokens_seen": 121897130,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6904346780496603,
+      "grad_norm": 2.76753980877324,
+      "learning_rate": 9.236676592036628e-07,
+      "loss": 0.6934,
+      "num_input_tokens_seen": 121916525,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6905549209402994,
+      "grad_norm": 2.1303453618743755,
+      "learning_rate": 9.230111933634228e-07,
+      "loss": 0.7232,
+      "num_input_tokens_seen": 121937840,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6906751638309385,
+      "grad_norm": 1.6923010245041947,
+      "learning_rate": 9.223548909015288e-07,
+      "loss": 0.8017,
+      "num_input_tokens_seen": 121959250,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6907954067215776,
+      "grad_norm": 3.130097704411959,
+      "learning_rate": 9.216987519175407e-07,
+      "loss": 0.7072,
+      "num_input_tokens_seen": 121979145,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6909156496122166,
+      "grad_norm": 2.501090917564905,
+      "learning_rate": 9.210427765109942e-07,
+      "loss": 0.6834,
+      "num_input_tokens_seen": 121998540,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6910358925028558,
+      "grad_norm": 2.726701896582755,
+      "learning_rate": 9.203869647814011e-07,
+      "loss": 0.8069,
+      "num_input_tokens_seen": 122016280,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6911561353934949,
+      "grad_norm": 3.140027158250298,
+      "learning_rate": 9.197313168282472e-07,
+      "loss": 0.8403,
+      "num_input_tokens_seen": 122033445,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6912763782841339,
+      "grad_norm": 5.375275732614449,
+      "learning_rate": 9.190758327509935e-07,
+      "loss": 0.7152,
+      "num_input_tokens_seen": 122051910,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6913966211747731,
+      "grad_norm": 0.9157513124102706,
+      "learning_rate": 9.184205126490761e-07,
+      "loss": 0.6711,
+      "num_input_tokens_seen": 122100525,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6915168640654121,
+      "grad_norm": 0.945610153369629,
+      "learning_rate": 9.177653566219075e-07,
+      "loss": 0.6404,
+      "num_input_tokens_seen": 122154970,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6916371069560512,
+      "grad_norm": 2.8989492381994553,
+      "learning_rate": 9.171103647688738e-07,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 122173430,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6917573498466904,
+      "grad_norm": 2.1568265652970733,
+      "learning_rate": 9.164555371893375e-07,
+      "loss": 0.6863,
+      "num_input_tokens_seen": 122193080,
+      "step": 5753
+    },
+    {
+      "epoch": 0.6918775927373294,
+      "grad_norm": 2.1854520104076025,
+      "learning_rate": 9.158008739826333e-07,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 122210400,
+      "step": 5754
+    },
+    {
+      "epoch": 0.6919978356279685,
+      "grad_norm": 1.7443003470287874,
+      "learning_rate": 9.15146375248075e-07,
+      "loss": 0.8532,
+      "num_input_tokens_seen": 122228850,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6921180785186076,
+      "grad_norm": 4.070780877577215,
+      "learning_rate": 9.144920410849493e-07,
+      "loss": 0.7947,
+      "num_input_tokens_seen": 122249805,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6922383214092467,
+      "grad_norm": 3.038691955820551,
+      "learning_rate": 9.138378715925176e-07,
+      "loss": 0.7957,
+      "num_input_tokens_seen": 122268620,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6923585642998857,
+      "grad_norm": 1.833583005051259,
+      "learning_rate": 9.131838668700167e-07,
+      "loss": 0.8057,
+      "num_input_tokens_seen": 122288410,
+      "step": 5758
+    },
+    {
+      "epoch": 0.6924788071905249,
+      "grad_norm": 3.469021084412814,
+      "learning_rate": 9.125300270166598e-07,
+      "loss": 0.8607,
+      "num_input_tokens_seen": 122308735,
+      "step": 5759
+    },
+    {
+      "epoch": 0.692599050081164,
+      "grad_norm": 1.920577696716791,
+      "learning_rate": 9.11876352131633e-07,
+      "loss": 0.8645,
+      "num_input_tokens_seen": 122329030,
+      "step": 5760
+    },
+    {
+      "epoch": 0.692719292971803,
+      "grad_norm": 1.8447309435312118,
+      "learning_rate": 9.112228423140987e-07,
+      "loss": 0.7534,
+      "num_input_tokens_seen": 122347670,
+      "step": 5761
+    },
+    {
+      "epoch": 0.6928395358624422,
+      "grad_norm": 2.603976261751542,
+      "learning_rate": 9.105694976631926e-07,
+      "loss": 0.856,
+      "num_input_tokens_seen": 122365300,
+      "step": 5762
+    },
+    {
+      "epoch": 0.6929597787530812,
+      "grad_norm": 2.760815713949478,
+      "learning_rate": 9.099163182780283e-07,
+      "loss": 0.7245,
+      "num_input_tokens_seen": 122383175,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6930800216437203,
+      "grad_norm": 3.018832470128621,
+      "learning_rate": 9.092633042576916e-07,
+      "loss": 0.4917,
+      "num_input_tokens_seen": 122400160,
+      "step": 5764
+    },
+    {
+      "epoch": 0.6932002645343595,
+      "grad_norm": 2.470595298487512,
+      "learning_rate": 9.086104557012446e-07,
+      "loss": 0.5699,
+      "num_input_tokens_seen": 122420450,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6933205074249985,
+      "grad_norm": 1.9883446091056796,
+      "learning_rate": 9.079577727077239e-07,
+      "loss": 0.6535,
+      "num_input_tokens_seen": 122439000,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6934407503156376,
+      "grad_norm": 2.8386987672233785,
+      "learning_rate": 9.073052553761404e-07,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 122458085,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6935609932062767,
+      "grad_norm": 3.5855480970972162,
+      "learning_rate": 9.066529038054811e-07,
+      "loss": 0.771,
+      "num_input_tokens_seen": 122477870,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6936812360969158,
+      "grad_norm": 2.015515930702198,
+      "learning_rate": 9.060007180947071e-07,
+      "loss": 0.7381,
+      "num_input_tokens_seen": 122495645,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6938014789875548,
+      "grad_norm": 2.20097854396206,
+      "learning_rate": 9.053486983427534e-07,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 122516615,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6939217218781939,
+      "grad_norm": 2.054274043760421,
+      "learning_rate": 9.046968446485326e-07,
+      "loss": 0.7021,
+      "num_input_tokens_seen": 122534740,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6940419647688331,
+      "grad_norm": 2.6846041985592817,
+      "learning_rate": 9.040451571109295e-07,
+      "loss": 0.692,
+      "num_input_tokens_seen": 122550080,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6941622076594721,
+      "grad_norm": 0.9973684702210112,
+      "learning_rate": 9.03393635828805e-07,
+      "loss": 0.6507,
+      "num_input_tokens_seen": 122603535,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6942824505501112,
+      "grad_norm": 1.8399738115287108,
+      "learning_rate": 9.02742280900993e-07,
+      "loss": 0.8167,
+      "num_input_tokens_seen": 122623200,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6944026934407503,
+      "grad_norm": 1.904868735645822,
+      "learning_rate": 9.020910924263054e-07,
+      "loss": 0.8266,
+      "num_input_tokens_seen": 122641445,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6945229363313894,
+      "grad_norm": 0.9423214519573473,
+      "learning_rate": 9.014400705035261e-07,
+      "loss": 0.6202,
+      "num_input_tokens_seen": 122698070,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6946431792220285,
+      "grad_norm": 5.401724289010299,
+      "learning_rate": 9.007892152314147e-07,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 122716185,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6947634221126676,
+      "grad_norm": 10.13534991058867,
+      "learning_rate": 9.001385267087047e-07,
+      "loss": 0.8203,
+      "num_input_tokens_seen": 122735050,
+      "step": 5778
+    },
+    {
+      "epoch": 0.6948836650033067,
+      "grad_norm": 2.31126092840529,
+      "learning_rate": 8.994880050341064e-07,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 122754875,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6950039078939457,
+      "grad_norm": 1.9673895105245716,
+      "learning_rate": 8.988376503063026e-07,
+      "loss": 0.7743,
+      "num_input_tokens_seen": 122775855,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6951241507845849,
+      "grad_norm": 2.5967403933541786,
+      "learning_rate": 8.981874626239521e-07,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 122794150,
+      "step": 5781
+    },
+    {
+      "epoch": 0.695244393675224,
+      "grad_norm": 2.696509190022972,
+      "learning_rate": 8.975374420856872e-07,
+      "loss": 0.8731,
+      "num_input_tokens_seen": 122810765,
+      "step": 5782
+    },
+    {
+      "epoch": 0.695364636565863,
+      "grad_norm": 2.361041820320028,
+      "learning_rate": 8.968875887901157e-07,
+      "loss": 0.7191,
+      "num_input_tokens_seen": 122827865,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6954848794565022,
+      "grad_norm": 2.6530293684449435,
+      "learning_rate": 8.9623790283582e-07,
+      "loss": 0.6246,
+      "num_input_tokens_seen": 122845465,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6956051223471412,
+      "grad_norm": 2.7323740234991307,
+      "learning_rate": 8.955883843213569e-07,
+      "loss": 0.759,
+      "num_input_tokens_seen": 122864200,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6957253652377803,
+      "grad_norm": 2.0600057429751257,
+      "learning_rate": 8.949390333452569e-07,
+      "loss": 0.8745,
+      "num_input_tokens_seen": 122881865,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6958456081284194,
+      "grad_norm": 2.0904027714530717,
+      "learning_rate": 8.942898500060279e-07,
+      "loss": 0.6737,
+      "num_input_tokens_seen": 122901300,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6959658510190585,
+      "grad_norm": 2.875111124277208,
+      "learning_rate": 8.936408344021493e-07,
+      "loss": 0.7222,
+      "num_input_tokens_seen": 122917935,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6960860939096976,
+      "grad_norm": 2.5961743687634904,
+      "learning_rate": 8.929919866320765e-07,
+      "loss": 0.7058,
+      "num_input_tokens_seen": 122938470,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6962063368003367,
+      "grad_norm": 3.278340845592569,
+      "learning_rate": 8.923433067942385e-07,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 122956755,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6963265796909758,
+      "grad_norm": 7.182160304427614,
+      "learning_rate": 8.916947949870409e-07,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 122976140,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6964468225816148,
+      "grad_norm": 0.8692665563349373,
+      "learning_rate": 8.910464513088615e-07,
+      "loss": 0.635,
+      "num_input_tokens_seen": 123039900,
+      "step": 5792
+    },
+    {
+      "epoch": 0.696567065472254,
+      "grad_norm": 2.078111162928809,
+      "learning_rate": 8.903982758580542e-07,
+      "loss": 0.785,
+      "num_input_tokens_seen": 123058560,
+      "step": 5793
+    },
+    {
+      "epoch": 0.696687308362893,
+      "grad_norm": 2.7022916338500855,
+      "learning_rate": 8.89750268732945e-07,
+      "loss": 0.801,
+      "num_input_tokens_seen": 123078080,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6968075512535321,
+      "grad_norm": 4.931016050874534,
+      "learning_rate": 8.891024300318382e-07,
+      "loss": 0.7961,
+      "num_input_tokens_seen": 123096370,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6969277941441713,
+      "grad_norm": 3.611687484505827,
+      "learning_rate": 8.884547598530103e-07,
+      "loss": 0.7502,
+      "num_input_tokens_seen": 123116660,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6970480370348103,
+      "grad_norm": 1.8958862359598438,
+      "learning_rate": 8.878072582947107e-07,
+      "loss": 0.7456,
+      "num_input_tokens_seen": 123134285,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6971682799254494,
+      "grad_norm": 3.378430523870369,
+      "learning_rate": 8.87159925455165e-07,
+      "loss": 0.7658,
+      "num_input_tokens_seen": 123153835,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6972885228160886,
+      "grad_norm": 2.3402846008513745,
+      "learning_rate": 8.865127614325745e-07,
+      "loss": 0.7276,
+      "num_input_tokens_seen": 123171985,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6974087657067276,
+      "grad_norm": 2.526308242551654,
+      "learning_rate": 8.85865766325113e-07,
+      "loss": 0.6703,
+      "num_input_tokens_seen": 123195635,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6975290085973667,
+      "grad_norm": 3.89629240584875,
+      "learning_rate": 8.852189402309287e-07,
+      "loss": 0.7201,
+      "num_input_tokens_seen": 123214540,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6976492514880057,
+      "grad_norm": 4.553674622989023,
+      "learning_rate": 8.845722832481441e-07,
+      "loss": 0.7328,
+      "num_input_tokens_seen": 123229690,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6977694943786449,
+      "grad_norm": 2.286421551144912,
+      "learning_rate": 8.83925795474858e-07,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 123249535,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6978897372692839,
+      "grad_norm": 2.5241828144253957,
+      "learning_rate": 8.832794770091414e-07,
+      "loss": 0.6059,
+      "num_input_tokens_seen": 123270090,
+      "step": 5804
+    },
+    {
+      "epoch": 0.698009980159923,
+      "grad_norm": 2.779768389048756,
+      "learning_rate": 8.826333279490401e-07,
+      "loss": 0.8236,
+      "num_input_tokens_seen": 123290445,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6981302230505622,
+      "grad_norm": 2.438945536551706,
+      "learning_rate": 8.81987348392574e-07,
+      "loss": 0.6807,
+      "num_input_tokens_seen": 123307285,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6982504659412012,
+      "grad_norm": 2.7875573101398365,
+      "learning_rate": 8.81341538437739e-07,
+      "loss": 0.7471,
+      "num_input_tokens_seen": 123325295,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6983707088318403,
+      "grad_norm": 2.4612091218673657,
+      "learning_rate": 8.80695898182503e-07,
+      "loss": 0.6812,
+      "num_input_tokens_seen": 123345995,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6984909517224794,
+      "grad_norm": 1.165436515940839,
+      "learning_rate": 8.800504277248093e-07,
+      "loss": 0.6885,
+      "num_input_tokens_seen": 123410465,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6986111946131185,
+      "grad_norm": 1.7971155093654674,
+      "learning_rate": 8.794051271625753e-07,
+      "loss": 0.7471,
+      "num_input_tokens_seen": 123427820,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6987314375037575,
+      "grad_norm": 1.99024624343518,
+      "learning_rate": 8.787599965936925e-07,
+      "loss": 0.8272,
+      "num_input_tokens_seen": 123448470,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6988516803943967,
+      "grad_norm": 1.6917874846028231,
+      "learning_rate": 8.781150361160268e-07,
+      "loss": 0.7109,
+      "num_input_tokens_seen": 123470100,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6989719232850358,
+      "grad_norm": 1.935236985201567,
+      "learning_rate": 8.774702458274181e-07,
+      "loss": 0.7289,
+      "num_input_tokens_seen": 123490225,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6990921661756748,
+      "grad_norm": 13.469183196271489,
+      "learning_rate": 8.768256258256799e-07,
+      "loss": 0.7016,
+      "num_input_tokens_seen": 123506570,
+      "step": 5814
+    },
+    {
+      "epoch": 0.699212409066314,
+      "grad_norm": 2.0798211833419034,
+      "learning_rate": 8.76181176208602e-07,
+      "loss": 0.7408,
+      "num_input_tokens_seen": 123524390,
+      "step": 5815
+    },
+    {
+      "epoch": 0.699332651956953,
+      "grad_norm": 2.0795534428333426,
+      "learning_rate": 8.755368970739461e-07,
+      "loss": 0.7288,
+      "num_input_tokens_seen": 123543470,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6994528948475921,
+      "grad_norm": 3.395109582403701,
+      "learning_rate": 8.748927885194489e-07,
+      "loss": 0.6144,
+      "num_input_tokens_seen": 123561495,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6995731377382313,
+      "grad_norm": 0.7578202913830686,
+      "learning_rate": 8.742488506428201e-07,
+      "loss": 0.5865,
+      "num_input_tokens_seen": 123620305,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6996933806288703,
+      "grad_norm": 2.1310537261469715,
+      "learning_rate": 8.736050835417466e-07,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 123640065,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6998136235195094,
+      "grad_norm": 2.843505119455081,
+      "learning_rate": 8.729614873138862e-07,
+      "loss": 0.6168,
+      "num_input_tokens_seen": 123657420,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6999338664101485,
+      "grad_norm": 2.4205749685245967,
+      "learning_rate": 8.723180620568722e-07,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 123676395,
+      "step": 5821
+    },
+    {
+      "epoch": 0.7000541093007876,
+      "grad_norm": 2.143577863912355,
+      "learning_rate": 8.716748078683107e-07,
+      "loss": 0.8562,
+      "num_input_tokens_seen": 123692890,
+      "step": 5822
+    },
+    {
+      "epoch": 0.7001743521914267,
+      "grad_norm": 2.453009840366571,
+      "learning_rate": 8.710317248457846e-07,
+      "loss": 0.685,
+      "num_input_tokens_seen": 123712225,
+      "step": 5823
+    },
+    {
+      "epoch": 0.7002945950820658,
+      "grad_norm": 2.295224754772024,
+      "learning_rate": 8.703888130868482e-07,
+      "loss": 0.7092,
+      "num_input_tokens_seen": 123733795,
+      "step": 5824
+    },
+    {
+      "epoch": 0.7004148379727049,
+      "grad_norm": 2.224199911024683,
+      "learning_rate": 8.697460726890307e-07,
+      "loss": 0.8128,
+      "num_input_tokens_seen": 123750660,
+      "step": 5825
+    },
+    {
+      "epoch": 0.7005350808633439,
+      "grad_norm": 2.4100870241860126,
+      "learning_rate": 8.691035037498354e-07,
+      "loss": 0.8991,
+      "num_input_tokens_seen": 123766370,
+      "step": 5826
+    },
+    {
+      "epoch": 0.7006553237539831,
+      "grad_norm": 3.264172559169138,
+      "learning_rate": 8.684611063667391e-07,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 123786555,
+      "step": 5827
+    },
+    {
+      "epoch": 0.7007755666446221,
+      "grad_norm": 2.9347790919562566,
+      "learning_rate": 8.678188806371935e-07,
+      "loss": 0.7581,
+      "num_input_tokens_seen": 123808310,
+      "step": 5828
+    },
+    {
+      "epoch": 0.7008958095352612,
+      "grad_norm": 2.7243321885056693,
+      "learning_rate": 8.671768266586234e-07,
+      "loss": 0.845,
+      "num_input_tokens_seen": 123826155,
+      "step": 5829
+    },
+    {
+      "epoch": 0.7010160524259004,
+      "grad_norm": 1.9201810822097873,
+      "learning_rate": 8.665349445284275e-07,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 123845615,
+      "step": 5830
+    },
+    {
+      "epoch": 0.7011362953165394,
+      "grad_norm": 1.5057307526088324,
+      "learning_rate": 8.658932343439799e-07,
+      "loss": 0.8041,
+      "num_input_tokens_seen": 123865120,
+      "step": 5831
+    },
+    {
+      "epoch": 0.7012565382071785,
+      "grad_norm": 2.1458664815422135,
+      "learning_rate": 8.65251696202627e-07,
+      "loss": 0.7627,
+      "num_input_tokens_seen": 123881220,
+      "step": 5832
+    },
+    {
+      "epoch": 0.7013767810978175,
+      "grad_norm": 3.8215086064687727,
+      "learning_rate": 8.646103302016896e-07,
+      "loss": 0.8735,
+      "num_input_tokens_seen": 123899910,
+      "step": 5833
+    },
+    {
+      "epoch": 0.7014970239884567,
+      "grad_norm": 1.8775991210681255,
+      "learning_rate": 8.639691364384614e-07,
+      "loss": 0.8826,
+      "num_input_tokens_seen": 123917255,
+      "step": 5834
+    },
+    {
+      "epoch": 0.7016172668790958,
+      "grad_norm": 2.199387815460875,
+      "learning_rate": 8.63328115010213e-07,
+      "loss": 0.7205,
+      "num_input_tokens_seen": 123933825,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7017375097697348,
+      "grad_norm": 2.7935219445064425,
+      "learning_rate": 8.626872660141855e-07,
+      "loss": 0.6746,
+      "num_input_tokens_seen": 123951455,
+      "step": 5836
+    },
+    {
+      "epoch": 0.701857752660374,
+      "grad_norm": 1.8000102053350795,
+      "learning_rate": 8.620465895475957e-07,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 123969395,
+      "step": 5837
+    },
+    {
+      "epoch": 0.701977995551013,
+      "grad_norm": 1.6164830658483664,
+      "learning_rate": 8.614060857076326e-07,
+      "loss": 0.7487,
+      "num_input_tokens_seen": 123989785,
+      "step": 5838
+    },
+    {
+      "epoch": 0.7020982384416521,
+      "grad_norm": 2.043937670169765,
+      "learning_rate": 8.607657545914626e-07,
+      "loss": 0.7384,
+      "num_input_tokens_seen": 124009200,
+      "step": 5839
+    },
+    {
+      "epoch": 0.7022184813322913,
+      "grad_norm": 2.149622226934407,
+      "learning_rate": 8.601255962962211e-07,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 124027930,
+      "step": 5840
+    },
+    {
+      "epoch": 0.7023387242229303,
+      "grad_norm": 2.7498104005409623,
+      "learning_rate": 8.594856109190203e-07,
+      "loss": 0.7205,
+      "num_input_tokens_seen": 124044680,
+      "step": 5841
+    },
+    {
+      "epoch": 0.7024589671135694,
+      "grad_norm": 3.8107181027584915,
+      "learning_rate": 8.588457985569446e-07,
+      "loss": 0.6908,
+      "num_input_tokens_seen": 124067310,
+      "step": 5842
+    },
+    {
+      "epoch": 0.7025792100042085,
+      "grad_norm": 2.4042925662751626,
+      "learning_rate": 8.582061593070548e-07,
+      "loss": 0.7062,
+      "num_input_tokens_seen": 124087760,
+      "step": 5843
+    },
+    {
+      "epoch": 0.7026994528948476,
+      "grad_norm": 2.2601402006842264,
+      "learning_rate": 8.57566693266383e-07,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 124105170,
+      "step": 5844
+    },
+    {
+      "epoch": 0.7028196957854866,
+      "grad_norm": 2.5387976499141582,
+      "learning_rate": 8.569274005319354e-07,
+      "loss": 0.694,
+      "num_input_tokens_seen": 124123290,
+      "step": 5845
+    },
+    {
+      "epoch": 0.7029399386761258,
+      "grad_norm": 2.2072677006018595,
+      "learning_rate": 8.562882812006913e-07,
+      "loss": 0.7957,
+      "num_input_tokens_seen": 124140500,
+      "step": 5846
+    },
+    {
+      "epoch": 0.7030601815667649,
+      "grad_norm": 2.0332175134146375,
+      "learning_rate": 8.556493353696066e-07,
+      "loss": 0.7639,
+      "num_input_tokens_seen": 124159220,
+      "step": 5847
+    },
+    {
+      "epoch": 0.7031804244574039,
+      "grad_norm": 2.8633720301501513,
+      "learning_rate": 8.550105631356077e-07,
+      "loss": 0.6747,
+      "num_input_tokens_seen": 124178665,
+      "step": 5848
+    },
+    {
+      "epoch": 0.7033006673480431,
+      "grad_norm": 2.3808748099815484,
+      "learning_rate": 8.543719645955961e-07,
+      "loss": 0.7597,
+      "num_input_tokens_seen": 124196715,
+      "step": 5849
+    },
+    {
+      "epoch": 0.7034209102386821,
+      "grad_norm": 1.6396057514756623,
+      "learning_rate": 8.537335398464458e-07,
+      "loss": 0.74,
+      "num_input_tokens_seen": 124216755,
+      "step": 5850
+    },
+    {
+      "epoch": 0.7035411531293212,
+      "grad_norm": 3.390915774377116,
+      "learning_rate": 8.53095288985007e-07,
+      "loss": 0.849,
+      "num_input_tokens_seen": 124230210,
+      "step": 5851
+    },
+    {
+      "epoch": 0.7036613960199604,
+      "grad_norm": 2.326238211452944,
+      "learning_rate": 8.524572121081009e-07,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 124250030,
+      "step": 5852
+    },
+    {
+      "epoch": 0.7037816389105994,
+      "grad_norm": 2.9606863369593506,
+      "learning_rate": 8.518193093125232e-07,
+      "loss": 0.6237,
+      "num_input_tokens_seen": 124268805,
+      "step": 5853
+    },
+    {
+      "epoch": 0.7039018818012385,
+      "grad_norm": 1.6307357093683321,
+      "learning_rate": 8.511815806950436e-07,
+      "loss": 0.7935,
+      "num_input_tokens_seen": 124289555,
+      "step": 5854
+    },
+    {
+      "epoch": 0.7040221246918776,
+      "grad_norm": 1.8601970997053092,
+      "learning_rate": 8.505440263524044e-07,
+      "loss": 0.771,
+      "num_input_tokens_seen": 124308120,
+      "step": 5855
+    },
+    {
+      "epoch": 0.7041423675825167,
+      "grad_norm": 4.216998523722609,
+      "learning_rate": 8.499066463813227e-07,
+      "loss": 0.8699,
+      "num_input_tokens_seen": 124320675,
+      "step": 5856
+    },
+    {
+      "epoch": 0.7042626104731557,
+      "grad_norm": 1.9165562234346236,
+      "learning_rate": 8.492694408784884e-07,
+      "loss": 0.712,
+      "num_input_tokens_seen": 124340650,
+      "step": 5857
+    },
+    {
+      "epoch": 0.7043828533637949,
+      "grad_norm": 7.135690130874974,
+      "learning_rate": 8.486324099405642e-07,
+      "loss": 0.6191,
+      "num_input_tokens_seen": 124357215,
+      "step": 5858
+    },
+    {
+      "epoch": 0.704503096254434,
+      "grad_norm": 1.7074100395502942,
+      "learning_rate": 8.479955536641887e-07,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 124378430,
+      "step": 5859
+    },
+    {
+      "epoch": 0.704623339145073,
+      "grad_norm": 2.4059061847083765,
+      "learning_rate": 8.473588721459716e-07,
+      "loss": 0.6517,
+      "num_input_tokens_seen": 124398060,
+      "step": 5860
+    },
+    {
+      "epoch": 0.7047435820357122,
+      "grad_norm": 2.929225172075417,
+      "learning_rate": 8.467223654824975e-07,
+      "loss": 0.7057,
+      "num_input_tokens_seen": 124417235,
+      "step": 5861
+    },
+    {
+      "epoch": 0.7048638249263512,
+      "grad_norm": 2.2320970155392987,
+      "learning_rate": 8.460860337703227e-07,
+      "loss": 0.6269,
+      "num_input_tokens_seen": 124437560,
+      "step": 5862
+    },
+    {
+      "epoch": 0.7049840678169903,
+      "grad_norm": 2.441890084258046,
+      "learning_rate": 8.454498771059797e-07,
+      "loss": 0.7073,
+      "num_input_tokens_seen": 124456655,
+      "step": 5863
+    },
+    {
+      "epoch": 0.7051043107076294,
+      "grad_norm": 2.3044616157977336,
+      "learning_rate": 8.448138955859725e-07,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 124472960,
+      "step": 5864
+    },
+    {
+      "epoch": 0.7052245535982685,
+      "grad_norm": 2.863272245794222,
+      "learning_rate": 8.44178089306779e-07,
+      "loss": 0.8954,
+      "num_input_tokens_seen": 124490615,
+      "step": 5865
+    },
+    {
+      "epoch": 0.7053447964889076,
+      "grad_norm": 1.9999592525800536,
+      "learning_rate": 8.435424583648494e-07,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 124508780,
+      "step": 5866
+    },
+    {
+      "epoch": 0.7054650393795466,
+      "grad_norm": 5.116959911996318,
+      "learning_rate": 8.429070028566101e-07,
+      "loss": 0.7216,
+      "num_input_tokens_seen": 124529810,
+      "step": 5867
+    },
+    {
+      "epoch": 0.7055852822701858,
+      "grad_norm": 2.0174898039859763,
+      "learning_rate": 8.422717228784586e-07,
+      "loss": 0.7439,
+      "num_input_tokens_seen": 124546405,
+      "step": 5868
+    },
+    {
+      "epoch": 0.7057055251608249,
+      "grad_norm": 1.9496718860492182,
+      "learning_rate": 8.416366185267663e-07,
+      "loss": 0.685,
+      "num_input_tokens_seen": 124563625,
+      "step": 5869
+    },
+    {
+      "epoch": 0.7058257680514639,
+      "grad_norm": 2.4153914601803,
+      "learning_rate": 8.410016898978778e-07,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 124580820,
+      "step": 5870
+    },
+    {
+      "epoch": 0.7059460109421031,
+      "grad_norm": 1.8264799683224753,
+      "learning_rate": 8.403669370881115e-07,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 124599275,
+      "step": 5871
+    },
+    {
+      "epoch": 0.7060662538327421,
+      "grad_norm": 1.702911751238476,
+      "learning_rate": 8.397323601937587e-07,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 124618895,
+      "step": 5872
+    },
+    {
+      "epoch": 0.7061864967233812,
+      "grad_norm": 2.7795835997745932,
+      "learning_rate": 8.390979593110845e-07,
+      "loss": 0.7669,
+      "num_input_tokens_seen": 124640745,
+      "step": 5873
+    },
+    {
+      "epoch": 0.7063067396140204,
+      "grad_norm": 2.628525904524467,
+      "learning_rate": 8.384637345363262e-07,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 124659655,
+      "step": 5874
+    },
+    {
+      "epoch": 0.7064269825046594,
+      "grad_norm": 2.2233804096283962,
+      "learning_rate": 8.378296859656964e-07,
+      "loss": 0.7635,
+      "num_input_tokens_seen": 124680530,
+      "step": 5875
+    },
+    {
+      "epoch": 0.7065472253952985,
+      "grad_norm": 2.4898693497953763,
+      "learning_rate": 8.371958136953792e-07,
+      "loss": 0.6727,
+      "num_input_tokens_seen": 124700280,
+      "step": 5876
+    },
+    {
+      "epoch": 0.7066674682859376,
+      "grad_norm": 2.9232944411278474,
+      "learning_rate": 8.365621178215326e-07,
+      "loss": 0.653,
+      "num_input_tokens_seen": 124716470,
+      "step": 5877
+    },
+    {
+      "epoch": 0.7067877111765767,
+      "grad_norm": 2.689336504326027,
+      "learning_rate": 8.359285984402871e-07,
+      "loss": 0.7455,
+      "num_input_tokens_seen": 124733455,
+      "step": 5878
+    },
+    {
+      "epoch": 0.7069079540672157,
+      "grad_norm": 2.156907680758512,
+      "learning_rate": 8.352952556477481e-07,
+      "loss": 0.737,
+      "num_input_tokens_seen": 124751085,
+      "step": 5879
+    },
+    {
+      "epoch": 0.7070281969578549,
+      "grad_norm": 2.0682501183752624,
+      "learning_rate": 8.34662089539993e-07,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 124770315,
+      "step": 5880
+    },
+    {
+      "epoch": 0.707148439848494,
+      "grad_norm": 3.187723238264632,
+      "learning_rate": 8.340291002130722e-07,
+      "loss": 0.786,
+      "num_input_tokens_seen": 124789225,
+      "step": 5881
+    },
+    {
+      "epoch": 0.707268682739133,
+      "grad_norm": 3.5202132722902872,
+      "learning_rate": 8.3339628776301e-07,
+      "loss": 0.7909,
+      "num_input_tokens_seen": 124807085,
+      "step": 5882
+    },
+    {
+      "epoch": 0.7073889256297722,
+      "grad_norm": 3.5763798492428855,
+      "learning_rate": 8.327636522858033e-07,
+      "loss": 0.5628,
+      "num_input_tokens_seen": 124826410,
+      "step": 5883
+    },
+    {
+      "epoch": 0.7075091685204112,
+      "grad_norm": 2.28302761732246,
+      "learning_rate": 8.321311938774225e-07,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 124845220,
+      "step": 5884
+    },
+    {
+      "epoch": 0.7076294114110503,
+      "grad_norm": 3.099892554351801,
+      "learning_rate": 8.31498912633811e-07,
+      "loss": 0.7839,
+      "num_input_tokens_seen": 124864950,
+      "step": 5885
+    },
+    {
+      "epoch": 0.7077496543016895,
+      "grad_norm": 1.8951283269948798,
+      "learning_rate": 8.308668086508847e-07,
+      "loss": 0.8389,
+      "num_input_tokens_seen": 124882750,
+      "step": 5886
+    },
+    {
+      "epoch": 0.7078698971923285,
+      "grad_norm": 2.189770710557123,
+      "learning_rate": 8.302348820245349e-07,
+      "loss": 0.7355,
+      "num_input_tokens_seen": 124905035,
+      "step": 5887
+    },
+    {
+      "epoch": 0.7079901400829676,
+      "grad_norm": 2.958003808312968,
+      "learning_rate": 8.296031328506232e-07,
+      "loss": 0.7008,
+      "num_input_tokens_seen": 124924505,
+      "step": 5888
+    },
+    {
+      "epoch": 0.7081103829736067,
+      "grad_norm": 2.4783905215207422,
+      "learning_rate": 8.289715612249857e-07,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 124944840,
+      "step": 5889
+    },
+    {
+      "epoch": 0.7082306258642458,
+      "grad_norm": 5.134662107702385,
+      "learning_rate": 8.283401672434305e-07,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 124959785,
+      "step": 5890
+    },
+    {
+      "epoch": 0.7083508687548848,
+      "grad_norm": 2.0960331757124306,
+      "learning_rate": 8.277089510017412e-07,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 124980310,
+      "step": 5891
+    },
+    {
+      "epoch": 0.708471111645524,
+      "grad_norm": 2.045932090879382,
+      "learning_rate": 8.270779125956719e-07,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 125000410,
+      "step": 5892
+    },
+    {
+      "epoch": 0.7085913545361631,
+      "grad_norm": 2.334997301712223,
+      "learning_rate": 8.264470521209505e-07,
+      "loss": 0.7908,
+      "num_input_tokens_seen": 125018495,
+      "step": 5893
+    },
+    {
+      "epoch": 0.7087115974268021,
+      "grad_norm": 3.274519333048488,
+      "learning_rate": 8.258163696732779e-07,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 125035805,
+      "step": 5894
+    },
+    {
+      "epoch": 0.7088318403174413,
+      "grad_norm": 2.4864310177079334,
+      "learning_rate": 8.251858653483288e-07,
+      "loss": 0.7667,
+      "num_input_tokens_seen": 125053690,
+      "step": 5895
+    },
+    {
+      "epoch": 0.7089520832080803,
+      "grad_norm": 2.236903660523214,
+      "learning_rate": 8.245555392417501e-07,
+      "loss": 0.8505,
+      "num_input_tokens_seen": 125068065,
+      "step": 5896
+    },
+    {
+      "epoch": 0.7090723260987194,
+      "grad_norm": 2.1861391571755275,
+      "learning_rate": 8.239253914491613e-07,
+      "loss": 0.7866,
+      "num_input_tokens_seen": 125086110,
+      "step": 5897
+    },
+    {
+      "epoch": 0.7091925689893585,
+      "grad_norm": 2.0798696769264153,
+      "learning_rate": 8.232954220661556e-07,
+      "loss": 0.7436,
+      "num_input_tokens_seen": 125108565,
+      "step": 5898
+    },
+    {
+      "epoch": 0.7093128118799976,
+      "grad_norm": 3.2558864905838205,
+      "learning_rate": 8.226656311882989e-07,
+      "loss": 0.6879,
+      "num_input_tokens_seen": 125127595,
+      "step": 5899
+    },
+    {
+      "epoch": 0.7094330547706367,
+      "grad_norm": 6.148354324078775,
+      "learning_rate": 8.220360189111298e-07,
+      "loss": 0.7619,
+      "num_input_tokens_seen": 125145345,
+      "step": 5900
+    },
+    {
+      "epoch": 0.7095532976612757,
+      "grad_norm": 3.6400111607903662,
+      "learning_rate": 8.214065853301599e-07,
+      "loss": 0.7953,
+      "num_input_tokens_seen": 125160595,
+      "step": 5901
+    },
+    {
+      "epoch": 0.7096735405519149,
+      "grad_norm": 0.8190312990669477,
+      "learning_rate": 8.207773305408734e-07,
+      "loss": 0.6026,
+      "num_input_tokens_seen": 125227535,
+      "step": 5902
+    },
+    {
+      "epoch": 0.709793783442554,
+      "grad_norm": 2.3134780933985466,
+      "learning_rate": 8.201482546387288e-07,
+      "loss": 0.7894,
+      "num_input_tokens_seen": 125246730,
+      "step": 5903
+    },
+    {
+      "epoch": 0.709914026333193,
+      "grad_norm": 1.8639194181343757,
+      "learning_rate": 8.195193577191559e-07,
+      "loss": 0.9075,
+      "num_input_tokens_seen": 125268280,
+      "step": 5904
+    },
+    {
+      "epoch": 0.7100342692238322,
+      "grad_norm": 1.917515582073336,
+      "learning_rate": 8.188906398775579e-07,
+      "loss": 0.8346,
+      "num_input_tokens_seen": 125288545,
+      "step": 5905
+    },
+    {
+      "epoch": 0.7101545121144712,
+      "grad_norm": 2.0250249711543935,
+      "learning_rate": 8.1826210120931e-07,
+      "loss": 0.6875,
+      "num_input_tokens_seen": 125307475,
+      "step": 5906
+    },
+    {
+      "epoch": 0.7102747550051103,
+      "grad_norm": 4.702882089338399,
+      "learning_rate": 8.176337418097626e-07,
+      "loss": 0.6945,
+      "num_input_tokens_seen": 125327665,
+      "step": 5907
+    },
+    {
+      "epoch": 0.7103949978957494,
+      "grad_norm": 2.2365564054696176,
+      "learning_rate": 8.170055617742364e-07,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 125344665,
+      "step": 5908
+    },
+    {
+      "epoch": 0.7105152407863885,
+      "grad_norm": 2.369596687233931,
+      "learning_rate": 8.163775611980259e-07,
+      "loss": 0.7049,
+      "num_input_tokens_seen": 125363495,
+      "step": 5909
+    },
+    {
+      "epoch": 0.7106354836770276,
+      "grad_norm": 3.2606317493847694,
+      "learning_rate": 8.157497401763976e-07,
+      "loss": 0.7844,
+      "num_input_tokens_seen": 125380880,
+      "step": 5910
+    },
+    {
+      "epoch": 0.7107557265676667,
+      "grad_norm": 2.2332964344941524,
+      "learning_rate": 8.151220988045928e-07,
+      "loss": 0.7738,
+      "num_input_tokens_seen": 125399855,
+      "step": 5911
+    },
+    {
+      "epoch": 0.7108759694583058,
+      "grad_norm": 2.269868914998486,
+      "learning_rate": 8.144946371778234e-07,
+      "loss": 0.8249,
+      "num_input_tokens_seen": 125419685,
+      "step": 5912
+    },
+    {
+      "epoch": 0.7109962123489448,
+      "grad_norm": 2.1570535202995544,
+      "learning_rate": 8.138673553912751e-07,
+      "loss": 0.771,
+      "num_input_tokens_seen": 125439965,
+      "step": 5913
+    },
+    {
+      "epoch": 0.711116455239584,
+      "grad_norm": 3.1449954650966006,
+      "learning_rate": 8.132402535401059e-07,
+      "loss": 0.5656,
+      "num_input_tokens_seen": 125460940,
+      "step": 5914
+    },
+    {
+      "epoch": 0.711236698130223,
+      "grad_norm": 1.8001175951779036,
+      "learning_rate": 8.126133317194465e-07,
+      "loss": 0.7433,
+      "num_input_tokens_seen": 125480850,
+      "step": 5915
+    },
+    {
+      "epoch": 0.7113569410208621,
+      "grad_norm": 2.8179978075944367,
+      "learning_rate": 8.11986590024401e-07,
+      "loss": 0.7325,
+      "num_input_tokens_seen": 125500310,
+      "step": 5916
+    },
+    {
+      "epoch": 0.7114771839115013,
+      "grad_norm": 1.9585462457148632,
+      "learning_rate": 8.113600285500448e-07,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 125520240,
+      "step": 5917
+    },
+    {
+      "epoch": 0.7115974268021403,
+      "grad_norm": 2.0153665071086633,
+      "learning_rate": 8.107336473914268e-07,
+      "loss": 0.733,
+      "num_input_tokens_seen": 125538590,
+      "step": 5918
+    },
+    {
+      "epoch": 0.7117176696927794,
+      "grad_norm": 0.9830103952212915,
+      "learning_rate": 8.101074466435694e-07,
+      "loss": 0.6027,
+      "num_input_tokens_seen": 125597785,
+      "step": 5919
+    },
+    {
+      "epoch": 0.7118379125834186,
+      "grad_norm": 2.755434164992065,
+      "learning_rate": 8.094814264014662e-07,
+      "loss": 0.6719,
+      "num_input_tokens_seen": 125616260,
+      "step": 5920
+    },
+    {
+      "epoch": 0.7119581554740576,
+      "grad_norm": 2.642542072806552,
+      "learning_rate": 8.088555867600844e-07,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 125632145,
+      "step": 5921
+    },
+    {
+      "epoch": 0.7120783983646967,
+      "grad_norm": 2.2963964485546855,
+      "learning_rate": 8.08229927814362e-07,
+      "loss": 0.6014,
+      "num_input_tokens_seen": 125654755,
+      "step": 5922
+    },
+    {
+      "epoch": 0.7121986412553358,
+      "grad_norm": 1.8868883795227693,
+      "learning_rate": 8.076044496592127e-07,
+      "loss": 0.6429,
+      "num_input_tokens_seen": 125676325,
+      "step": 5923
+    },
+    {
+      "epoch": 0.7123188841459749,
+      "grad_norm": 2.3393160007064018,
+      "learning_rate": 8.069791523895204e-07,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 125692495,
+      "step": 5924
+    },
+    {
+      "epoch": 0.7124391270366139,
+      "grad_norm": 2.1264029617422993,
+      "learning_rate": 8.063540361001422e-07,
+      "loss": 0.7751,
+      "num_input_tokens_seen": 125710785,
+      "step": 5925
+    },
+    {
+      "epoch": 0.7125593699272531,
+      "grad_norm": 4.318504675685387,
+      "learning_rate": 8.057291008859075e-07,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 125728665,
+      "step": 5926
+    },
+    {
+      "epoch": 0.7126796128178922,
+      "grad_norm": 2.0216689069860263,
+      "learning_rate": 8.051043468416187e-07,
+      "loss": 0.6735,
+      "num_input_tokens_seen": 125749635,
+      "step": 5927
+    },
+    {
+      "epoch": 0.7127998557085312,
+      "grad_norm": 3.190755346379996,
+      "learning_rate": 8.044797740620506e-07,
+      "loss": 0.8195,
+      "num_input_tokens_seen": 125767960,
+      "step": 5928
+    },
+    {
+      "epoch": 0.7129200985991703,
+      "grad_norm": 2.2690636682016407,
+      "learning_rate": 8.0385538264195e-07,
+      "loss": 0.7793,
+      "num_input_tokens_seen": 125786390,
+      "step": 5929
+    },
+    {
+      "epoch": 0.7130403414898094,
+      "grad_norm": 1.9017918148792083,
+      "learning_rate": 8.032311726760364e-07,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 125807330,
+      "step": 5930
+    },
+    {
+      "epoch": 0.7131605843804485,
+      "grad_norm": 3.0389515470673842,
+      "learning_rate": 8.026071442590028e-07,
+      "loss": 0.6874,
+      "num_input_tokens_seen": 125833980,
+      "step": 5931
+    },
+    {
+      "epoch": 0.7132808272710875,
+      "grad_norm": 2.2587520762098476,
+      "learning_rate": 8.019832974855134e-07,
+      "loss": 0.8127,
+      "num_input_tokens_seen": 125851660,
+      "step": 5932
+    },
+    {
+      "epoch": 0.7134010701617267,
+      "grad_norm": 3.210069659489253,
+      "learning_rate": 8.013596324502052e-07,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 125869845,
+      "step": 5933
+    },
+    {
+      "epoch": 0.7135213130523658,
+      "grad_norm": 7.198537849727589,
+      "learning_rate": 8.007361492476872e-07,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 125888890,
+      "step": 5934
+    },
+    {
+      "epoch": 0.7136415559430048,
+      "grad_norm": 1.5668746728659018,
+      "learning_rate": 8.001128479725426e-07,
+      "loss": 0.78,
+      "num_input_tokens_seen": 125910515,
+      "step": 5935
+    },
+    {
+      "epoch": 0.713761798833644,
+      "grad_norm": 2.437616022721662,
+      "learning_rate": 7.994897287193248e-07,
+      "loss": 0.8013,
+      "num_input_tokens_seen": 125929615,
+      "step": 5936
+    },
+    {
+      "epoch": 0.713882041724283,
+      "grad_norm": 2.3862970082853283,
+      "learning_rate": 7.988667915825605e-07,
+      "loss": 0.839,
+      "num_input_tokens_seen": 125946400,
+      "step": 5937
+    },
+    {
+      "epoch": 0.7140022846149221,
+      "grad_norm": 3.3095438249269162,
+      "learning_rate": 7.982440366567485e-07,
+      "loss": 0.7516,
+      "num_input_tokens_seen": 125964610,
+      "step": 5938
+    },
+    {
+      "epoch": 0.7141225275055613,
+      "grad_norm": 1.8872712834661791,
+      "learning_rate": 7.97621464036361e-07,
+      "loss": 0.7497,
+      "num_input_tokens_seen": 125986090,
+      "step": 5939
+    },
+    {
+      "epoch": 0.7142427703962003,
+      "grad_norm": 2.7646027532956055,
+      "learning_rate": 7.969990738158417e-07,
+      "loss": 0.6787,
+      "num_input_tokens_seen": 126004220,
+      "step": 5940
+    },
+    {
+      "epoch": 0.7143630132868394,
+      "grad_norm": 2.454530292894201,
+      "learning_rate": 7.963768660896062e-07,
+      "loss": 0.8344,
+      "num_input_tokens_seen": 126022350,
+      "step": 5941
+    },
+    {
+      "epoch": 0.7144832561774785,
+      "grad_norm": 2.478187868166997,
+      "learning_rate": 7.957548409520432e-07,
+      "loss": 0.8194,
+      "num_input_tokens_seen": 126041295,
+      "step": 5942
+    },
+    {
+      "epoch": 0.7146034990681176,
+      "grad_norm": 2.0563859570904963,
+      "learning_rate": 7.951329984975135e-07,
+      "loss": 0.8405,
+      "num_input_tokens_seen": 126057955,
+      "step": 5943
+    },
+    {
+      "epoch": 0.7147237419587567,
+      "grad_norm": 0.7367238794582575,
+      "learning_rate": 7.945113388203497e-07,
+      "loss": 0.5617,
+      "num_input_tokens_seen": 126119980,
+      "step": 5944
+    },
+    {
+      "epoch": 0.7148439848493958,
+      "grad_norm": 3.2841200464990017,
+      "learning_rate": 7.938898620148575e-07,
+      "loss": 0.7807,
+      "num_input_tokens_seen": 126137460,
+      "step": 5945
+    },
+    {
+      "epoch": 0.7149642277400349,
+      "grad_norm": 2.6035488552783077,
+      "learning_rate": 7.932685681753135e-07,
+      "loss": 0.7029,
+      "num_input_tokens_seen": 126154460,
+      "step": 5946
+    },
+    {
+      "epoch": 0.7150844706306739,
+      "grad_norm": 2.566913677399377,
+      "learning_rate": 7.92647457395969e-07,
+      "loss": 0.6211,
+      "num_input_tokens_seen": 126176005,
+      "step": 5947
+    },
+    {
+      "epoch": 0.7152047135213131,
+      "grad_norm": 3.6668935326290213,
+      "learning_rate": 7.920265297710451e-07,
+      "loss": 0.735,
+      "num_input_tokens_seen": 126193115,
+      "step": 5948
+    },
+    {
+      "epoch": 0.7153249564119522,
+      "grad_norm": 1.9827154045629922,
+      "learning_rate": 7.914057853947363e-07,
+      "loss": 0.7173,
+      "num_input_tokens_seen": 126212015,
+      "step": 5949
+    },
+    {
+      "epoch": 0.7154451993025912,
+      "grad_norm": 1.8503598201708058,
+      "learning_rate": 7.907852243612083e-07,
+      "loss": 0.6291,
+      "num_input_tokens_seen": 126232140,
+      "step": 5950
+    },
+    {
+      "epoch": 0.7155654421932304,
+      "grad_norm": 2.2508461389296377,
+      "learning_rate": 7.901648467646009e-07,
+      "loss": 0.7137,
+      "num_input_tokens_seen": 126250800,
+      "step": 5951
+    },
+    {
+      "epoch": 0.7156856850838694,
+      "grad_norm": 1.7510362527105234,
+      "learning_rate": 7.895446526990244e-07,
+      "loss": 0.7187,
+      "num_input_tokens_seen": 126270535,
+      "step": 5952
+    },
+    {
+      "epoch": 0.7158059279745085,
+      "grad_norm": 1.6663469947012228,
+      "learning_rate": 7.889246422585616e-07,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 126289640,
+      "step": 5953
+    },
+    {
+      "epoch": 0.7159261708651476,
+      "grad_norm": 1.8898179682963823,
+      "learning_rate": 7.883048155372669e-07,
+      "loss": 0.7327,
+      "num_input_tokens_seen": 126307875,
+      "step": 5954
+    },
+    {
+      "epoch": 0.7160464137557867,
+      "grad_norm": 2.5333450951029226,
+      "learning_rate": 7.876851726291691e-07,
+      "loss": 0.7047,
+      "num_input_tokens_seen": 126325895,
+      "step": 5955
+    },
+    {
+      "epoch": 0.7161666566464258,
+      "grad_norm": 2.4790833965747487,
+      "learning_rate": 7.870657136282666e-07,
+      "loss": 0.7807,
+      "num_input_tokens_seen": 126344475,
+      "step": 5956
+    },
+    {
+      "epoch": 0.7162868995370649,
+      "grad_norm": 1.7677675413444143,
+      "learning_rate": 7.86446438628531e-07,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 126365265,
+      "step": 5957
+    },
+    {
+      "epoch": 0.716407142427704,
+      "grad_norm": 0.8362269586633847,
+      "learning_rate": 7.858273477239059e-07,
+      "loss": 0.5977,
+      "num_input_tokens_seen": 126433405,
+      "step": 5958
+    },
+    {
+      "epoch": 0.716527385318343,
+      "grad_norm": 1.9054625662871856,
+      "learning_rate": 7.852084410083067e-07,
+      "loss": 0.713,
+      "num_input_tokens_seen": 126451945,
+      "step": 5959
+    },
+    {
+      "epoch": 0.7166476282089821,
+      "grad_norm": 1.7789717298763061,
+      "learning_rate": 7.84589718575621e-07,
+      "loss": 0.628,
+      "num_input_tokens_seen": 126472110,
+      "step": 5960
+    },
+    {
+      "epoch": 0.7167678710996213,
+      "grad_norm": 2.333946754621867,
+      "learning_rate": 7.839711805197087e-07,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 126490685,
+      "step": 5961
+    },
+    {
+      "epoch": 0.7168881139902603,
+      "grad_norm": 2.935143616386971,
+      "learning_rate": 7.833528269344008e-07,
+      "loss": 0.7433,
+      "num_input_tokens_seen": 126510310,
+      "step": 5962
+    },
+    {
+      "epoch": 0.7170083568808994,
+      "grad_norm": 3.432118378663752,
+      "learning_rate": 7.827346579135023e-07,
+      "loss": 0.7725,
+      "num_input_tokens_seen": 126527370,
+      "step": 5963
+    },
+    {
+      "epoch": 0.7171285997715385,
+      "grad_norm": 2.7708960766688504,
+      "learning_rate": 7.821166735507885e-07,
+      "loss": 0.8279,
+      "num_input_tokens_seen": 126546120,
+      "step": 5964
+    },
+    {
+      "epoch": 0.7172488426621776,
+      "grad_norm": 44.32332475648218,
+      "learning_rate": 7.81498873940007e-07,
+      "loss": 0.6812,
+      "num_input_tokens_seen": 126563055,
+      "step": 5965
+    },
+    {
+      "epoch": 0.7173690855528166,
+      "grad_norm": 10.181157326224408,
+      "learning_rate": 7.808812591748768e-07,
+      "loss": 0.7645,
+      "num_input_tokens_seen": 126583155,
+      "step": 5966
+    },
+    {
+      "epoch": 0.7174893284434558,
+      "grad_norm": 2.751923502991377,
+      "learning_rate": 7.802638293490908e-07,
+      "loss": 0.649,
+      "num_input_tokens_seen": 126602520,
+      "step": 5967
+    },
+    {
+      "epoch": 0.7176095713340949,
+      "grad_norm": 2.501735916087633,
+      "learning_rate": 7.796465845563123e-07,
+      "loss": 0.7735,
+      "num_input_tokens_seen": 126621115,
+      "step": 5968
+    },
+    {
+      "epoch": 0.7177298142247339,
+      "grad_norm": 2.250301904435027,
+      "learning_rate": 7.790295248901766e-07,
+      "loss": 0.7968,
+      "num_input_tokens_seen": 126641965,
+      "step": 5969
+    },
+    {
+      "epoch": 0.7178500571153731,
+      "grad_norm": 2.0454599439646612,
+      "learning_rate": 7.78412650444291e-07,
+      "loss": 0.6241,
+      "num_input_tokens_seen": 126664915,
+      "step": 5970
+    },
+    {
+      "epoch": 0.7179703000060121,
+      "grad_norm": 1.939846278150598,
+      "learning_rate": 7.777959613122351e-07,
+      "loss": 0.6713,
+      "num_input_tokens_seen": 126684460,
+      "step": 5971
+    },
+    {
+      "epoch": 0.7180905428966512,
+      "grad_norm": 1.9554624487795176,
+      "learning_rate": 7.771794575875604e-07,
+      "loss": 0.773,
+      "num_input_tokens_seen": 126706050,
+      "step": 5972
+    },
+    {
+      "epoch": 0.7182107857872904,
+      "grad_norm": 3.1908312814053374,
+      "learning_rate": 7.765631393637894e-07,
+      "loss": 0.7719,
+      "num_input_tokens_seen": 126723965,
+      "step": 5973
+    },
+    {
+      "epoch": 0.7183310286779294,
+      "grad_norm": 3.0188150875785333,
+      "learning_rate": 7.75947006734417e-07,
+      "loss": 0.4717,
+      "num_input_tokens_seen": 126741465,
+      "step": 5974
+    },
+    {
+      "epoch": 0.7184512715685685,
+      "grad_norm": 2.252119704361162,
+      "learning_rate": 7.753310597929108e-07,
+      "loss": 0.821,
+      "num_input_tokens_seen": 126757825,
+      "step": 5975
+    },
+    {
+      "epoch": 0.7185715144592076,
+      "grad_norm": 0.8101836214369763,
+      "learning_rate": 7.747152986327095e-07,
+      "loss": 0.5865,
+      "num_input_tokens_seen": 126818090,
+      "step": 5976
+    },
+    {
+      "epoch": 0.7186917573498467,
+      "grad_norm": 2.0468655191869036,
+      "learning_rate": 7.740997233472228e-07,
+      "loss": 0.6764,
+      "num_input_tokens_seen": 126835430,
+      "step": 5977
+    },
+    {
+      "epoch": 0.7188120002404857,
+      "grad_norm": 2.5491412583530177,
+      "learning_rate": 7.734843340298329e-07,
+      "loss": 0.7033,
+      "num_input_tokens_seen": 126854975,
+      "step": 5978
+    },
+    {
+      "epoch": 0.7189322431311249,
+      "grad_norm": 2.2320724391262643,
+      "learning_rate": 7.72869130773895e-07,
+      "loss": 0.7556,
+      "num_input_tokens_seen": 126875295,
+      "step": 5979
+    },
+    {
+      "epoch": 0.719052486021764,
+      "grad_norm": 0.8149386333211902,
+      "learning_rate": 7.722541136727343e-07,
+      "loss": 0.6088,
+      "num_input_tokens_seen": 126931030,
+      "step": 5980
+    },
+    {
+      "epoch": 0.719172728912403,
+      "grad_norm": 2.176573895538796,
+      "learning_rate": 7.716392828196483e-07,
+      "loss": 0.8061,
+      "num_input_tokens_seen": 126948550,
+      "step": 5981
+    },
+    {
+      "epoch": 0.7192929718030422,
+      "grad_norm": 3.2493379123144703,
+      "learning_rate": 7.710246383079057e-07,
+      "loss": 0.7648,
+      "num_input_tokens_seen": 126963655,
+      "step": 5982
+    },
+    {
+      "epoch": 0.7194132146936812,
+      "grad_norm": 2.6072413778681582,
+      "learning_rate": 7.704101802307492e-07,
+      "loss": 0.9128,
+      "num_input_tokens_seen": 126975675,
+      "step": 5983
+    },
+    {
+      "epoch": 0.7195334575843203,
+      "grad_norm": 2.452466744129387,
+      "learning_rate": 7.697959086813906e-07,
+      "loss": 0.8713,
+      "num_input_tokens_seen": 126991560,
+      "step": 5984
+    },
+    {
+      "epoch": 0.7196537004749595,
+      "grad_norm": 1.9079363467480595,
+      "learning_rate": 7.691818237530145e-07,
+      "loss": 0.7955,
+      "num_input_tokens_seen": 127010140,
+      "step": 5985
+    },
+    {
+      "epoch": 0.7197739433655985,
+      "grad_norm": 3.080624927405265,
+      "learning_rate": 7.685679255387774e-07,
+      "loss": 0.773,
+      "num_input_tokens_seen": 127028175,
+      "step": 5986
+    },
+    {
+      "epoch": 0.7198941862562376,
+      "grad_norm": 2.337735198455269,
+      "learning_rate": 7.679542141318065e-07,
+      "loss": 0.7647,
+      "num_input_tokens_seen": 127045000,
+      "step": 5987
+    },
+    {
+      "epoch": 0.7200144291468767,
+      "grad_norm": 1.8635313567547989,
+      "learning_rate": 7.67340689625202e-07,
+      "loss": 0.7591,
+      "num_input_tokens_seen": 127066095,
+      "step": 5988
+    },
+    {
+      "epoch": 0.7201346720375158,
+      "grad_norm": 1.6116332146169243,
+      "learning_rate": 7.667273521120347e-07,
+      "loss": 0.7771,
+      "num_input_tokens_seen": 127085375,
+      "step": 5989
+    },
+    {
+      "epoch": 0.7202549149281549,
+      "grad_norm": 2.262156162347373,
+      "learning_rate": 7.661142016853468e-07,
+      "loss": 0.7934,
+      "num_input_tokens_seen": 127102455,
+      "step": 5990
+    },
+    {
+      "epoch": 0.7203751578187939,
+      "grad_norm": 2.0128542565893155,
+      "learning_rate": 7.655012384381543e-07,
+      "loss": 0.7439,
+      "num_input_tokens_seen": 127121660,
+      "step": 5991
+    },
+    {
+      "epoch": 0.7204954007094331,
+      "grad_norm": 1.9066759849473476,
+      "learning_rate": 7.648884624634422e-07,
+      "loss": 0.8086,
+      "num_input_tokens_seen": 127139930,
+      "step": 5992
+    },
+    {
+      "epoch": 0.7206156436000721,
+      "grad_norm": 1.9261636133864954,
+      "learning_rate": 7.642758738541683e-07,
+      "loss": 0.8779,
+      "num_input_tokens_seen": 127156230,
+      "step": 5993
+    },
+    {
+      "epoch": 0.7207358864907112,
+      "grad_norm": 0.8004235625357764,
+      "learning_rate": 7.636634727032613e-07,
+      "loss": 0.6186,
+      "num_input_tokens_seen": 127213055,
+      "step": 5994
+    },
+    {
+      "epoch": 0.7208561293813504,
+      "grad_norm": 2.592785243194575,
+      "learning_rate": 7.630512591036231e-07,
+      "loss": 0.7904,
+      "num_input_tokens_seen": 127232085,
+      "step": 5995
+    },
+    {
+      "epoch": 0.7209763722719894,
+      "grad_norm": 7.701021332679958,
+      "learning_rate": 7.624392331481255e-07,
+      "loss": 0.6483,
+      "num_input_tokens_seen": 127249460,
+      "step": 5996
+    },
+    {
+      "epoch": 0.7210966151626285,
+      "grad_norm": 0.7452960272034347,
+      "learning_rate": 7.618273949296121e-07,
+      "loss": 0.5415,
+      "num_input_tokens_seen": 127308690,
+      "step": 5997
+    },
+    {
+      "epoch": 0.7212168580532676,
+      "grad_norm": 2.2312588143596397,
+      "learning_rate": 7.612157445408977e-07,
+      "loss": 0.6805,
+      "num_input_tokens_seen": 127326220,
+      "step": 5998
+    },
+    {
+      "epoch": 0.7213371009439067,
+      "grad_norm": 3.1190137399266855,
+      "learning_rate": 7.606042820747709e-07,
+      "loss": 0.7373,
+      "num_input_tokens_seen": 127342345,
+      "step": 5999
+    },
+    {
+      "epoch": 0.7214573438345457,
+      "grad_norm": 1.8428060407422815,
+      "learning_rate": 7.599930076239889e-07,
+      "loss": 0.8471,
+      "num_input_tokens_seen": 127359350,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7215775867251849,
+      "grad_norm": 1.932430630127869,
+      "learning_rate": 7.593819212812818e-07,
+      "loss": 0.7065,
+      "num_input_tokens_seen": 127380650,
+      "step": 6001
+    },
+    {
+      "epoch": 0.721697829615824,
+      "grad_norm": 3.019556453587002,
+      "learning_rate": 7.587710231393508e-07,
+      "loss": 0.7231,
+      "num_input_tokens_seen": 127398725,
+      "step": 6002
+    },
+    {
+      "epoch": 0.721818072506463,
+      "grad_norm": 2.693911332211301,
+      "learning_rate": 7.581603132908685e-07,
+      "loss": 0.8292,
+      "num_input_tokens_seen": 127416415,
+      "step": 6003
+    },
+    {
+      "epoch": 0.7219383153971022,
+      "grad_norm": 3.948674689801315,
+      "learning_rate": 7.575497918284795e-07,
+      "loss": 0.7734,
+      "num_input_tokens_seen": 127433680,
+      "step": 6004
+    },
+    {
+      "epoch": 0.7220585582877412,
+      "grad_norm": 2.3923381033490476,
+      "learning_rate": 7.569394588447992e-07,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 127450415,
+      "step": 6005
+    },
+    {
+      "epoch": 0.7221788011783803,
+      "grad_norm": 4.883190723785703,
+      "learning_rate": 7.563293144324139e-07,
+      "loss": 0.774,
+      "num_input_tokens_seen": 127465685,
+      "step": 6006
+    },
+    {
+      "epoch": 0.7222990440690195,
+      "grad_norm": 2.171551114174458,
+      "learning_rate": 7.557193586838834e-07,
+      "loss": 0.7974,
+      "num_input_tokens_seen": 127480770,
+      "step": 6007
+    },
+    {
+      "epoch": 0.7224192869596585,
+      "grad_norm": 2.429654985195171,
+      "learning_rate": 7.551095916917371e-07,
+      "loss": 0.7025,
+      "num_input_tokens_seen": 127497820,
+      "step": 6008
+    },
+    {
+      "epoch": 0.7225395298502976,
+      "grad_norm": 4.001595389917697,
+      "learning_rate": 7.545000135484758e-07,
+      "loss": 0.6615,
+      "num_input_tokens_seen": 127514975,
+      "step": 6009
+    },
+    {
+      "epoch": 0.7226597727409367,
+      "grad_norm": 2.282572083469935,
+      "learning_rate": 7.538906243465714e-07,
+      "loss": 0.6249,
+      "num_input_tokens_seen": 127534830,
+      "step": 6010
+    },
+    {
+      "epoch": 0.7227800156315758,
+      "grad_norm": 2.257202023658201,
+      "learning_rate": 7.532814241784693e-07,
+      "loss": 0.7762,
+      "num_input_tokens_seen": 127551315,
+      "step": 6011
+    },
+    {
+      "epoch": 0.7229002585222148,
+      "grad_norm": 1.962389888024602,
+      "learning_rate": 7.526724131365838e-07,
+      "loss": 0.688,
+      "num_input_tokens_seen": 127571990,
+      "step": 6012
+    },
+    {
+      "epoch": 0.723020501412854,
+      "grad_norm": 3.1777409189049,
+      "learning_rate": 7.520635913133017e-07,
+      "loss": 0.6979,
+      "num_input_tokens_seen": 127590340,
+      "step": 6013
+    },
+    {
+      "epoch": 0.7231407443034931,
+      "grad_norm": 3.8676112323254332,
+      "learning_rate": 7.514549588009804e-07,
+      "loss": 0.8145,
+      "num_input_tokens_seen": 127610935,
+      "step": 6014
+    },
+    {
+      "epoch": 0.7232609871941321,
+      "grad_norm": 2.142341987214896,
+      "learning_rate": 7.508465156919492e-07,
+      "loss": 0.7036,
+      "num_input_tokens_seen": 127634165,
+      "step": 6015
+    },
+    {
+      "epoch": 0.7233812300847713,
+      "grad_norm": 3.271809125721559,
+      "learning_rate": 7.502382620785083e-07,
+      "loss": 0.6215,
+      "num_input_tokens_seen": 127650435,
+      "step": 6016
+    },
+    {
+      "epoch": 0.7235014729754103,
+      "grad_norm": 0.8811509860409997,
+      "learning_rate": 7.496301980529296e-07,
+      "loss": 0.6887,
+      "num_input_tokens_seen": 127713365,
+      "step": 6017
+    },
+    {
+      "epoch": 0.7236217158660494,
+      "grad_norm": 2.4687664104283806,
+      "learning_rate": 7.490223237074547e-07,
+      "loss": 0.741,
+      "num_input_tokens_seen": 127732795,
+      "step": 6018
+    },
+    {
+      "epoch": 0.7237419587566886,
+      "grad_norm": 3.480842721464305,
+      "learning_rate": 7.484146391342996e-07,
+      "loss": 0.6526,
+      "num_input_tokens_seen": 127752310,
+      "step": 6019
+    },
+    {
+      "epoch": 0.7238622016473276,
+      "grad_norm": 4.125243374523593,
+      "learning_rate": 7.478071444256484e-07,
+      "loss": 0.567,
+      "num_input_tokens_seen": 127769790,
+      "step": 6020
+    },
+    {
+      "epoch": 0.7239824445379667,
+      "grad_norm": 2.947257535081137,
+      "learning_rate": 7.471998396736579e-07,
+      "loss": 0.7804,
+      "num_input_tokens_seen": 127789890,
+      "step": 6021
+    },
+    {
+      "epoch": 0.7241026874286057,
+      "grad_norm": 1.7203305881308386,
+      "learning_rate": 7.465927249704549e-07,
+      "loss": 0.7466,
+      "num_input_tokens_seen": 127807495,
+      "step": 6022
+    },
+    {
+      "epoch": 0.7242229303192449,
+      "grad_norm": 1.781019346072082,
+      "learning_rate": 7.459858004081398e-07,
+      "loss": 0.7651,
+      "num_input_tokens_seen": 127825185,
+      "step": 6023
+    },
+    {
+      "epoch": 0.724343173209884,
+      "grad_norm": 1.4043195305498013,
+      "learning_rate": 7.453790660787815e-07,
+      "loss": 0.5991,
+      "num_input_tokens_seen": 127893000,
+      "step": 6024
+    },
+    {
+      "epoch": 0.724463416100523,
+      "grad_norm": 2.346777454868482,
+      "learning_rate": 7.447725220744214e-07,
+      "loss": 0.6317,
+      "num_input_tokens_seen": 127914965,
+      "step": 6025
+    },
+    {
+      "epoch": 0.7245836589911622,
+      "grad_norm": 2.6436261936115035,
+      "learning_rate": 7.44166168487071e-07,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 127934940,
+      "step": 6026
+    },
+    {
+      "epoch": 0.7247039018818012,
+      "grad_norm": 1.9341843939480519,
+      "learning_rate": 7.435600054087152e-07,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 127956825,
+      "step": 6027
+    },
+    {
+      "epoch": 0.7248241447724403,
+      "grad_norm": 2.5057000785449564,
+      "learning_rate": 7.429540329313074e-07,
+      "loss": 0.7425,
+      "num_input_tokens_seen": 127977585,
+      "step": 6028
+    },
+    {
+      "epoch": 0.7249443876630794,
+      "grad_norm": 2.821246576160181,
+      "learning_rate": 7.423482511467733e-07,
+      "loss": 0.7348,
+      "num_input_tokens_seen": 127998075,
+      "step": 6029
+    },
+    {
+      "epoch": 0.7250646305537185,
+      "grad_norm": 2.7519864013125233,
+      "learning_rate": 7.417426601470099e-07,
+      "loss": 0.6494,
+      "num_input_tokens_seen": 128018155,
+      "step": 6030
+    },
+    {
+      "epoch": 0.7251848734443576,
+      "grad_norm": 4.728718352846354,
+      "learning_rate": 7.411372600238841e-07,
+      "loss": 0.7788,
+      "num_input_tokens_seen": 128038490,
+      "step": 6031
+    },
+    {
+      "epoch": 0.7253051163349967,
+      "grad_norm": 2.9919865971125645,
+      "learning_rate": 7.405320508692352e-07,
+      "loss": 0.734,
+      "num_input_tokens_seen": 128056950,
+      "step": 6032
+    },
+    {
+      "epoch": 0.7254253592256358,
+      "grad_norm": 2.294866685366273,
+      "learning_rate": 7.399270327748727e-07,
+      "loss": 0.7509,
+      "num_input_tokens_seen": 128074330,
+      "step": 6033
+    },
+    {
+      "epoch": 0.7255456021162748,
+      "grad_norm": 2.573428054503482,
+      "learning_rate": 7.39322205832577e-07,
+      "loss": 0.7407,
+      "num_input_tokens_seen": 128094940,
+      "step": 6034
+    },
+    {
+      "epoch": 0.725665845006914,
+      "grad_norm": 2.4812717277175724,
+      "learning_rate": 7.387175701341009e-07,
+      "loss": 0.798,
+      "num_input_tokens_seen": 128113330,
+      "step": 6035
+    },
+    {
+      "epoch": 0.7257860878975531,
+      "grad_norm": 3.1899022408767865,
+      "learning_rate": 7.381131257711666e-07,
+      "loss": 0.7333,
+      "num_input_tokens_seen": 128130155,
+      "step": 6036
+    },
+    {
+      "epoch": 0.7259063307881921,
+      "grad_norm": 2.156626242718935,
+      "learning_rate": 7.375088728354677e-07,
+      "loss": 0.8318,
+      "num_input_tokens_seen": 128144905,
+      "step": 6037
+    },
+    {
+      "epoch": 0.7260265736788313,
+      "grad_norm": 3.436898749245299,
+      "learning_rate": 7.369048114186685e-07,
+      "loss": 0.6653,
+      "num_input_tokens_seen": 128165670,
+      "step": 6038
+    },
+    {
+      "epoch": 0.7261468165694703,
+      "grad_norm": 1.9228060106765044,
+      "learning_rate": 7.363009416124055e-07,
+      "loss": 0.828,
+      "num_input_tokens_seen": 128184715,
+      "step": 6039
+    },
+    {
+      "epoch": 0.7262670594601094,
+      "grad_norm": 2.84561545439624,
+      "learning_rate": 7.356972635082852e-07,
+      "loss": 0.6314,
+      "num_input_tokens_seen": 128203290,
+      "step": 6040
+    },
+    {
+      "epoch": 0.7263873023507486,
+      "grad_norm": 4.085337008285573,
+      "learning_rate": 7.350937771978847e-07,
+      "loss": 0.7476,
+      "num_input_tokens_seen": 128223080,
+      "step": 6041
+    },
+    {
+      "epoch": 0.7265075452413876,
+      "grad_norm": 3.6370958890663085,
+      "learning_rate": 7.344904827727519e-07,
+      "loss": 0.8457,
+      "num_input_tokens_seen": 128239980,
+      "step": 6042
+    },
+    {
+      "epoch": 0.7266277881320267,
+      "grad_norm": 2.978948940095128,
+      "learning_rate": 7.33887380324407e-07,
+      "loss": 0.7256,
+      "num_input_tokens_seen": 128254935,
+      "step": 6043
+    },
+    {
+      "epoch": 0.7267480310226658,
+      "grad_norm": 1.8045441250723215,
+      "learning_rate": 7.332844699443401e-07,
+      "loss": 0.7952,
+      "num_input_tokens_seen": 128273255,
+      "step": 6044
+    },
+    {
+      "epoch": 0.7268682739133049,
+      "grad_norm": 2.891263978504804,
+      "learning_rate": 7.326817517240121e-07,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 128294680,
+      "step": 6045
+    },
+    {
+      "epoch": 0.7269885168039439,
+      "grad_norm": 4.3115409786657555,
+      "learning_rate": 7.320792257548545e-07,
+      "loss": 0.8312,
+      "num_input_tokens_seen": 128315575,
+      "step": 6046
+    },
+    {
+      "epoch": 0.7271087596945831,
+      "grad_norm": 3.3347672501136456,
+      "learning_rate": 7.314768921282704e-07,
+      "loss": 0.7662,
+      "num_input_tokens_seen": 128335950,
+      "step": 6047
+    },
+    {
+      "epoch": 0.7272290025852222,
+      "grad_norm": 2.7446353530954477,
+      "learning_rate": 7.30874750935633e-07,
+      "loss": 0.7098,
+      "num_input_tokens_seen": 128355355,
+      "step": 6048
+    },
+    {
+      "epoch": 0.7273492454758612,
+      "grad_norm": 2.2518717189056576,
+      "learning_rate": 7.302728022682869e-07,
+      "loss": 0.7836,
+      "num_input_tokens_seen": 128372070,
+      "step": 6049
+    },
+    {
+      "epoch": 0.7274694883665004,
+      "grad_norm": 3.1131514248189145,
+      "learning_rate": 7.296710462175464e-07,
+      "loss": 0.7572,
+      "num_input_tokens_seen": 128390900,
+      "step": 6050
+    },
+    {
+      "epoch": 0.7275897312571394,
+      "grad_norm": 2.6819960937452736,
+      "learning_rate": 7.290694828746988e-07,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 128410285,
+      "step": 6051
+    },
+    {
+      "epoch": 0.7277099741477785,
+      "grad_norm": 3.520379790824816,
+      "learning_rate": 7.284681123310004e-07,
+      "loss": 0.8554,
+      "num_input_tokens_seen": 128428720,
+      "step": 6052
+    },
+    {
+      "epoch": 0.7278302170384175,
+      "grad_norm": 5.851062002035487,
+      "learning_rate": 7.27866934677678e-07,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 128448110,
+      "step": 6053
+    },
+    {
+      "epoch": 0.7279504599290567,
+      "grad_norm": 1.9870206113182822,
+      "learning_rate": 7.272659500059297e-07,
+      "loss": 0.7735,
+      "num_input_tokens_seen": 128465170,
+      "step": 6054
+    },
+    {
+      "epoch": 0.7280707028196958,
+      "grad_norm": 2.0897166472497037,
+      "learning_rate": 7.266651584069256e-07,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 128482555,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7281909457103348,
+      "grad_norm": 1.7866459548501132,
+      "learning_rate": 7.260645599718045e-07,
+      "loss": 0.5687,
+      "num_input_tokens_seen": 128508630,
+      "step": 6056
+    },
+    {
+      "epoch": 0.728311188600974,
+      "grad_norm": 3.220914906018947,
+      "learning_rate": 7.254641547916767e-07,
+      "loss": 0.6661,
+      "num_input_tokens_seen": 128525845,
+      "step": 6057
+    },
+    {
+      "epoch": 0.728431431491613,
+      "grad_norm": 2.0605856054673386,
+      "learning_rate": 7.248639429576234e-07,
+      "loss": 0.6863,
+      "num_input_tokens_seen": 128545020,
+      "step": 6058
+    },
+    {
+      "epoch": 0.7285516743822521,
+      "grad_norm": 2.6631959703913024,
+      "learning_rate": 7.242639245606959e-07,
+      "loss": 0.7232,
+      "num_input_tokens_seen": 128564530,
+      "step": 6059
+    },
+    {
+      "epoch": 0.7286719172728913,
+      "grad_norm": 1.989737404280523,
+      "learning_rate": 7.236640996919168e-07,
+      "loss": 0.8245,
+      "num_input_tokens_seen": 128583295,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7287921601635303,
+      "grad_norm": 2.00108215447217,
+      "learning_rate": 7.230644684422789e-07,
+      "loss": 0.7062,
+      "num_input_tokens_seen": 128603245,
+      "step": 6061
+    },
+    {
+      "epoch": 0.7289124030541694,
+      "grad_norm": 2.051632005993278,
+      "learning_rate": 7.224650309027451e-07,
+      "loss": 0.8138,
+      "num_input_tokens_seen": 128622715,
+      "step": 6062
+    },
+    {
+      "epoch": 0.7290326459448085,
+      "grad_norm": 2.6032836186423465,
+      "learning_rate": 7.218657871642512e-07,
+      "loss": 0.6767,
+      "num_input_tokens_seen": 128641240,
+      "step": 6063
+    },
+    {
+      "epoch": 0.7291528888354476,
+      "grad_norm": 3.3610758079283727,
+      "learning_rate": 7.212667373177012e-07,
+      "loss": 0.6113,
+      "num_input_tokens_seen": 128655955,
+      "step": 6064
+    },
+    {
+      "epoch": 0.7292731317260867,
+      "grad_norm": 2.200698457498222,
+      "learning_rate": 7.206678814539704e-07,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 128673975,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7293933746167258,
+      "grad_norm": 2.226091148949567,
+      "learning_rate": 7.20069219663904e-07,
+      "loss": 0.7241,
+      "num_input_tokens_seen": 128693580,
+      "step": 6066
+    },
+    {
+      "epoch": 0.7295136175073649,
+      "grad_norm": 2.4209448392144757,
+      "learning_rate": 7.1947075203832e-07,
+      "loss": 0.7952,
+      "num_input_tokens_seen": 128713280,
+      "step": 6067
+    },
+    {
+      "epoch": 0.7296338603980039,
+      "grad_norm": 1.1248126970217327,
+      "learning_rate": 7.188724786680049e-07,
+      "loss": 0.6134,
+      "num_input_tokens_seen": 128773470,
+      "step": 6068
+    },
+    {
+      "epoch": 0.7297541032886431,
+      "grad_norm": 1.9710274931489284,
+      "learning_rate": 7.182743996437162e-07,
+      "loss": 0.7444,
+      "num_input_tokens_seen": 128792725,
+      "step": 6069
+    },
+    {
+      "epoch": 0.7298743461792822,
+      "grad_norm": 2.2890650565063684,
+      "learning_rate": 7.176765150561812e-07,
+      "loss": 0.6804,
+      "num_input_tokens_seen": 128811050,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7299945890699212,
+      "grad_norm": 2.4091861150171057,
+      "learning_rate": 7.170788249961002e-07,
+      "loss": 0.7909,
+      "num_input_tokens_seen": 128829280,
+      "step": 6071
+    },
+    {
+      "epoch": 0.7301148319605604,
+      "grad_norm": 2.313346913900591,
+      "learning_rate": 7.164813295541412e-07,
+      "loss": 0.8797,
+      "num_input_tokens_seen": 128848565,
+      "step": 6072
+    },
+    {
+      "epoch": 0.7302350748511994,
+      "grad_norm": 1.97986860280755,
+      "learning_rate": 7.15884028820944e-07,
+      "loss": 0.6945,
+      "num_input_tokens_seen": 128867340,
+      "step": 6073
+    },
+    {
+      "epoch": 0.7303553177418385,
+      "grad_norm": 2.3942194688134135,
+      "learning_rate": 7.152869228871185e-07,
+      "loss": 0.5954,
+      "num_input_tokens_seen": 128889545,
+      "step": 6074
+    },
+    {
+      "epoch": 0.7304755606324776,
+      "grad_norm": 2.4428380423786877,
+      "learning_rate": 7.146900118432457e-07,
+      "loss": 0.715,
+      "num_input_tokens_seen": 128909010,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7305958035231167,
+      "grad_norm": 1.8614967404043345,
+      "learning_rate": 7.140932957798759e-07,
+      "loss": 0.8486,
+      "num_input_tokens_seen": 128927170,
+      "step": 6076
+    },
+    {
+      "epoch": 0.7307160464137558,
+      "grad_norm": 1.9805047861872567,
+      "learning_rate": 7.134967747875309e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 128945100,
+      "step": 6077
+    },
+    {
+      "epoch": 0.7308362893043949,
+      "grad_norm": 2.4014442499278386,
+      "learning_rate": 7.129004489567014e-07,
+      "loss": 0.8063,
+      "num_input_tokens_seen": 128962300,
+      "step": 6078
+    },
+    {
+      "epoch": 0.730956532195034,
+      "grad_norm": 41.363265287859434,
+      "learning_rate": 7.123043183778512e-07,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 128979350,
+      "step": 6079
+    },
+    {
+      "epoch": 0.731076775085673,
+      "grad_norm": 1.6456871501273476,
+      "learning_rate": 7.117083831414122e-07,
+      "loss": 0.6405,
+      "num_input_tokens_seen": 128998345,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7311970179763122,
+      "grad_norm": 2.4294938073202097,
+      "learning_rate": 7.11112643337787e-07,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 129017110,
+      "step": 6081
+    },
+    {
+      "epoch": 0.7313172608669513,
+      "grad_norm": 2.450763682414957,
+      "learning_rate": 7.105170990573484e-07,
+      "loss": 0.7574,
+      "num_input_tokens_seen": 129033780,
+      "step": 6082
+    },
+    {
+      "epoch": 0.7314375037575903,
+      "grad_norm": 3.102626626507124,
+      "learning_rate": 7.099217503904411e-07,
+      "loss": 0.6097,
+      "num_input_tokens_seen": 129051355,
+      "step": 6083
+    },
+    {
+      "epoch": 0.7315577466482295,
+      "grad_norm": 9.44806186414828,
+      "learning_rate": 7.093265974273788e-07,
+      "loss": 0.8978,
+      "num_input_tokens_seen": 129068970,
+      "step": 6084
+    },
+    {
+      "epoch": 0.7316779895388685,
+      "grad_norm": 2.009054671319429,
+      "learning_rate": 7.087316402584453e-07,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 129087515,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7317982324295076,
+      "grad_norm": 2.016204630139146,
+      "learning_rate": 7.081368789738947e-07,
+      "loss": 0.8483,
+      "num_input_tokens_seen": 129104435,
+      "step": 6086
+    },
+    {
+      "epoch": 0.7319184753201466,
+      "grad_norm": 2.8238217052094017,
+      "learning_rate": 7.075423136639531e-07,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 129123410,
+      "step": 6087
+    },
+    {
+      "epoch": 0.7320387182107858,
+      "grad_norm": 2.455112707077979,
+      "learning_rate": 7.069479444188149e-07,
+      "loss": 0.7408,
+      "num_input_tokens_seen": 129143720,
+      "step": 6088
+    },
+    {
+      "epoch": 0.7321589611014249,
+      "grad_norm": 2.417089871530588,
+      "learning_rate": 7.063537713286461e-07,
+      "loss": 0.8213,
+      "num_input_tokens_seen": 129161120,
+      "step": 6089
+    },
+    {
+      "epoch": 0.7322792039920639,
+      "grad_norm": 1.9645444849915148,
+      "learning_rate": 7.057597944835803e-07,
+      "loss": 0.803,
+      "num_input_tokens_seen": 129180115,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7323994468827031,
+      "grad_norm": 2.064679686669884,
+      "learning_rate": 7.051660139737253e-07,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 129198055,
+      "step": 6091
+    },
+    {
+      "epoch": 0.7325196897733421,
+      "grad_norm": 1.9606667900970483,
+      "learning_rate": 7.045724298891565e-07,
+      "loss": 0.7554,
+      "num_input_tokens_seen": 129217245,
+      "step": 6092
+    },
+    {
+      "epoch": 0.7326399326639812,
+      "grad_norm": 2.189003655264423,
+      "learning_rate": 7.039790423199198e-07,
+      "loss": 0.6905,
+      "num_input_tokens_seen": 129236605,
+      "step": 6093
+    },
+    {
+      "epoch": 0.7327601755546204,
+      "grad_norm": 2.385283375288238,
+      "learning_rate": 7.033858513560316e-07,
+      "loss": 0.7796,
+      "num_input_tokens_seen": 129252620,
+      "step": 6094
+    },
+    {
+      "epoch": 0.7328804184452594,
+      "grad_norm": 3.082280751904044,
+      "learning_rate": 7.027928570874794e-07,
+      "loss": 0.7711,
+      "num_input_tokens_seen": 129270530,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7330006613358985,
+      "grad_norm": 1.9275694232687495,
+      "learning_rate": 7.022000596042194e-07,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 129287350,
+      "step": 6096
+    },
+    {
+      "epoch": 0.7331209042265376,
+      "grad_norm": 3.919498117069979,
+      "learning_rate": 7.016074589961784e-07,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 129305635,
+      "step": 6097
+    },
+    {
+      "epoch": 0.7332411471171767,
+      "grad_norm": 2.580664585653789,
+      "learning_rate": 7.01015055353253e-07,
+      "loss": 0.6596,
+      "num_input_tokens_seen": 129327780,
+      "step": 6098
+    },
+    {
+      "epoch": 0.7333613900078157,
+      "grad_norm": 2.008725279965256,
+      "learning_rate": 7.004228487653116e-07,
+      "loss": 0.7698,
+      "num_input_tokens_seen": 129348305,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7334816328984549,
+      "grad_norm": 2.2350749318053045,
+      "learning_rate": 6.998308393221906e-07,
+      "loss": 0.7746,
+      "num_input_tokens_seen": 129366430,
+      "step": 6100
+    },
+    {
+      "epoch": 0.733601875789094,
+      "grad_norm": 3.3344335588747582,
+      "learning_rate": 6.992390271136977e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 129381860,
+      "step": 6101
+    },
+    {
+      "epoch": 0.733722118679733,
+      "grad_norm": 2.124030678251846,
+      "learning_rate": 6.9864741222961e-07,
+      "loss": 0.8545,
+      "num_input_tokens_seen": 129400695,
+      "step": 6102
+    },
+    {
+      "epoch": 0.7338423615703722,
+      "grad_norm": 2.204175519286464,
+      "learning_rate": 6.980559947596751e-07,
+      "loss": 0.7238,
+      "num_input_tokens_seen": 129418955,
+      "step": 6103
+    },
+    {
+      "epoch": 0.7339626044610112,
+      "grad_norm": 2.6375335083590876,
+      "learning_rate": 6.974647747936109e-07,
+      "loss": 0.7578,
+      "num_input_tokens_seen": 129437060,
+      "step": 6104
+    },
+    {
+      "epoch": 0.7340828473516503,
+      "grad_norm": 2.2435324145256006,
+      "learning_rate": 6.968737524211046e-07,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 129453590,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7342030902422895,
+      "grad_norm": 2.6369733767013317,
+      "learning_rate": 6.962829277318132e-07,
+      "loss": 0.7981,
+      "num_input_tokens_seen": 129472905,
+      "step": 6106
+    },
+    {
+      "epoch": 0.7343233331329285,
+      "grad_norm": 2.376747445052866,
+      "learning_rate": 6.956923008153659e-07,
+      "loss": 0.8244,
+      "num_input_tokens_seen": 129492390,
+      "step": 6107
+    },
+    {
+      "epoch": 0.7344435760235676,
+      "grad_norm": 2.4372994325587625,
+      "learning_rate": 6.951018717613593e-07,
+      "loss": 0.8416,
+      "num_input_tokens_seen": 129511125,
+      "step": 6108
+    },
+    {
+      "epoch": 0.7345638189142067,
+      "grad_norm": 1.9203232744908312,
+      "learning_rate": 6.945116406593614e-07,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 129529700,
+      "step": 6109
+    },
+    {
+      "epoch": 0.7346840618048458,
+      "grad_norm": 22.235917557628696,
+      "learning_rate": 6.939216075989089e-07,
+      "loss": 0.7364,
+      "num_input_tokens_seen": 129547350,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7348043046954849,
+      "grad_norm": 2.2380365069196477,
+      "learning_rate": 6.933317726695109e-07,
+      "loss": 0.6586,
+      "num_input_tokens_seen": 129568300,
+      "step": 6111
+    },
+    {
+      "epoch": 0.734924547586124,
+      "grad_norm": 2.89637256690497,
+      "learning_rate": 6.92742135960644e-07,
+      "loss": 0.7912,
+      "num_input_tokens_seen": 129585720,
+      "step": 6112
+    },
+    {
+      "epoch": 0.7350447904767631,
+      "grad_norm": 0.9017764840904604,
+      "learning_rate": 6.921526975617556e-07,
+      "loss": 0.5997,
+      "num_input_tokens_seen": 129644900,
+      "step": 6113
+    },
+    {
+      "epoch": 0.7351650333674021,
+      "grad_norm": 1.8310775935069004,
+      "learning_rate": 6.915634575622625e-07,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 129663135,
+      "step": 6114
+    },
+    {
+      "epoch": 0.7352852762580413,
+      "grad_norm": 2.834489786226201,
+      "learning_rate": 6.909744160515532e-07,
+      "loss": 0.7119,
+      "num_input_tokens_seen": 129680995,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7354055191486804,
+      "grad_norm": 2.537388304611972,
+      "learning_rate": 6.903855731189843e-07,
+      "loss": 0.6792,
+      "num_input_tokens_seen": 129703350,
+      "step": 6116
+    },
+    {
+      "epoch": 0.7355257620393194,
+      "grad_norm": 2.3967596416043957,
+      "learning_rate": 6.897969288538825e-07,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 129721015,
+      "step": 6117
+    },
+    {
+      "epoch": 0.7356460049299585,
+      "grad_norm": 2.42968571050616,
+      "learning_rate": 6.892084833455452e-07,
+      "loss": 0.8069,
+      "num_input_tokens_seen": 129740305,
+      "step": 6118
+    },
+    {
+      "epoch": 0.7357662478205976,
+      "grad_norm": 1.7636107275414388,
+      "learning_rate": 6.886202366832384e-07,
+      "loss": 0.8381,
+      "num_input_tokens_seen": 129761710,
+      "step": 6119
+    },
+    {
+      "epoch": 0.7358864907112367,
+      "grad_norm": 2.430707664843906,
+      "learning_rate": 6.880321889561993e-07,
+      "loss": 0.7371,
+      "num_input_tokens_seen": 129779405,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7360067336018757,
+      "grad_norm": 2.3829015864766547,
+      "learning_rate": 6.874443402536338e-07,
+      "loss": 0.659,
+      "num_input_tokens_seen": 129798215,
+      "step": 6121
+    },
+    {
+      "epoch": 0.7361269764925149,
+      "grad_norm": 1.8503110283205804,
+      "learning_rate": 6.868566906647177e-07,
+      "loss": 0.7964,
+      "num_input_tokens_seen": 129818885,
+      "step": 6122
+    },
+    {
+      "epoch": 0.736247219383154,
+      "grad_norm": 1.9334544049962614,
+      "learning_rate": 6.862692402785984e-07,
+      "loss": 0.8233,
+      "num_input_tokens_seen": 129838855,
+      "step": 6123
+    },
+    {
+      "epoch": 0.736367462273793,
+      "grad_norm": 0.706384267566739,
+      "learning_rate": 6.856819891843905e-07,
+      "loss": 0.5159,
+      "num_input_tokens_seen": 129903280,
+      "step": 6124
+    },
+    {
+      "epoch": 0.7364877051644322,
+      "grad_norm": 2.4974335458414822,
+      "learning_rate": 6.8509493747118e-07,
+      "loss": 0.722,
+      "num_input_tokens_seen": 129921810,
+      "step": 6125
+    },
+    {
+      "epoch": 0.7366079480550712,
+      "grad_norm": 4.167968514206307,
+      "learning_rate": 6.845080852280213e-07,
+      "loss": 0.8793,
+      "num_input_tokens_seen": 129938600,
+      "step": 6126
+    },
+    {
+      "epoch": 0.7367281909457103,
+      "grad_norm": 1.80061949395276,
+      "learning_rate": 6.839214325439409e-07,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 129956015,
+      "step": 6127
+    },
+    {
+      "epoch": 0.7368484338363495,
+      "grad_norm": 1.6820896811155979,
+      "learning_rate": 6.833349795079327e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 129974845,
+      "step": 6128
+    },
+    {
+      "epoch": 0.7369686767269885,
+      "grad_norm": 2.0374385285789915,
+      "learning_rate": 6.827487262089613e-07,
+      "loss": 0.6802,
+      "num_input_tokens_seen": 129995070,
+      "step": 6129
+    },
+    {
+      "epoch": 0.7370889196176276,
+      "grad_norm": 0.9163140055993958,
+      "learning_rate": 6.8216267273596e-07,
+      "loss": 0.6001,
+      "num_input_tokens_seen": 130060350,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7372091625082667,
+      "grad_norm": 4.372017083883281,
+      "learning_rate": 6.815768191778342e-07,
+      "loss": 0.7747,
+      "num_input_tokens_seen": 130078150,
+      "step": 6131
+    },
+    {
+      "epoch": 0.7373294053989058,
+      "grad_norm": 1.9991369768946987,
+      "learning_rate": 6.809911656234575e-07,
+      "loss": 0.7307,
+      "num_input_tokens_seen": 130099845,
+      "step": 6132
+    },
+    {
+      "epoch": 0.7374496482895448,
+      "grad_norm": 2.7718994211318235,
+      "learning_rate": 6.804057121616713e-07,
+      "loss": 0.7835,
+      "num_input_tokens_seen": 130117770,
+      "step": 6133
+    },
+    {
+      "epoch": 0.737569891180184,
+      "grad_norm": 2.091485395373309,
+      "learning_rate": 6.798204588812888e-07,
+      "loss": 0.7171,
+      "num_input_tokens_seen": 130136905,
+      "step": 6134
+    },
+    {
+      "epoch": 0.7376901340708231,
+      "grad_norm": 2.1228414387598336,
+      "learning_rate": 6.792354058710937e-07,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 130154095,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7378103769614621,
+      "grad_norm": 5.784542254557947,
+      "learning_rate": 6.786505532198374e-07,
+      "loss": 0.6553,
+      "num_input_tokens_seen": 130172760,
+      "step": 6136
+    },
+    {
+      "epoch": 0.7379306198521013,
+      "grad_norm": 2.19090763445533,
+      "learning_rate": 6.780659010162417e-07,
+      "loss": 0.8445,
+      "num_input_tokens_seen": 130191430,
+      "step": 6137
+    },
+    {
+      "epoch": 0.7380508627427403,
+      "grad_norm": 1.849804107875879,
+      "learning_rate": 6.774814493489969e-07,
+      "loss": 0.8252,
+      "num_input_tokens_seen": 130208825,
+      "step": 6138
+    },
+    {
+      "epoch": 0.7381711056333794,
+      "grad_norm": 2.114469309547878,
+      "learning_rate": 6.768971983067655e-07,
+      "loss": 0.661,
+      "num_input_tokens_seen": 130228875,
+      "step": 6139
+    },
+    {
+      "epoch": 0.7382913485240186,
+      "grad_norm": 1.0841302311203715,
+      "learning_rate": 6.763131479781772e-07,
+      "loss": 0.7047,
+      "num_input_tokens_seen": 130278355,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7384115914146576,
+      "grad_norm": 3.246889732029036,
+      "learning_rate": 6.757292984518316e-07,
+      "loss": 0.774,
+      "num_input_tokens_seen": 130297475,
+      "step": 6141
+    },
+    {
+      "epoch": 0.7385318343052967,
+      "grad_norm": 0.8745190247783067,
+      "learning_rate": 6.751456498162981e-07,
+      "loss": 0.615,
+      "num_input_tokens_seen": 130356230,
+      "step": 6142
+    },
+    {
+      "epoch": 0.7386520771959358,
+      "grad_norm": 2.164061349621602,
+      "learning_rate": 6.745622021601167e-07,
+      "loss": 0.8491,
+      "num_input_tokens_seen": 130372975,
+      "step": 6143
+    },
+    {
+      "epoch": 0.7387723200865749,
+      "grad_norm": 2.9392690147388443,
+      "learning_rate": 6.739789555717954e-07,
+      "loss": 0.704,
+      "num_input_tokens_seen": 130389670,
+      "step": 6144
+    },
+    {
+      "epoch": 0.738892562977214,
+      "grad_norm": 2.438339528576635,
+      "learning_rate": 6.733959101398124e-07,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 130407520,
+      "step": 6145
+    },
+    {
+      "epoch": 0.7390128058678531,
+      "grad_norm": 2.0102901008691823,
+      "learning_rate": 6.72813065952615e-07,
+      "loss": 0.8091,
+      "num_input_tokens_seen": 130425050,
+      "step": 6146
+    },
+    {
+      "epoch": 0.7391330487584922,
+      "grad_norm": 2.7105356116443384,
+      "learning_rate": 6.7223042309862e-07,
+      "loss": 0.7025,
+      "num_input_tokens_seen": 130444970,
+      "step": 6147
+    },
+    {
+      "epoch": 0.7392532916491312,
+      "grad_norm": 2.4451829151999136,
+      "learning_rate": 6.716479816662144e-07,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 130466420,
+      "step": 6148
+    },
+    {
+      "epoch": 0.7393735345397703,
+      "grad_norm": 2.072380817722059,
+      "learning_rate": 6.710657417437537e-07,
+      "loss": 0.728,
+      "num_input_tokens_seen": 130485845,
+      "step": 6149
+    },
+    {
+      "epoch": 0.7394937774304094,
+      "grad_norm": 4.157813101906924,
+      "learning_rate": 6.704837034195628e-07,
+      "loss": 0.7874,
+      "num_input_tokens_seen": 130504030,
+      "step": 6150
+    },
+    {
+      "epoch": 0.7396140203210485,
+      "grad_norm": 1.8346196805104165,
+      "learning_rate": 6.699018667819376e-07,
+      "loss": 0.845,
+      "num_input_tokens_seen": 130523150,
+      "step": 6151
+    },
+    {
+      "epoch": 0.7397342632116876,
+      "grad_norm": 2.038622815045475,
+      "learning_rate": 6.693202319191415e-07,
+      "loss": 0.7203,
+      "num_input_tokens_seen": 130544605,
+      "step": 6152
+    },
+    {
+      "epoch": 0.7398545061023267,
+      "grad_norm": 1.9652806355905164,
+      "learning_rate": 6.687387989194084e-07,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 130563840,
+      "step": 6153
+    },
+    {
+      "epoch": 0.7399747489929658,
+      "grad_norm": 2.2835269910615086,
+      "learning_rate": 6.681575678709404e-07,
+      "loss": 0.7958,
+      "num_input_tokens_seen": 130582250,
+      "step": 6154
+    },
+    {
+      "epoch": 0.7400949918836048,
+      "grad_norm": 2.1808278160421475,
+      "learning_rate": 6.67576538861911e-07,
+      "loss": 0.6989,
+      "num_input_tokens_seen": 130600545,
+      "step": 6155
+    },
+    {
+      "epoch": 0.740215234774244,
+      "grad_norm": 1.7902682349162393,
+      "learning_rate": 6.669957119804612e-07,
+      "loss": 0.8192,
+      "num_input_tokens_seen": 130621900,
+      "step": 6156
+    },
+    {
+      "epoch": 0.7403354776648831,
+      "grad_norm": 3.2241967800642453,
+      "learning_rate": 6.66415087314702e-07,
+      "loss": 0.7087,
+      "num_input_tokens_seen": 130636575,
+      "step": 6157
+    },
+    {
+      "epoch": 0.7404557205555221,
+      "grad_norm": 2.750336163648373,
+      "learning_rate": 6.658346649527133e-07,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 130653745,
+      "step": 6158
+    },
+    {
+      "epoch": 0.7405759634461613,
+      "grad_norm": 2.0806408185318297,
+      "learning_rate": 6.652544449825457e-07,
+      "loss": 0.7493,
+      "num_input_tokens_seen": 130673720,
+      "step": 6159
+    },
+    {
+      "epoch": 0.7406962063368003,
+      "grad_norm": 1.7759439799619565,
+      "learning_rate": 6.646744274922176e-07,
+      "loss": 0.7578,
+      "num_input_tokens_seen": 130691885,
+      "step": 6160
+    },
+    {
+      "epoch": 0.7408164492274394,
+      "grad_norm": 6.522752526671609,
+      "learning_rate": 6.640946125697171e-07,
+      "loss": 0.7544,
+      "num_input_tokens_seen": 130709135,
+      "step": 6161
+    },
+    {
+      "epoch": 0.7409366921180786,
+      "grad_norm": 2.888736431926943,
+      "learning_rate": 6.635150003030017e-07,
+      "loss": 0.7567,
+      "num_input_tokens_seen": 130727380,
+      "step": 6162
+    },
+    {
+      "epoch": 0.7410569350087176,
+      "grad_norm": 2.613375665571984,
+      "learning_rate": 6.629355907799981e-07,
+      "loss": 0.8561,
+      "num_input_tokens_seen": 130746905,
+      "step": 6163
+    },
+    {
+      "epoch": 0.7411771778993567,
+      "grad_norm": 2.1444508926581087,
+      "learning_rate": 6.623563840886022e-07,
+      "loss": 0.6979,
+      "num_input_tokens_seen": 130767550,
+      "step": 6164
+    },
+    {
+      "epoch": 0.7412974207899958,
+      "grad_norm": 2.6413550402617556,
+      "learning_rate": 6.617773803166795e-07,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 130785595,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7414176636806349,
+      "grad_norm": 2.3734336601310226,
+      "learning_rate": 6.611985795520634e-07,
+      "loss": 0.8176,
+      "num_input_tokens_seen": 130803860,
+      "step": 6166
+    },
+    {
+      "epoch": 0.7415379065712739,
+      "grad_norm": 2.3865155461574923,
+      "learning_rate": 6.606199818825588e-07,
+      "loss": 0.7603,
+      "num_input_tokens_seen": 130824035,
+      "step": 6167
+    },
+    {
+      "epoch": 0.7416581494619131,
+      "grad_norm": 2.35215065475073,
+      "learning_rate": 6.600415873959384e-07,
+      "loss": 0.806,
+      "num_input_tokens_seen": 130841630,
+      "step": 6168
+    },
+    {
+      "epoch": 0.7417783923525522,
+      "grad_norm": 2.965403989764952,
+      "learning_rate": 6.594633961799437e-07,
+      "loss": 0.6441,
+      "num_input_tokens_seen": 130860390,
+      "step": 6169
+    },
+    {
+      "epoch": 0.7418986352431912,
+      "grad_norm": 2.230060175680007,
+      "learning_rate": 6.58885408322285e-07,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 130879545,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7420188781338304,
+      "grad_norm": 2.2890149871782937,
+      "learning_rate": 6.583076239106444e-07,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 130897770,
+      "step": 6171
+    },
+    {
+      "epoch": 0.7421391210244694,
+      "grad_norm": 2.7357449577245387,
+      "learning_rate": 6.577300430326707e-07,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 130912435,
+      "step": 6172
+    },
+    {
+      "epoch": 0.7422593639151085,
+      "grad_norm": 2.670546819823486,
+      "learning_rate": 6.571526657759821e-07,
+      "loss": 0.7135,
+      "num_input_tokens_seen": 130927895,
+      "step": 6173
+    },
+    {
+      "epoch": 0.7423796068057477,
+      "grad_norm": 3.842777133018692,
+      "learning_rate": 6.565754922281657e-07,
+      "loss": 0.7093,
+      "num_input_tokens_seen": 130949860,
+      "step": 6174
+    },
+    {
+      "epoch": 0.7424998496963867,
+      "grad_norm": 1.897159116202194,
+      "learning_rate": 6.559985224767801e-07,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 130967455,
+      "step": 6175
+    },
+    {
+      "epoch": 0.7426200925870258,
+      "grad_norm": 2.7516521150344198,
+      "learning_rate": 6.554217566093496e-07,
+      "loss": 0.7465,
+      "num_input_tokens_seen": 130985430,
+      "step": 6176
+    },
+    {
+      "epoch": 0.7427403354776649,
+      "grad_norm": 3.0097296119077526,
+      "learning_rate": 6.548451947133698e-07,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 131006100,
+      "step": 6177
+    },
+    {
+      "epoch": 0.742860578368304,
+      "grad_norm": 2.196136065511712,
+      "learning_rate": 6.542688368763034e-07,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 131024225,
+      "step": 6178
+    },
+    {
+      "epoch": 0.742980821258943,
+      "grad_norm": 1.901052777838635,
+      "learning_rate": 6.536926831855854e-07,
+      "loss": 0.7697,
+      "num_input_tokens_seen": 131043110,
+      "step": 6179
+    },
+    {
+      "epoch": 0.7431010641495821,
+      "grad_norm": 2.746691240813948,
+      "learning_rate": 6.531167337286165e-07,
+      "loss": 0.7266,
+      "num_input_tokens_seen": 131062850,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7432213070402213,
+      "grad_norm": 1.6885432823452455,
+      "learning_rate": 6.525409885927686e-07,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 131083590,
+      "step": 6181
+    },
+    {
+      "epoch": 0.7433415499308603,
+      "grad_norm": 2.0787481087993265,
+      "learning_rate": 6.519654478653806e-07,
+      "loss": 0.8332,
+      "num_input_tokens_seen": 131101675,
+      "step": 6182
+    },
+    {
+      "epoch": 0.7434617928214994,
+      "grad_norm": 0.7801077334796087,
+      "learning_rate": 6.51390111633763e-07,
+      "loss": 0.5811,
+      "num_input_tokens_seen": 131166670,
+      "step": 6183
+    },
+    {
+      "epoch": 0.7435820357121385,
+      "grad_norm": 2.0190433369509146,
+      "learning_rate": 6.508149799851932e-07,
+      "loss": 0.7552,
+      "num_input_tokens_seen": 131188055,
+      "step": 6184
+    },
+    {
+      "epoch": 0.7437022786027776,
+      "grad_norm": 2.309625367798866,
+      "learning_rate": 6.502400530069183e-07,
+      "loss": 0.6165,
+      "num_input_tokens_seen": 131207660,
+      "step": 6185
+    },
+    {
+      "epoch": 0.7438225214934167,
+      "grad_norm": 2.1925091187398533,
+      "learning_rate": 6.496653307861535e-07,
+      "loss": 0.68,
+      "num_input_tokens_seen": 131228050,
+      "step": 6186
+    },
+    {
+      "epoch": 0.7439427643840558,
+      "grad_norm": 1.9996484675225485,
+      "learning_rate": 6.49090813410085e-07,
+      "loss": 0.6531,
+      "num_input_tokens_seen": 131246235,
+      "step": 6187
+    },
+    {
+      "epoch": 0.7440630072746949,
+      "grad_norm": 2.2901946076725483,
+      "learning_rate": 6.48516500965866e-07,
+      "loss": 0.6906,
+      "num_input_tokens_seen": 131265890,
+      "step": 6188
+    },
+    {
+      "epoch": 0.7441832501653339,
+      "grad_norm": 1.8128198889936369,
+      "learning_rate": 6.479423935406192e-07,
+      "loss": 0.8118,
+      "num_input_tokens_seen": 131285595,
+      "step": 6189
+    },
+    {
+      "epoch": 0.7443034930559731,
+      "grad_norm": 0.9215484257310398,
+      "learning_rate": 6.473684912214363e-07,
+      "loss": 0.7075,
+      "num_input_tokens_seen": 131348875,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7444237359466122,
+      "grad_norm": 2.307173296223332,
+      "learning_rate": 6.467947940953778e-07,
+      "loss": 0.6909,
+      "num_input_tokens_seen": 131367120,
+      "step": 6191
+    },
+    {
+      "epoch": 0.7445439788372512,
+      "grad_norm": 2.478181516923358,
+      "learning_rate": 6.462213022494732e-07,
+      "loss": 0.7186,
+      "num_input_tokens_seen": 131386085,
+      "step": 6192
+    },
+    {
+      "epoch": 0.7446642217278904,
+      "grad_norm": 0.8337086010664749,
+      "learning_rate": 6.456480157707207e-07,
+      "loss": 0.6557,
+      "num_input_tokens_seen": 131450580,
+      "step": 6193
+    },
+    {
+      "epoch": 0.7447844646185294,
+      "grad_norm": 2.0336677824027034,
+      "learning_rate": 6.450749347460866e-07,
+      "loss": 0.8509,
+      "num_input_tokens_seen": 131467275,
+      "step": 6194
+    },
+    {
+      "epoch": 0.7449047075091685,
+      "grad_norm": 1.8449635820557584,
+      "learning_rate": 6.445020592625083e-07,
+      "loss": 0.7871,
+      "num_input_tokens_seen": 131487645,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7450249503998077,
+      "grad_norm": 2.2208985712182034,
+      "learning_rate": 6.4392938940689e-07,
+      "loss": 0.8044,
+      "num_input_tokens_seen": 131502780,
+      "step": 6196
+    },
+    {
+      "epoch": 0.7451451932904467,
+      "grad_norm": 2.5486925102851847,
+      "learning_rate": 6.433569252661049e-07,
+      "loss": 0.7006,
+      "num_input_tokens_seen": 131520500,
+      "step": 6197
+    },
+    {
+      "epoch": 0.7452654361810858,
+      "grad_norm": 2.1374560083020366,
+      "learning_rate": 6.427846669269952e-07,
+      "loss": 0.7058,
+      "num_input_tokens_seen": 131537840,
+      "step": 6198
+    },
+    {
+      "epoch": 0.7453856790717249,
+      "grad_norm": 1.9680436428453851,
+      "learning_rate": 6.422126144763729e-07,
+      "loss": 0.818,
+      "num_input_tokens_seen": 131556950,
+      "step": 6199
+    },
+    {
+      "epoch": 0.745505921962364,
+      "grad_norm": 2.252747848797241,
+      "learning_rate": 6.416407680010174e-07,
+      "loss": 0.7718,
+      "num_input_tokens_seen": 131571030,
+      "step": 6200
+    },
+    {
+      "epoch": 0.745626164853003,
+      "grad_norm": 2.074550575069196,
+      "learning_rate": 6.410691275876774e-07,
+      "loss": 0.8076,
+      "num_input_tokens_seen": 131590170,
+      "step": 6201
+    },
+    {
+      "epoch": 0.7457464077436422,
+      "grad_norm": 7.644304659026054,
+      "learning_rate": 6.404976933230696e-07,
+      "loss": 0.7584,
+      "num_input_tokens_seen": 131606410,
+      "step": 6202
+    },
+    {
+      "epoch": 0.7458666506342813,
+      "grad_norm": 2.153671702211841,
+      "learning_rate": 6.399264652938813e-07,
+      "loss": 0.7226,
+      "num_input_tokens_seen": 131627035,
+      "step": 6203
+    },
+    {
+      "epoch": 0.7459868935249203,
+      "grad_norm": 2.1276632030606537,
+      "learning_rate": 6.393554435867672e-07,
+      "loss": 0.7374,
+      "num_input_tokens_seen": 131647605,
+      "step": 6204
+    },
+    {
+      "epoch": 0.7461071364155595,
+      "grad_norm": 2.5335659087954143,
+      "learning_rate": 6.387846282883502e-07,
+      "loss": 0.8255,
+      "num_input_tokens_seen": 131663855,
+      "step": 6205
+    },
+    {
+      "epoch": 0.7462273793061985,
+      "grad_norm": 2.443258329221693,
+      "learning_rate": 6.38214019485223e-07,
+      "loss": 0.7656,
+      "num_input_tokens_seen": 131682400,
+      "step": 6206
+    },
+    {
+      "epoch": 0.7463476221968376,
+      "grad_norm": 2.185481759928335,
+      "learning_rate": 6.376436172639461e-07,
+      "loss": 0.7106,
+      "num_input_tokens_seen": 131699965,
+      "step": 6207
+    },
+    {
+      "epoch": 0.7464678650874768,
+      "grad_norm": 3.725363092407717,
+      "learning_rate": 6.370734217110494e-07,
+      "loss": 0.6447,
+      "num_input_tokens_seen": 131718430,
+      "step": 6208
+    },
+    {
+      "epoch": 0.7465881079781158,
+      "grad_norm": 19.256440167348703,
+      "learning_rate": 6.36503432913031e-07,
+      "loss": 0.6384,
+      "num_input_tokens_seen": 131741295,
+      "step": 6209
+    },
+    {
+      "epoch": 0.7467083508687549,
+      "grad_norm": 1.9628870552078443,
+      "learning_rate": 6.359336509563569e-07,
+      "loss": 0.6761,
+      "num_input_tokens_seen": 131757035,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7468285937593939,
+      "grad_norm": 2.071013534539995,
+      "learning_rate": 6.353640759274641e-07,
+      "loss": 0.8013,
+      "num_input_tokens_seen": 131775645,
+      "step": 6211
+    },
+    {
+      "epoch": 0.7469488366500331,
+      "grad_norm": 3.4086281004602874,
+      "learning_rate": 6.34794707912756e-07,
+      "loss": 0.7437,
+      "num_input_tokens_seen": 131793265,
+      "step": 6212
+    },
+    {
+      "epoch": 0.7470690795406721,
+      "grad_norm": 2.50271502462732,
+      "learning_rate": 6.342255469986053e-07,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 131811730,
+      "step": 6213
+    },
+    {
+      "epoch": 0.7471893224313112,
+      "grad_norm": 2.0791887329503833,
+      "learning_rate": 6.336565932713527e-07,
+      "loss": 0.773,
+      "num_input_tokens_seen": 131830875,
+      "step": 6214
+    },
+    {
+      "epoch": 0.7473095653219504,
+      "grad_norm": 1.7651660965190452,
+      "learning_rate": 6.330878468173088e-07,
+      "loss": 0.7732,
+      "num_input_tokens_seen": 131850660,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7474298082125894,
+      "grad_norm": 1.9421687644878305,
+      "learning_rate": 6.32519307722752e-07,
+      "loss": 0.7249,
+      "num_input_tokens_seen": 131868275,
+      "step": 6216
+    },
+    {
+      "epoch": 0.7475500511032285,
+      "grad_norm": 0.8144640220043486,
+      "learning_rate": 6.31950976073929e-07,
+      "loss": 0.5802,
+      "num_input_tokens_seen": 131922085,
+      "step": 6217
+    },
+    {
+      "epoch": 0.7476702939938676,
+      "grad_norm": 2.653064343602838,
+      "learning_rate": 6.31382851957055e-07,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 131938625,
+      "step": 6218
+    },
+    {
+      "epoch": 0.7477905368845067,
+      "grad_norm": 10.286464837577718,
+      "learning_rate": 6.308149354583143e-07,
+      "loss": 0.7037,
+      "num_input_tokens_seen": 131957750,
+      "step": 6219
+    },
+    {
+      "epoch": 0.7479107797751458,
+      "grad_norm": 2.950586283588881,
+      "learning_rate": 6.302472266638592e-07,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 131978010,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7480310226657849,
+      "grad_norm": 2.1960507784738965,
+      "learning_rate": 6.296797256598107e-07,
+      "loss": 0.6991,
+      "num_input_tokens_seen": 131999210,
+      "step": 6221
+    },
+    {
+      "epoch": 0.748151265556424,
+      "grad_norm": 2.621399342220879,
+      "learning_rate": 6.291124325322576e-07,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 132019055,
+      "step": 6222
+    },
+    {
+      "epoch": 0.748271508447063,
+      "grad_norm": 1.8982444736940969,
+      "learning_rate": 6.285453473672595e-07,
+      "loss": 0.619,
+      "num_input_tokens_seen": 132041345,
+      "step": 6223
+    },
+    {
+      "epoch": 0.7483917513377022,
+      "grad_norm": 4.290460144155665,
+      "learning_rate": 6.279784702508415e-07,
+      "loss": 0.7567,
+      "num_input_tokens_seen": 132061815,
+      "step": 6224
+    },
+    {
+      "epoch": 0.7485119942283412,
+      "grad_norm": 0.8330735577887192,
+      "learning_rate": 6.274118012689987e-07,
+      "loss": 0.6456,
+      "num_input_tokens_seen": 132123435,
+      "step": 6225
+    },
+    {
+      "epoch": 0.7486322371189803,
+      "grad_norm": 2.0693548885609068,
+      "learning_rate": 6.268453405076937e-07,
+      "loss": 0.6787,
+      "num_input_tokens_seen": 132145550,
+      "step": 6226
+    },
+    {
+      "epoch": 0.7487524800096195,
+      "grad_norm": 2.1435393605445463,
+      "learning_rate": 6.262790880528592e-07,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 132162890,
+      "step": 6227
+    },
+    {
+      "epoch": 0.7488727229002585,
+      "grad_norm": 3.4252993367512374,
+      "learning_rate": 6.257130439903951e-07,
+      "loss": 0.7912,
+      "num_input_tokens_seen": 132179105,
+      "step": 6228
+    },
+    {
+      "epoch": 0.7489929657908976,
+      "grad_norm": 2.143155115115386,
+      "learning_rate": 6.251472084061695e-07,
+      "loss": 0.8011,
+      "num_input_tokens_seen": 132197745,
+      "step": 6229
+    },
+    {
+      "epoch": 0.7491132086815367,
+      "grad_norm": 2.17579804041867,
+      "learning_rate": 6.245815813860184e-07,
+      "loss": 0.8936,
+      "num_input_tokens_seen": 132212975,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7492334515721758,
+      "grad_norm": 5.344181930879343,
+      "learning_rate": 6.240161630157487e-07,
+      "loss": 0.6909,
+      "num_input_tokens_seen": 132232050,
+      "step": 6231
+    },
+    {
+      "epoch": 0.7493536944628149,
+      "grad_norm": 2.563653592848581,
+      "learning_rate": 6.23450953381133e-07,
+      "loss": 0.6937,
+      "num_input_tokens_seen": 132249860,
+      "step": 6232
+    },
+    {
+      "epoch": 0.749473937353454,
+      "grad_norm": 2.14881737953873,
+      "learning_rate": 6.228859525679131e-07,
+      "loss": 0.6751,
+      "num_input_tokens_seen": 132263995,
+      "step": 6233
+    },
+    {
+      "epoch": 0.7495941802440931,
+      "grad_norm": 2.1536892318808394,
+      "learning_rate": 6.223211606617993e-07,
+      "loss": 0.7959,
+      "num_input_tokens_seen": 132282135,
+      "step": 6234
+    },
+    {
+      "epoch": 0.7497144231347321,
+      "grad_norm": 1.8020189379318385,
+      "learning_rate": 6.217565777484701e-07,
+      "loss": 0.8229,
+      "num_input_tokens_seen": 132300950,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7498346660253713,
+      "grad_norm": 1.7648875950356748,
+      "learning_rate": 6.211922039135722e-07,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 132320815,
+      "step": 6236
+    },
+    {
+      "epoch": 0.7499549089160104,
+      "grad_norm": 2.6101216125564037,
+      "learning_rate": 6.206280392427208e-07,
+      "loss": 0.7996,
+      "num_input_tokens_seen": 132340120,
+      "step": 6237
+    },
+    {
+      "epoch": 0.7500751518066494,
+      "grad_norm": 1.8213365546893499,
+      "learning_rate": 6.200640838214983e-07,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 132362615,
+      "step": 6238
+    },
+    {
+      "epoch": 0.7501953946972886,
+      "grad_norm": 2.1930796733989286,
+      "learning_rate": 6.195003377354578e-07,
+      "loss": 0.6683,
+      "num_input_tokens_seen": 132381605,
+      "step": 6239
+    },
+    {
+      "epoch": 0.7503156375879276,
+      "grad_norm": 3.4603565266700507,
+      "learning_rate": 6.189368010701183e-07,
+      "loss": 0.728,
+      "num_input_tokens_seen": 132398385,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7504358804785667,
+      "grad_norm": 2.0900273184119484,
+      "learning_rate": 6.183734739109683e-07,
+      "loss": 0.753,
+      "num_input_tokens_seen": 132415925,
+      "step": 6241
+    },
+    {
+      "epoch": 0.7505561233692057,
+      "grad_norm": 2.5381114127317614,
+      "learning_rate": 6.178103563434629e-07,
+      "loss": 0.6837,
+      "num_input_tokens_seen": 132434645,
+      "step": 6242
+    },
+    {
+      "epoch": 0.7506763662598449,
+      "grad_norm": 1.7457561623563378,
+      "learning_rate": 6.172474484530283e-07,
+      "loss": 0.8338,
+      "num_input_tokens_seen": 132453100,
+      "step": 6243
+    },
+    {
+      "epoch": 0.750796609150484,
+      "grad_norm": 1.9637201652485372,
+      "learning_rate": 6.166847503250563e-07,
+      "loss": 0.75,
+      "num_input_tokens_seen": 132475060,
+      "step": 6244
+    },
+    {
+      "epoch": 0.750916852041123,
+      "grad_norm": 3.1392941287431344,
+      "learning_rate": 6.161222620449078e-07,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 132493555,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7510370949317622,
+      "grad_norm": 4.0444749547274945,
+      "learning_rate": 6.155599836979111e-07,
+      "loss": 0.8016,
+      "num_input_tokens_seen": 132511960,
+      "step": 6246
+    },
+    {
+      "epoch": 0.7511573378224012,
+      "grad_norm": 2.18621966273293,
+      "learning_rate": 6.149979153693649e-07,
+      "loss": 0.8045,
+      "num_input_tokens_seen": 132528935,
+      "step": 6247
+    },
+    {
+      "epoch": 0.7512775807130403,
+      "grad_norm": 7.65615635466436,
+      "learning_rate": 6.144360571445337e-07,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 132547800,
+      "step": 6248
+    },
+    {
+      "epoch": 0.7513978236036795,
+      "grad_norm": 2.1028147941652455,
+      "learning_rate": 6.138744091086509e-07,
+      "loss": 0.7912,
+      "num_input_tokens_seen": 132567105,
+      "step": 6249
+    },
+    {
+      "epoch": 0.7515180664943185,
+      "grad_norm": 9.293610789068884,
+      "learning_rate": 6.133129713469183e-07,
+      "loss": 0.7285,
+      "num_input_tokens_seen": 132586030,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7516383093849576,
+      "grad_norm": 3.501417704553043,
+      "learning_rate": 6.127517439445053e-07,
+      "loss": 0.6323,
+      "num_input_tokens_seen": 132606595,
+      "step": 6251
+    },
+    {
+      "epoch": 0.7517585522755967,
+      "grad_norm": 2.2087797664612903,
+      "learning_rate": 6.121907269865498e-07,
+      "loss": 0.8223,
+      "num_input_tokens_seen": 132625805,
+      "step": 6252
+    },
+    {
+      "epoch": 0.7518787951662358,
+      "grad_norm": 0.9794511107462682,
+      "learning_rate": 6.116299205581577e-07,
+      "loss": 0.7298,
+      "num_input_tokens_seen": 132680355,
+      "step": 6253
+    },
+    {
+      "epoch": 0.7519990380568748,
+      "grad_norm": 12.222096153516329,
+      "learning_rate": 6.110693247444018e-07,
+      "loss": 0.6843,
+      "num_input_tokens_seen": 132701910,
+      "step": 6254
+    },
+    {
+      "epoch": 0.752119280947514,
+      "grad_norm": 2.6435225226564913,
+      "learning_rate": 6.105089396303258e-07,
+      "loss": 0.8152,
+      "num_input_tokens_seen": 132720020,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7522395238381531,
+      "grad_norm": 2.543789548541477,
+      "learning_rate": 6.09948765300939e-07,
+      "loss": 0.753,
+      "num_input_tokens_seen": 132739085,
+      "step": 6256
+    },
+    {
+      "epoch": 0.7523597667287921,
+      "grad_norm": 2.436047192948676,
+      "learning_rate": 6.093888018412192e-07,
+      "loss": 0.83,
+      "num_input_tokens_seen": 132754995,
+      "step": 6257
+    },
+    {
+      "epoch": 0.7524800096194313,
+      "grad_norm": 0.7746759661345062,
+      "learning_rate": 6.088290493361119e-07,
+      "loss": 0.5791,
+      "num_input_tokens_seen": 132819600,
+      "step": 6258
+    },
+    {
+      "epoch": 0.7526002525100703,
+      "grad_norm": 3.3903081888929067,
+      "learning_rate": 6.082695078705322e-07,
+      "loss": 0.7157,
+      "num_input_tokens_seen": 132836800,
+      "step": 6259
+    },
+    {
+      "epoch": 0.7527204954007094,
+      "grad_norm": 5.823284895127585,
+      "learning_rate": 6.077101775293618e-07,
+      "loss": 0.6779,
+      "num_input_tokens_seen": 132855345,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7528407382913486,
+      "grad_norm": 2.5486632210317373,
+      "learning_rate": 6.071510583974504e-07,
+      "loss": 0.8224,
+      "num_input_tokens_seen": 132870250,
+      "step": 6261
+    },
+    {
+      "epoch": 0.7529609811819876,
+      "grad_norm": 2.0739337043893973,
+      "learning_rate": 6.065921505596161e-07,
+      "loss": 0.7117,
+      "num_input_tokens_seen": 132888250,
+      "step": 6262
+    },
+    {
+      "epoch": 0.7530812240726267,
+      "grad_norm": 2.030532307697674,
+      "learning_rate": 6.060334541006445e-07,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 132906465,
+      "step": 6263
+    },
+    {
+      "epoch": 0.7532014669632658,
+      "grad_norm": 1.5603893390424808,
+      "learning_rate": 6.054749691052896e-07,
+      "loss": 0.6772,
+      "num_input_tokens_seen": 132929175,
+      "step": 6264
+    },
+    {
+      "epoch": 0.7533217098539049,
+      "grad_norm": 2.6149667986905802,
+      "learning_rate": 6.049166956582732e-07,
+      "loss": 0.7295,
+      "num_input_tokens_seen": 132947160,
+      "step": 6265
+    },
+    {
+      "epoch": 0.753441952744544,
+      "grad_norm": 2.792615985038066,
+      "learning_rate": 6.043586338442841e-07,
+      "loss": 0.8612,
+      "num_input_tokens_seen": 132965935,
+      "step": 6266
+    },
+    {
+      "epoch": 0.7535621956351831,
+      "grad_norm": 1.4436900666320727,
+      "learning_rate": 6.038007837479815e-07,
+      "loss": 0.7231,
+      "num_input_tokens_seen": 132986760,
+      "step": 6267
+    },
+    {
+      "epoch": 0.7536824385258222,
+      "grad_norm": 2.3453243820403724,
+      "learning_rate": 6.032431454539897e-07,
+      "loss": 0.6303,
+      "num_input_tokens_seen": 133005325,
+      "step": 6268
+    },
+    {
+      "epoch": 0.7538026814164612,
+      "grad_norm": 2.2050019327520127,
+      "learning_rate": 6.026857190469022e-07,
+      "loss": 0.8037,
+      "num_input_tokens_seen": 133026800,
+      "step": 6269
+    },
+    {
+      "epoch": 0.7539229243071004,
+      "grad_norm": 2.9678922303506803,
+      "learning_rate": 6.021285046112794e-07,
+      "loss": 0.7417,
+      "num_input_tokens_seen": 133045640,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7540431671977395,
+      "grad_norm": 2.350658345750166,
+      "learning_rate": 6.015715022316516e-07,
+      "loss": 0.7505,
+      "num_input_tokens_seen": 133063340,
+      "step": 6271
+    },
+    {
+      "epoch": 0.7541634100883785,
+      "grad_norm": 2.8055025198113523,
+      "learning_rate": 6.010147119925154e-07,
+      "loss": 0.7783,
+      "num_input_tokens_seen": 133080815,
+      "step": 6272
+    },
+    {
+      "epoch": 0.7542836529790176,
+      "grad_norm": 2.3274050714884793,
+      "learning_rate": 6.004581339783348e-07,
+      "loss": 0.6489,
+      "num_input_tokens_seen": 133098855,
+      "step": 6273
+    },
+    {
+      "epoch": 0.7544038958696567,
+      "grad_norm": 2.9049557388342926,
+      "learning_rate": 5.999017682735419e-07,
+      "loss": 0.6763,
+      "num_input_tokens_seen": 133114965,
+      "step": 6274
+    },
+    {
+      "epoch": 0.7545241387602958,
+      "grad_norm": 2.2010233390423077,
+      "learning_rate": 5.993456149625382e-07,
+      "loss": 0.6582,
+      "num_input_tokens_seen": 133135835,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7546443816509348,
+      "grad_norm": 2.186178802353043,
+      "learning_rate": 5.987896741296909e-07,
+      "loss": 0.8092,
+      "num_input_tokens_seen": 133153295,
+      "step": 6276
+    },
+    {
+      "epoch": 0.754764624541574,
+      "grad_norm": 2.3978810453707293,
+      "learning_rate": 5.982339458593361e-07,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 133172955,
+      "step": 6277
+    },
+    {
+      "epoch": 0.7548848674322131,
+      "grad_norm": 3.6616735672445806,
+      "learning_rate": 5.976784302357773e-07,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 133193240,
+      "step": 6278
+    },
+    {
+      "epoch": 0.7550051103228521,
+      "grad_norm": 10.076436554867167,
+      "learning_rate": 5.971231273432855e-07,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 133212445,
+      "step": 6279
+    },
+    {
+      "epoch": 0.7551253532134913,
+      "grad_norm": 0.839676214768804,
+      "learning_rate": 5.965680372661e-07,
+      "loss": 0.5816,
+      "num_input_tokens_seen": 133269730,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7552455961041303,
+      "grad_norm": 1.9735119962533987,
+      "learning_rate": 5.960131600884273e-07,
+      "loss": 0.5609,
+      "num_input_tokens_seen": 133288720,
+      "step": 6281
+    },
+    {
+      "epoch": 0.7553658389947694,
+      "grad_norm": 2.4628861132618143,
+      "learning_rate": 5.954584958944413e-07,
+      "loss": 0.7513,
+      "num_input_tokens_seen": 133307105,
+      "step": 6282
+    },
+    {
+      "epoch": 0.7554860818854086,
+      "grad_norm": 2.63888668777687,
+      "learning_rate": 5.949040447682854e-07,
+      "loss": 0.8064,
+      "num_input_tokens_seen": 133326650,
+      "step": 6283
+    },
+    {
+      "epoch": 0.7556063247760476,
+      "grad_norm": 2.766594810246037,
+      "learning_rate": 5.943498067940686e-07,
+      "loss": 0.6852,
+      "num_input_tokens_seen": 133343395,
+      "step": 6284
+    },
+    {
+      "epoch": 0.7557265676666867,
+      "grad_norm": 1.899263174678764,
+      "learning_rate": 5.937957820558686e-07,
+      "loss": 0.8126,
+      "num_input_tokens_seen": 133362460,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7558468105573258,
+      "grad_norm": 0.9006774994163477,
+      "learning_rate": 5.932419706377296e-07,
+      "loss": 0.6776,
+      "num_input_tokens_seen": 133420485,
+      "step": 6286
+    },
+    {
+      "epoch": 0.7559670534479649,
+      "grad_norm": 2.6403849656575433,
+      "learning_rate": 5.92688372623666e-07,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 133438910,
+      "step": 6287
+    },
+    {
+      "epoch": 0.7560872963386039,
+      "grad_norm": 2.6468925376835033,
+      "learning_rate": 5.921349880976574e-07,
+      "loss": 0.7339,
+      "num_input_tokens_seen": 133456465,
+      "step": 6288
+    },
+    {
+      "epoch": 0.7562075392292431,
+      "grad_norm": 2.0771945384229156,
+      "learning_rate": 5.915818171436515e-07,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 133475520,
+      "step": 6289
+    },
+    {
+      "epoch": 0.7563277821198822,
+      "grad_norm": 2.3145208987277868,
+      "learning_rate": 5.910288598455637e-07,
+      "loss": 0.7385,
+      "num_input_tokens_seen": 133494590,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7564480250105212,
+      "grad_norm": 2.9164621711051555,
+      "learning_rate": 5.90476116287278e-07,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 133511910,
+      "step": 6291
+    },
+    {
+      "epoch": 0.7565682679011604,
+      "grad_norm": 2.0452487514656696,
+      "learning_rate": 5.899235865526448e-07,
+      "loss": 0.6733,
+      "num_input_tokens_seen": 133530925,
+      "step": 6292
+    },
+    {
+      "epoch": 0.7566885107917994,
+      "grad_norm": 1.704670311944046,
+      "learning_rate": 5.893712707254825e-07,
+      "loss": 0.8086,
+      "num_input_tokens_seen": 133548105,
+      "step": 6293
+    },
+    {
+      "epoch": 0.7568087536824385,
+      "grad_norm": 6.10587203076784,
+      "learning_rate": 5.888191688895769e-07,
+      "loss": 0.652,
+      "num_input_tokens_seen": 133565085,
+      "step": 6294
+    },
+    {
+      "epoch": 0.7569289965730777,
+      "grad_norm": 2.4617329476836214,
+      "learning_rate": 5.882672811286813e-07,
+      "loss": 0.6164,
+      "num_input_tokens_seen": 133581085,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7570492394637167,
+      "grad_norm": 2.1563886958390728,
+      "learning_rate": 5.877156075265166e-07,
+      "loss": 0.6989,
+      "num_input_tokens_seen": 133597070,
+      "step": 6296
+    },
+    {
+      "epoch": 0.7571694823543558,
+      "grad_norm": 2.975366519091935,
+      "learning_rate": 5.871641481667715e-07,
+      "loss": 0.6873,
+      "num_input_tokens_seen": 133611235,
+      "step": 6297
+    },
+    {
+      "epoch": 0.7572897252449949,
+      "grad_norm": 2.0131947287565217,
+      "learning_rate": 5.866129031331011e-07,
+      "loss": 0.8368,
+      "num_input_tokens_seen": 133630610,
+      "step": 6298
+    },
+    {
+      "epoch": 0.757409968135634,
+      "grad_norm": 3.063705082157713,
+      "learning_rate": 5.8606187250913e-07,
+      "loss": 0.8253,
+      "num_input_tokens_seen": 133648380,
+      "step": 6299
+    },
+    {
+      "epoch": 0.757530211026273,
+      "grad_norm": 2.2738609262671976,
+      "learning_rate": 5.855110563784488e-07,
+      "loss": 0.8367,
+      "num_input_tokens_seen": 133666635,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7576504539169122,
+      "grad_norm": 2.06961783322287,
+      "learning_rate": 5.849604548246156e-07,
+      "loss": 0.6371,
+      "num_input_tokens_seen": 133687465,
+      "step": 6301
+    },
+    {
+      "epoch": 0.7577706968075513,
+      "grad_norm": 2.8338165550698586,
+      "learning_rate": 5.844100679311559e-07,
+      "loss": 0.7941,
+      "num_input_tokens_seen": 133706145,
+      "step": 6302
+    },
+    {
+      "epoch": 0.7578909396981903,
+      "grad_norm": 2.2453796282595073,
+      "learning_rate": 5.838598957815637e-07,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 133723095,
+      "step": 6303
+    },
+    {
+      "epoch": 0.7580111825888295,
+      "grad_norm": 1.5239320188420045,
+      "learning_rate": 5.833099384592996e-07,
+      "loss": 0.8522,
+      "num_input_tokens_seen": 133743390,
+      "step": 6304
+    },
+    {
+      "epoch": 0.7581314254794685,
+      "grad_norm": 2.174260235563152,
+      "learning_rate": 5.827601960477913e-07,
+      "loss": 0.6993,
+      "num_input_tokens_seen": 133761035,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7582516683701076,
+      "grad_norm": 2.0347380545057954,
+      "learning_rate": 5.822106686304344e-07,
+      "loss": 0.7045,
+      "num_input_tokens_seen": 133780045,
+      "step": 6306
+    },
+    {
+      "epoch": 0.7583719112607467,
+      "grad_norm": 2.2108613534897463,
+      "learning_rate": 5.816613562905919e-07,
+      "loss": 0.5712,
+      "num_input_tokens_seen": 133800950,
+      "step": 6307
+    },
+    {
+      "epoch": 0.7584921541513858,
+      "grad_norm": 1.6758909541225415,
+      "learning_rate": 5.81112259111594e-07,
+      "loss": 0.7057,
+      "num_input_tokens_seen": 133821655,
+      "step": 6308
+    },
+    {
+      "epoch": 0.7586123970420249,
+      "grad_norm": 3.8633948154619615,
+      "learning_rate": 5.805633771767382e-07,
+      "loss": 0.7164,
+      "num_input_tokens_seen": 133838770,
+      "step": 6309
+    },
+    {
+      "epoch": 0.7587326399326639,
+      "grad_norm": 2.185368817915525,
+      "learning_rate": 5.800147105692888e-07,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 133858065,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7588528828233031,
+      "grad_norm": 2.0287861015399997,
+      "learning_rate": 5.794662593724795e-07,
+      "loss": 0.7918,
+      "num_input_tokens_seen": 133876790,
+      "step": 6311
+    },
+    {
+      "epoch": 0.7589731257139422,
+      "grad_norm": 3.0625557041677363,
+      "learning_rate": 5.789180236695091e-07,
+      "loss": 0.748,
+      "num_input_tokens_seen": 133893365,
+      "step": 6312
+    },
+    {
+      "epoch": 0.7590933686045812,
+      "grad_norm": 2.0374945047089454,
+      "learning_rate": 5.783700035435446e-07,
+      "loss": 0.8472,
+      "num_input_tokens_seen": 133911840,
+      "step": 6313
+    },
+    {
+      "epoch": 0.7592136114952204,
+      "grad_norm": 2.161303217221468,
+      "learning_rate": 5.778221990777197e-07,
+      "loss": 0.8308,
+      "num_input_tokens_seen": 133929300,
+      "step": 6314
+    },
+    {
+      "epoch": 0.7593338543858594,
+      "grad_norm": 2.0704439596362607,
+      "learning_rate": 5.772746103551372e-07,
+      "loss": 0.8179,
+      "num_input_tokens_seen": 133944415,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7594540972764985,
+      "grad_norm": 1.7474266966923675,
+      "learning_rate": 5.767272374588648e-07,
+      "loss": 0.7164,
+      "num_input_tokens_seen": 133965540,
+      "step": 6316
+    },
+    {
+      "epoch": 0.7595743401671377,
+      "grad_norm": 3.3065380836355547,
+      "learning_rate": 5.76180080471939e-07,
+      "loss": 0.7779,
+      "num_input_tokens_seen": 133988430,
+      "step": 6317
+    },
+    {
+      "epoch": 0.7596945830577767,
+      "grad_norm": 3.366626368727072,
+      "learning_rate": 5.756331394773623e-07,
+      "loss": 0.7111,
+      "num_input_tokens_seen": 134004365,
+      "step": 6318
+    },
+    {
+      "epoch": 0.7598148259484158,
+      "grad_norm": 2.30264668025299,
+      "learning_rate": 5.750864145581065e-07,
+      "loss": 0.7619,
+      "num_input_tokens_seen": 134023305,
+      "step": 6319
+    },
+    {
+      "epoch": 0.7599350688390549,
+      "grad_norm": 5.985573394851742,
+      "learning_rate": 5.745399057971085e-07,
+      "loss": 0.846,
+      "num_input_tokens_seen": 134044160,
+      "step": 6320
+    },
+    {
+      "epoch": 0.760055311729694,
+      "grad_norm": 2.2643468370251956,
+      "learning_rate": 5.739936132772738e-07,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 134062445,
+      "step": 6321
+    },
+    {
+      "epoch": 0.760175554620333,
+      "grad_norm": 2.8697174241943224,
+      "learning_rate": 5.734475370814737e-07,
+      "loss": 0.7444,
+      "num_input_tokens_seen": 134081845,
+      "step": 6322
+    },
+    {
+      "epoch": 0.7602957975109722,
+      "grad_norm": 1.90973632003252,
+      "learning_rate": 5.729016772925483e-07,
+      "loss": 0.7714,
+      "num_input_tokens_seen": 134103140,
+      "step": 6323
+    },
+    {
+      "epoch": 0.7604160404016113,
+      "grad_norm": 1.7974974356915343,
+      "learning_rate": 5.723560339933038e-07,
+      "loss": 0.7011,
+      "num_input_tokens_seen": 134123195,
+      "step": 6324
+    },
+    {
+      "epoch": 0.7605362832922503,
+      "grad_norm": 2.1259658033756375,
+      "learning_rate": 5.718106072665136e-07,
+      "loss": 0.641,
+      "num_input_tokens_seen": 134141500,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7606565261828895,
+      "grad_norm": 2.744882177130346,
+      "learning_rate": 5.712653971949184e-07,
+      "loss": 0.5986,
+      "num_input_tokens_seen": 134159340,
+      "step": 6326
+    },
+    {
+      "epoch": 0.7607767690735285,
+      "grad_norm": 2.6108161945061474,
+      "learning_rate": 5.707204038612268e-07,
+      "loss": 0.759,
+      "num_input_tokens_seen": 134176490,
+      "step": 6327
+    },
+    {
+      "epoch": 0.7608970119641676,
+      "grad_norm": 3.1723513692747405,
+      "learning_rate": 5.701756273481138e-07,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 134193630,
+      "step": 6328
+    },
+    {
+      "epoch": 0.7610172548548068,
+      "grad_norm": 1.6257986294807119,
+      "learning_rate": 5.696310677382212e-07,
+      "loss": 0.7305,
+      "num_input_tokens_seen": 134214745,
+      "step": 6329
+    },
+    {
+      "epoch": 0.7611374977454458,
+      "grad_norm": 0.8682098760207265,
+      "learning_rate": 5.690867251141576e-07,
+      "loss": 0.6511,
+      "num_input_tokens_seen": 134281120,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7612577406360849,
+      "grad_norm": 2.213534489430463,
+      "learning_rate": 5.685425995585009e-07,
+      "loss": 0.9084,
+      "num_input_tokens_seen": 134298765,
+      "step": 6331
+    },
+    {
+      "epoch": 0.761377983526724,
+      "grad_norm": 0.7863598053935742,
+      "learning_rate": 5.679986911537935e-07,
+      "loss": 0.6236,
+      "num_input_tokens_seen": 134366015,
+      "step": 6332
+    },
+    {
+      "epoch": 0.7614982264173631,
+      "grad_norm": 2.126246227801962,
+      "learning_rate": 5.674549999825462e-07,
+      "loss": 0.6739,
+      "num_input_tokens_seen": 134388550,
+      "step": 6333
+    },
+    {
+      "epoch": 0.7616184693080021,
+      "grad_norm": 1.0510224346764374,
+      "learning_rate": 5.669115261272359e-07,
+      "loss": 0.7738,
+      "num_input_tokens_seen": 134448590,
+      "step": 6334
+    },
+    {
+      "epoch": 0.7617387121986413,
+      "grad_norm": 2.3599441280967044,
+      "learning_rate": 5.663682696703081e-07,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 134466575,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7618589550892804,
+      "grad_norm": 2.11980395708863,
+      "learning_rate": 5.65825230694174e-07,
+      "loss": 0.8251,
+      "num_input_tokens_seen": 134485615,
+      "step": 6336
+    },
+    {
+      "epoch": 0.7619791979799194,
+      "grad_norm": 2.8284980776438275,
+      "learning_rate": 5.65282409281212e-07,
+      "loss": 0.758,
+      "num_input_tokens_seen": 134502800,
+      "step": 6337
+    },
+    {
+      "epoch": 0.7620994408705585,
+      "grad_norm": 2.8148622652435833,
+      "learning_rate": 5.64739805513768e-07,
+      "loss": 0.6917,
+      "num_input_tokens_seen": 134520065,
+      "step": 6338
+    },
+    {
+      "epoch": 0.7622196837611976,
+      "grad_norm": 0.8457363162198092,
+      "learning_rate": 5.641974194741541e-07,
+      "loss": 0.5855,
+      "num_input_tokens_seen": 134575470,
+      "step": 6339
+    },
+    {
+      "epoch": 0.7623399266518367,
+      "grad_norm": 0.7661169667163336,
+      "learning_rate": 5.636552512446502e-07,
+      "loss": 0.6509,
+      "num_input_tokens_seen": 134636245,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7624601695424758,
+      "grad_norm": 2.08578809747483,
+      "learning_rate": 5.631133009075027e-07,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 134655150,
+      "step": 6341
+    },
+    {
+      "epoch": 0.7625804124331149,
+      "grad_norm": 2.231950114295946,
+      "learning_rate": 5.625715685449242e-07,
+      "loss": 0.6905,
+      "num_input_tokens_seen": 134672975,
+      "step": 6342
+    },
+    {
+      "epoch": 0.762700655323754,
+      "grad_norm": 1.7613145502041365,
+      "learning_rate": 5.620300542390966e-07,
+      "loss": 0.7133,
+      "num_input_tokens_seen": 134693740,
+      "step": 6343
+    },
+    {
+      "epoch": 0.762820898214393,
+      "grad_norm": 2.313018382918317,
+      "learning_rate": 5.614887580721666e-07,
+      "loss": 0.8484,
+      "num_input_tokens_seen": 134713605,
+      "step": 6344
+    },
+    {
+      "epoch": 0.7629411411050322,
+      "grad_norm": 2.730684423706499,
+      "learning_rate": 5.609476801262481e-07,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 134728185,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7630613839956712,
+      "grad_norm": 3.195982214878335,
+      "learning_rate": 5.604068204834215e-07,
+      "loss": 0.6444,
+      "num_input_tokens_seen": 134744800,
+      "step": 6346
+    },
+    {
+      "epoch": 0.7631816268863103,
+      "grad_norm": 3.463030339719599,
+      "learning_rate": 5.598661792257367e-07,
+      "loss": 0.7472,
+      "num_input_tokens_seen": 134761565,
+      "step": 6347
+    },
+    {
+      "epoch": 0.7633018697769495,
+      "grad_norm": 2.1872474599821197,
+      "learning_rate": 5.593257564352071e-07,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 134779725,
+      "step": 6348
+    },
+    {
+      "epoch": 0.7634221126675885,
+      "grad_norm": 1.6377286611500093,
+      "learning_rate": 5.58785552193815e-07,
+      "loss": 0.7506,
+      "num_input_tokens_seen": 134799690,
+      "step": 6349
+    },
+    {
+      "epoch": 0.7635423555582276,
+      "grad_norm": 2.1302826171642923,
+      "learning_rate": 5.582455665835086e-07,
+      "loss": 0.751,
+      "num_input_tokens_seen": 134819705,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7636625984488667,
+      "grad_norm": 5.179631870952496,
+      "learning_rate": 5.577057996862036e-07,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 134837050,
+      "step": 6351
+    },
+    {
+      "epoch": 0.7637828413395058,
+      "grad_norm": 2.4177652355882375,
+      "learning_rate": 5.571662515837818e-07,
+      "loss": 0.7536,
+      "num_input_tokens_seen": 134858730,
+      "step": 6352
+    },
+    {
+      "epoch": 0.7639030842301449,
+      "grad_norm": 2.0240989134851124,
+      "learning_rate": 5.566269223580926e-07,
+      "loss": 0.8352,
+      "num_input_tokens_seen": 134880160,
+      "step": 6353
+    },
+    {
+      "epoch": 0.764023327120784,
+      "grad_norm": 1.6718112131723575,
+      "learning_rate": 5.560878120909511e-07,
+      "loss": 0.7403,
+      "num_input_tokens_seen": 134902480,
+      "step": 6354
+    },
+    {
+      "epoch": 0.7641435700114231,
+      "grad_norm": 0.9038247306596882,
+      "learning_rate": 5.55548920864141e-07,
+      "loss": 0.6478,
+      "num_input_tokens_seen": 134962855,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7642638129020621,
+      "grad_norm": 1.8227317918894237,
+      "learning_rate": 5.550102487594113e-07,
+      "loss": 0.7637,
+      "num_input_tokens_seen": 134981245,
+      "step": 6356
+    },
+    {
+      "epoch": 0.7643840557927013,
+      "grad_norm": 1.9256008769006938,
+      "learning_rate": 5.544717958584776e-07,
+      "loss": 0.7111,
+      "num_input_tokens_seen": 135001035,
+      "step": 6357
+    },
+    {
+      "epoch": 0.7645042986833404,
+      "grad_norm": 2.6707409407821396,
+      "learning_rate": 5.539335622430227e-07,
+      "loss": 0.8288,
+      "num_input_tokens_seen": 135019375,
+      "step": 6358
+    },
+    {
+      "epoch": 0.7646245415739794,
+      "grad_norm": 2.0996248617636364,
+      "learning_rate": 5.533955479946975e-07,
+      "loss": 0.7407,
+      "num_input_tokens_seen": 135037875,
+      "step": 6359
+    },
+    {
+      "epoch": 0.7647447844646186,
+      "grad_norm": 0.8931092337628109,
+      "learning_rate": 5.528577531951173e-07,
+      "loss": 0.6762,
+      "num_input_tokens_seen": 135098000,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7648650273552576,
+      "grad_norm": 2.494854709978254,
+      "learning_rate": 5.523201779258653e-07,
+      "loss": 0.7418,
+      "num_input_tokens_seen": 135116695,
+      "step": 6361
+    },
+    {
+      "epoch": 0.7649852702458967,
+      "grad_norm": 1.989201457230547,
+      "learning_rate": 5.517828222684906e-07,
+      "loss": 0.8398,
+      "num_input_tokens_seen": 135137070,
+      "step": 6362
+    },
+    {
+      "epoch": 0.7651055131365359,
+      "grad_norm": 0.793777499821597,
+      "learning_rate": 5.512456863045109e-07,
+      "loss": 0.6047,
+      "num_input_tokens_seen": 135197480,
+      "step": 6363
+    },
+    {
+      "epoch": 0.7652257560271749,
+      "grad_norm": 2.2512515513123526,
+      "learning_rate": 5.507087701154089e-07,
+      "loss": 0.7335,
+      "num_input_tokens_seen": 135217120,
+      "step": 6364
+    },
+    {
+      "epoch": 0.765345998917814,
+      "grad_norm": 2.4449229420965257,
+      "learning_rate": 5.50172073782634e-07,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 135234820,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7654662418084531,
+      "grad_norm": 2.469374293426904,
+      "learning_rate": 5.49635597387603e-07,
+      "loss": 0.8621,
+      "num_input_tokens_seen": 135253795,
+      "step": 6366
+    },
+    {
+      "epoch": 0.7655864846990922,
+      "grad_norm": 1.9648030689397533,
+      "learning_rate": 5.490993410116984e-07,
+      "loss": 0.7087,
+      "num_input_tokens_seen": 135276505,
+      "step": 6367
+    },
+    {
+      "epoch": 0.7657067275897312,
+      "grad_norm": 1.9197839390277576,
+      "learning_rate": 5.485633047362704e-07,
+      "loss": 0.7009,
+      "num_input_tokens_seen": 135298230,
+      "step": 6368
+    },
+    {
+      "epoch": 0.7658269704803703,
+      "grad_norm": 2.154869834587019,
+      "learning_rate": 5.480274886426349e-07,
+      "loss": 0.7773,
+      "num_input_tokens_seen": 135314590,
+      "step": 6369
+    },
+    {
+      "epoch": 0.7659472133710095,
+      "grad_norm": 2.707946763944997,
+      "learning_rate": 5.474918928120744e-07,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 135330805,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7660674562616485,
+      "grad_norm": 1.8624321786494242,
+      "learning_rate": 5.469565173258392e-07,
+      "loss": 0.8722,
+      "num_input_tokens_seen": 135349040,
+      "step": 6371
+    },
+    {
+      "epoch": 0.7661876991522876,
+      "grad_norm": 1.9315928820242891,
+      "learning_rate": 5.464213622651454e-07,
+      "loss": 0.6391,
+      "num_input_tokens_seen": 135366575,
+      "step": 6372
+    },
+    {
+      "epoch": 0.7663079420429267,
+      "grad_norm": 2.131534258398,
+      "learning_rate": 5.458864277111753e-07,
+      "loss": 0.8352,
+      "num_input_tokens_seen": 135384130,
+      "step": 6373
+    },
+    {
+      "epoch": 0.7664281849335658,
+      "grad_norm": 2.7329928404504593,
+      "learning_rate": 5.453517137450769e-07,
+      "loss": 0.6867,
+      "num_input_tokens_seen": 135400425,
+      "step": 6374
+    },
+    {
+      "epoch": 0.7665484278242048,
+      "grad_norm": 1.9557421048202888,
+      "learning_rate": 5.448172204479677e-07,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 135419425,
+      "step": 6375
+    },
+    {
+      "epoch": 0.766668670714844,
+      "grad_norm": 2.1165919264466795,
+      "learning_rate": 5.442829479009294e-07,
+      "loss": 0.7421,
+      "num_input_tokens_seen": 135437925,
+      "step": 6376
+    },
+    {
+      "epoch": 0.7667889136054831,
+      "grad_norm": 2.7468194789400147,
+      "learning_rate": 5.437488961850103e-07,
+      "loss": 0.7134,
+      "num_input_tokens_seen": 135457445,
+      "step": 6377
+    },
+    {
+      "epoch": 0.7669091564961221,
+      "grad_norm": 2.3414434543991227,
+      "learning_rate": 5.432150653812253e-07,
+      "loss": 0.7527,
+      "num_input_tokens_seen": 135477200,
+      "step": 6378
+    },
+    {
+      "epoch": 0.7670293993867613,
+      "grad_norm": 2.6579669881075523,
+      "learning_rate": 5.42681455570557e-07,
+      "loss": 0.8282,
+      "num_input_tokens_seen": 135493450,
+      "step": 6379
+    },
+    {
+      "epoch": 0.7671496422774003,
+      "grad_norm": 2.179744610404988,
+      "learning_rate": 5.421480668339533e-07,
+      "loss": 0.6442,
+      "num_input_tokens_seen": 135512415,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7672698851680394,
+      "grad_norm": 2.212817525249641,
+      "learning_rate": 5.416148992523289e-07,
+      "loss": 0.7572,
+      "num_input_tokens_seen": 135530710,
+      "step": 6381
+    },
+    {
+      "epoch": 0.7673901280586786,
+      "grad_norm": 1.8596636510581415,
+      "learning_rate": 5.410819529065644e-07,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 135548385,
+      "step": 6382
+    },
+    {
+      "epoch": 0.7675103709493176,
+      "grad_norm": 2.8472899830798735,
+      "learning_rate": 5.405492278775079e-07,
+      "loss": 0.6477,
+      "num_input_tokens_seen": 135567885,
+      "step": 6383
+    },
+    {
+      "epoch": 0.7676306138399567,
+      "grad_norm": 4.507056836227131,
+      "learning_rate": 5.400167242459732e-07,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 135586565,
+      "step": 6384
+    },
+    {
+      "epoch": 0.7677508567305958,
+      "grad_norm": 1.7234290338446134,
+      "learning_rate": 5.394844420927405e-07,
+      "loss": 0.7978,
+      "num_input_tokens_seen": 135605895,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7678710996212349,
+      "grad_norm": 2.263809891326768,
+      "learning_rate": 5.389523814985562e-07,
+      "loss": 0.7201,
+      "num_input_tokens_seen": 135625035,
+      "step": 6386
+    },
+    {
+      "epoch": 0.767991342511874,
+      "grad_norm": 8.227022186249583,
+      "learning_rate": 5.384205425441344e-07,
+      "loss": 0.754,
+      "num_input_tokens_seen": 135645665,
+      "step": 6387
+    },
+    {
+      "epoch": 0.7681115854025131,
+      "grad_norm": 2.6188421603062584,
+      "learning_rate": 5.378889253101542e-07,
+      "loss": 0.8362,
+      "num_input_tokens_seen": 135665940,
+      "step": 6388
+    },
+    {
+      "epoch": 0.7682318282931522,
+      "grad_norm": 1.7735668920390248,
+      "learning_rate": 5.373575298772617e-07,
+      "loss": 0.7951,
+      "num_input_tokens_seen": 135684780,
+      "step": 6389
+    },
+    {
+      "epoch": 0.7683520711837912,
+      "grad_norm": 0.7374682030310578,
+      "learning_rate": 5.368263563260682e-07,
+      "loss": 0.62,
+      "num_input_tokens_seen": 135749635,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7684723140744304,
+      "grad_norm": 2.557667045463059,
+      "learning_rate": 5.362954047371537e-07,
+      "loss": 0.6383,
+      "num_input_tokens_seen": 135768465,
+      "step": 6391
+    },
+    {
+      "epoch": 0.7685925569650695,
+      "grad_norm": 2.499681258431957,
+      "learning_rate": 5.357646751910627e-07,
+      "loss": 0.7163,
+      "num_input_tokens_seen": 135789365,
+      "step": 6392
+    },
+    {
+      "epoch": 0.7687127998557085,
+      "grad_norm": 4.272197991257629,
+      "learning_rate": 5.352341677683061e-07,
+      "loss": 0.7948,
+      "num_input_tokens_seen": 135810385,
+      "step": 6393
+    },
+    {
+      "epoch": 0.7688330427463477,
+      "grad_norm": 1.995636035652429,
+      "learning_rate": 5.347038825493617e-07,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 135831635,
+      "step": 6394
+    },
+    {
+      "epoch": 0.7689532856369867,
+      "grad_norm": 2.6701079900926112,
+      "learning_rate": 5.341738196146732e-07,
+      "loss": 0.6731,
+      "num_input_tokens_seen": 135849700,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7690735285276258,
+      "grad_norm": 2.459553700301667,
+      "learning_rate": 5.33643979044651e-07,
+      "loss": 0.728,
+      "num_input_tokens_seen": 135868520,
+      "step": 6396
+    },
+    {
+      "epoch": 0.769193771418265,
+      "grad_norm": 2.0732176520505803,
+      "learning_rate": 5.331143609196711e-07,
+      "loss": 0.6224,
+      "num_input_tokens_seen": 135892055,
+      "step": 6397
+    },
+    {
+      "epoch": 0.769314014308904,
+      "grad_norm": 1.976923347515186,
+      "learning_rate": 5.325849653200758e-07,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 135915725,
+      "step": 6398
+    },
+    {
+      "epoch": 0.7694342571995431,
+      "grad_norm": 2.6583830368322254,
+      "learning_rate": 5.32055792326175e-07,
+      "loss": 0.7563,
+      "num_input_tokens_seen": 135933870,
+      "step": 6399
+    },
+    {
+      "epoch": 0.7695545000901821,
+      "grad_norm": 3.6193365256209304,
+      "learning_rate": 5.315268420182437e-07,
+      "loss": 0.7229,
+      "num_input_tokens_seen": 135952265,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7696747429808213,
+      "grad_norm": 2.0388274747217854,
+      "learning_rate": 5.309981144765225e-07,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 135972130,
+      "step": 6401
+    },
+    {
+      "epoch": 0.7697949858714603,
+      "grad_norm": 2.6418384952674465,
+      "learning_rate": 5.304696097812191e-07,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 135988450,
+      "step": 6402
+    },
+    {
+      "epoch": 0.7699152287620994,
+      "grad_norm": 3.8519158979226367,
+      "learning_rate": 5.299413280125078e-07,
+      "loss": 0.5921,
+      "num_input_tokens_seen": 136006480,
+      "step": 6403
+    },
+    {
+      "epoch": 0.7700354716527386,
+      "grad_norm": 5.576130817088323,
+      "learning_rate": 5.294132692505284e-07,
+      "loss": 0.7258,
+      "num_input_tokens_seen": 136024610,
+      "step": 6404
+    },
+    {
+      "epoch": 0.7701557145433776,
+      "grad_norm": 2.350712528257476,
+      "learning_rate": 5.288854335753868e-07,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 136042590,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7702759574340167,
+      "grad_norm": 3.4448411809848025,
+      "learning_rate": 5.283578210671545e-07,
+      "loss": 0.7575,
+      "num_input_tokens_seen": 136064550,
+      "step": 6406
+    },
+    {
+      "epoch": 0.7703962003246558,
+      "grad_norm": 2.393596155989619,
+      "learning_rate": 5.278304318058713e-07,
+      "loss": 0.7614,
+      "num_input_tokens_seen": 136082125,
+      "step": 6407
+    },
+    {
+      "epoch": 0.7705164432152949,
+      "grad_norm": 2.2142736632796027,
+      "learning_rate": 5.273032658715411e-07,
+      "loss": 0.7915,
+      "num_input_tokens_seen": 136104655,
+      "step": 6408
+    },
+    {
+      "epoch": 0.7706366861059339,
+      "grad_norm": 6.339771914321342,
+      "learning_rate": 5.267763233441347e-07,
+      "loss": 0.7664,
+      "num_input_tokens_seen": 136125005,
+      "step": 6409
+    },
+    {
+      "epoch": 0.7707569289965731,
+      "grad_norm": 2.4807597346937857,
+      "learning_rate": 5.262496043035885e-07,
+      "loss": 0.694,
+      "num_input_tokens_seen": 136143230,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7708771718872122,
+      "grad_norm": 2.373775723891977,
+      "learning_rate": 5.257231088298057e-07,
+      "loss": 0.7752,
+      "num_input_tokens_seen": 136161360,
+      "step": 6411
+    },
+    {
+      "epoch": 0.7709974147778512,
+      "grad_norm": 1.2244398579784608,
+      "learning_rate": 5.25196837002655e-07,
+      "loss": 0.5723,
+      "num_input_tokens_seen": 136220790,
+      "step": 6412
+    },
+    {
+      "epoch": 0.7711176576684904,
+      "grad_norm": 2.117101016430545,
+      "learning_rate": 5.246707889019715e-07,
+      "loss": 0.6871,
+      "num_input_tokens_seen": 136243600,
+      "step": 6413
+    },
+    {
+      "epoch": 0.7712379005591294,
+      "grad_norm": 2.4156496445622158,
+      "learning_rate": 5.241449646075557e-07,
+      "loss": 0.681,
+      "num_input_tokens_seen": 136266545,
+      "step": 6414
+    },
+    {
+      "epoch": 0.7713581434497685,
+      "grad_norm": 5.062791395610434,
+      "learning_rate": 5.236193641991762e-07,
+      "loss": 0.7196,
+      "num_input_tokens_seen": 136284195,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7714783863404077,
+      "grad_norm": 3.5130192356481333,
+      "learning_rate": 5.23093987756565e-07,
+      "loss": 0.7024,
+      "num_input_tokens_seen": 136302610,
+      "step": 6416
+    },
+    {
+      "epoch": 0.7715986292310467,
+      "grad_norm": 1.9611568210320338,
+      "learning_rate": 5.225688353594217e-07,
+      "loss": 0.7558,
+      "num_input_tokens_seen": 136321960,
+      "step": 6417
+    },
+    {
+      "epoch": 0.7717188721216858,
+      "grad_norm": 3.1634505213160145,
+      "learning_rate": 5.220439070874108e-07,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 136340920,
+      "step": 6418
+    },
+    {
+      "epoch": 0.7718391150123249,
+      "grad_norm": 2.0177819571696576,
+      "learning_rate": 5.215192030201645e-07,
+      "loss": 0.7116,
+      "num_input_tokens_seen": 136361630,
+      "step": 6419
+    },
+    {
+      "epoch": 0.771959357902964,
+      "grad_norm": 3.428402878299086,
+      "learning_rate": 5.209947232372798e-07,
+      "loss": 0.8588,
+      "num_input_tokens_seen": 136378840,
+      "step": 6420
+    },
+    {
+      "epoch": 0.772079600793603,
+      "grad_norm": 1.8664875569644048,
+      "learning_rate": 5.204704678183196e-07,
+      "loss": 0.8009,
+      "num_input_tokens_seen": 136397295,
+      "step": 6421
+    },
+    {
+      "epoch": 0.7721998436842422,
+      "grad_norm": 2.2075968706803404,
+      "learning_rate": 5.199464368428124e-07,
+      "loss": 0.8419,
+      "num_input_tokens_seen": 136414145,
+      "step": 6422
+    },
+    {
+      "epoch": 0.7723200865748813,
+      "grad_norm": 1.910598148910355,
+      "learning_rate": 5.194226303902546e-07,
+      "loss": 0.68,
+      "num_input_tokens_seen": 136433600,
+      "step": 6423
+    },
+    {
+      "epoch": 0.7724403294655203,
+      "grad_norm": 1.8156191521669256,
+      "learning_rate": 5.188990485401066e-07,
+      "loss": 0.706,
+      "num_input_tokens_seen": 136452525,
+      "step": 6424
+    },
+    {
+      "epoch": 0.7725605723561595,
+      "grad_norm": 2.325674322472472,
+      "learning_rate": 5.183756913717958e-07,
+      "loss": 0.8531,
+      "num_input_tokens_seen": 136472020,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7726808152467985,
+      "grad_norm": 3.2641462093122073,
+      "learning_rate": 5.178525589647136e-07,
+      "loss": 0.7309,
+      "num_input_tokens_seen": 136493380,
+      "step": 6426
+    },
+    {
+      "epoch": 0.7728010581374376,
+      "grad_norm": 2.232113012897507,
+      "learning_rate": 5.173296513982201e-07,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 136511625,
+      "step": 6427
+    },
+    {
+      "epoch": 0.7729213010280768,
+      "grad_norm": 3.3984708156592336,
+      "learning_rate": 5.168069687516398e-07,
+      "loss": 0.6484,
+      "num_input_tokens_seen": 136531115,
+      "step": 6428
+    },
+    {
+      "epoch": 0.7730415439187158,
+      "grad_norm": 2.461511298269615,
+      "learning_rate": 5.16284511104263e-07,
+      "loss": 0.7127,
+      "num_input_tokens_seen": 136549970,
+      "step": 6429
+    },
+    {
+      "epoch": 0.7731617868093549,
+      "grad_norm": 3.394391373177479,
+      "learning_rate": 5.157622785353457e-07,
+      "loss": 0.8005,
+      "num_input_tokens_seen": 136567805,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7732820296999939,
+      "grad_norm": 0.6800467868587783,
+      "learning_rate": 5.152402711241113e-07,
+      "loss": 0.6218,
+      "num_input_tokens_seen": 136635430,
+      "step": 6431
+    },
+    {
+      "epoch": 0.7734022725906331,
+      "grad_norm": 2.0045139502992817,
+      "learning_rate": 5.147184889497471e-07,
+      "loss": 0.8312,
+      "num_input_tokens_seen": 136654620,
+      "step": 6432
+    },
+    {
+      "epoch": 0.7735225154812722,
+      "grad_norm": 3.5323197025817774,
+      "learning_rate": 5.141969320914072e-07,
+      "loss": 0.7902,
+      "num_input_tokens_seen": 136671845,
+      "step": 6433
+    },
+    {
+      "epoch": 0.7736427583719112,
+      "grad_norm": 3.749421956043653,
+      "learning_rate": 5.136756006282108e-07,
+      "loss": 0.6299,
+      "num_input_tokens_seen": 136690230,
+      "step": 6434
+    },
+    {
+      "epoch": 0.7737630012625504,
+      "grad_norm": 2.612260793205545,
+      "learning_rate": 5.131544946392446e-07,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 136705230,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7738832441531894,
+      "grad_norm": 3.074551949162788,
+      "learning_rate": 5.126336142035592e-07,
+      "loss": 0.6384,
+      "num_input_tokens_seen": 136724985,
+      "step": 6436
+    },
+    {
+      "epoch": 0.7740034870438285,
+      "grad_norm": 3.385516497942366,
+      "learning_rate": 5.121129594001721e-07,
+      "loss": 0.7123,
+      "num_input_tokens_seen": 136738970,
+      "step": 6437
+    },
+    {
+      "epoch": 0.7741237299344677,
+      "grad_norm": 1.9370530758292988,
+      "learning_rate": 5.115925303080661e-07,
+      "loss": 0.8106,
+      "num_input_tokens_seen": 136758400,
+      "step": 6438
+    },
+    {
+      "epoch": 0.7742439728251067,
+      "grad_norm": 3.7438483385105057,
+      "learning_rate": 5.110723270061899e-07,
+      "loss": 0.7855,
+      "num_input_tokens_seen": 136774610,
+      "step": 6439
+    },
+    {
+      "epoch": 0.7743642157157458,
+      "grad_norm": 2.4063009868142498,
+      "learning_rate": 5.105523495734576e-07,
+      "loss": 0.7914,
+      "num_input_tokens_seen": 136791730,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7744844586063849,
+      "grad_norm": 1.5599379362955605,
+      "learning_rate": 5.100325980887499e-07,
+      "loss": 0.736,
+      "num_input_tokens_seen": 136811375,
+      "step": 6441
+    },
+    {
+      "epoch": 0.774604701497024,
+      "grad_norm": 2.175839337697373,
+      "learning_rate": 5.095130726309116e-07,
+      "loss": 0.8241,
+      "num_input_tokens_seen": 136831270,
+      "step": 6442
+    },
+    {
+      "epoch": 0.774724944387663,
+      "grad_norm": 0.8999989931838549,
+      "learning_rate": 5.089937732787559e-07,
+      "loss": 0.6675,
+      "num_input_tokens_seen": 136895550,
+      "step": 6443
+    },
+    {
+      "epoch": 0.7748451872783022,
+      "grad_norm": 4.730448172191009,
+      "learning_rate": 5.084747001110592e-07,
+      "loss": 0.6623,
+      "num_input_tokens_seen": 136914895,
+      "step": 6444
+    },
+    {
+      "epoch": 0.7749654301689413,
+      "grad_norm": 1.9356166607084275,
+      "learning_rate": 5.079558532065646e-07,
+      "loss": 0.6898,
+      "num_input_tokens_seen": 136939320,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7750856730595803,
+      "grad_norm": 1.955323697211826,
+      "learning_rate": 5.074372326439802e-07,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 136962050,
+      "step": 6446
+    },
+    {
+      "epoch": 0.7752059159502195,
+      "grad_norm": 3.216235087068157,
+      "learning_rate": 5.069188385019814e-07,
+      "loss": 0.7354,
+      "num_input_tokens_seen": 136979470,
+      "step": 6447
+    },
+    {
+      "epoch": 0.7753261588408585,
+      "grad_norm": 3.7768500726745438,
+      "learning_rate": 5.064006708592077e-07,
+      "loss": 0.6157,
+      "num_input_tokens_seen": 136995435,
+      "step": 6448
+    },
+    {
+      "epoch": 0.7754464017314976,
+      "grad_norm": 2.8616665406475312,
+      "learning_rate": 5.058827297942647e-07,
+      "loss": 0.7502,
+      "num_input_tokens_seen": 137010260,
+      "step": 6449
+    },
+    {
+      "epoch": 0.7755666446221368,
+      "grad_norm": 4.600530760313983,
+      "learning_rate": 5.053650153857229e-07,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 137028990,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7756868875127758,
+      "grad_norm": 1.755189198874425,
+      "learning_rate": 5.048475277121207e-07,
+      "loss": 0.6943,
+      "num_input_tokens_seen": 137045925,
+      "step": 6451
+    },
+    {
+      "epoch": 0.7758071304034149,
+      "grad_norm": 1.9175797372736918,
+      "learning_rate": 5.043302668519598e-07,
+      "loss": 0.7641,
+      "num_input_tokens_seen": 137064980,
+      "step": 6452
+    },
+    {
+      "epoch": 0.775927373294054,
+      "grad_norm": 1.9843691088440378,
+      "learning_rate": 5.038132328837079e-07,
+      "loss": 0.7188,
+      "num_input_tokens_seen": 137083090,
+      "step": 6453
+    },
+    {
+      "epoch": 0.7760476161846931,
+      "grad_norm": 2.069665213275956,
+      "learning_rate": 5.032964258857993e-07,
+      "loss": 0.7323,
+      "num_input_tokens_seen": 137102905,
+      "step": 6454
+    },
+    {
+      "epoch": 0.7761678590753321,
+      "grad_norm": 2.7698051160249983,
+      "learning_rate": 5.027798459366329e-07,
+      "loss": 0.6818,
+      "num_input_tokens_seen": 137127990,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7762881019659713,
+      "grad_norm": 3.2233197082646234,
+      "learning_rate": 5.02263493114573e-07,
+      "loss": 0.6316,
+      "num_input_tokens_seen": 137149505,
+      "step": 6456
+    },
+    {
+      "epoch": 0.7764083448566104,
+      "grad_norm": 2.847564099744299,
+      "learning_rate": 5.017473674979509e-07,
+      "loss": 0.7634,
+      "num_input_tokens_seen": 137165250,
+      "step": 6457
+    },
+    {
+      "epoch": 0.7765285877472494,
+      "grad_norm": 0.7789420496969789,
+      "learning_rate": 5.01231469165061e-07,
+      "loss": 0.5974,
+      "num_input_tokens_seen": 137220795,
+      "step": 6458
+    },
+    {
+      "epoch": 0.7766488306378886,
+      "grad_norm": 1.892357678647014,
+      "learning_rate": 5.007157981941663e-07,
+      "loss": 0.6145,
+      "num_input_tokens_seen": 137285875,
+      "step": 6459
+    },
+    {
+      "epoch": 0.7767690735285276,
+      "grad_norm": 0.9138976902236747,
+      "learning_rate": 5.002003546634928e-07,
+      "loss": 0.6994,
+      "num_input_tokens_seen": 137341695,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7768893164191667,
+      "grad_norm": 2.6393848820023265,
+      "learning_rate": 4.996851386512331e-07,
+      "loss": 0.7559,
+      "num_input_tokens_seen": 137360120,
+      "step": 6461
+    },
+    {
+      "epoch": 0.7770095593098058,
+      "grad_norm": 1.7820880241764918,
+      "learning_rate": 4.991701502355444e-07,
+      "loss": 0.8253,
+      "num_input_tokens_seen": 137380305,
+      "step": 6462
+    },
+    {
+      "epoch": 0.7771298022004449,
+      "grad_norm": 1.9623330895479028,
+      "learning_rate": 4.986553894945511e-07,
+      "loss": 0.7586,
+      "num_input_tokens_seen": 137401235,
+      "step": 6463
+    },
+    {
+      "epoch": 0.777250045091084,
+      "grad_norm": 3.4448093715462997,
+      "learning_rate": 4.981408565063416e-07,
+      "loss": 0.8614,
+      "num_input_tokens_seen": 137420900,
+      "step": 6464
+    },
+    {
+      "epoch": 0.777370287981723,
+      "grad_norm": 2.429690082460131,
+      "learning_rate": 4.976265513489701e-07,
+      "loss": 0.7537,
+      "num_input_tokens_seen": 137440590,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7774905308723622,
+      "grad_norm": 2.7243379708507773,
+      "learning_rate": 4.971124741004558e-07,
+      "loss": 0.8043,
+      "num_input_tokens_seen": 137459310,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7776107737630013,
+      "grad_norm": 2.0105311463703166,
+      "learning_rate": 4.965986248387846e-07,
+      "loss": 0.7571,
+      "num_input_tokens_seen": 137477345,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7777310166536403,
+      "grad_norm": 1.6472463248806992,
+      "learning_rate": 4.960850036419073e-07,
+      "loss": 0.7627,
+      "num_input_tokens_seen": 137496165,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7778512595442795,
+      "grad_norm": 2.2132686495040432,
+      "learning_rate": 4.955716105877386e-07,
+      "loss": 0.7906,
+      "num_input_tokens_seen": 137514655,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7779715024349185,
+      "grad_norm": 2.0253479672148753,
+      "learning_rate": 4.950584457541598e-07,
+      "loss": 0.8271,
+      "num_input_tokens_seen": 137532840,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7780917453255576,
+      "grad_norm": 1.6716101950465727,
+      "learning_rate": 4.945455092190187e-07,
+      "loss": 0.8174,
+      "num_input_tokens_seen": 137553815,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7782119882161967,
+      "grad_norm": 0.73251137181082,
+      "learning_rate": 4.940328010601271e-07,
+      "loss": 0.5756,
+      "num_input_tokens_seen": 137618450,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7783322311068358,
+      "grad_norm": 2.4365344690004163,
+      "learning_rate": 4.935203213552621e-07,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 137641910,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7784524739974749,
+      "grad_norm": 2.28594232703087,
+      "learning_rate": 4.930080701821662e-07,
+      "loss": 0.6544,
+      "num_input_tokens_seen": 137659095,
+      "step": 6474
+    },
+    {
+      "epoch": 0.778572716888114,
+      "grad_norm": 2.0915009232905843,
+      "learning_rate": 4.92496047618548e-07,
+      "loss": 0.767,
+      "num_input_tokens_seen": 137678575,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7786929597787531,
+      "grad_norm": 2.6397487718854613,
+      "learning_rate": 4.919842537420811e-07,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 137695410,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7788132026693921,
+      "grad_norm": 3.4215923177787007,
+      "learning_rate": 4.91472688630404e-07,
+      "loss": 0.7938,
+      "num_input_tokens_seen": 137715870,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7789334455600313,
+      "grad_norm": 2.6457797796889,
+      "learning_rate": 4.909613523611198e-07,
+      "loss": 0.7383,
+      "num_input_tokens_seen": 137732470,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7790536884506704,
+      "grad_norm": 2.1519768105567856,
+      "learning_rate": 4.904502450117991e-07,
+      "loss": 0.743,
+      "num_input_tokens_seen": 137753150,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7791739313413094,
+      "grad_norm": 2.4090559515328316,
+      "learning_rate": 4.899393666599762e-07,
+      "loss": 0.7253,
+      "num_input_tokens_seen": 137769445,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7792941742319486,
+      "grad_norm": 4.288459063556391,
+      "learning_rate": 4.894287173831506e-07,
+      "loss": 0.7148,
+      "num_input_tokens_seen": 137785125,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7794144171225876,
+      "grad_norm": 2.774863982885005,
+      "learning_rate": 4.889182972587877e-07,
+      "loss": 0.8375,
+      "num_input_tokens_seen": 137804140,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7795346600132267,
+      "grad_norm": 2.317646733257671,
+      "learning_rate": 4.884081063643177e-07,
+      "loss": 0.6648,
+      "num_input_tokens_seen": 137822520,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7796549029038659,
+      "grad_norm": 0.9363650201554515,
+      "learning_rate": 4.87898144777136e-07,
+      "loss": 0.5688,
+      "num_input_tokens_seen": 137876620,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7797751457945049,
+      "grad_norm": 2.3436783865556774,
+      "learning_rate": 4.873884125746035e-07,
+      "loss": 0.7271,
+      "num_input_tokens_seen": 137898015,
+      "step": 6485
+    },
+    {
+      "epoch": 0.779895388685144,
+      "grad_norm": 2.882342492659769,
+      "learning_rate": 4.868789098340456e-07,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 137915640,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7800156315757831,
+      "grad_norm": 2.9574250572369576,
+      "learning_rate": 4.863696366327543e-07,
+      "loss": 0.7103,
+      "num_input_tokens_seen": 137934530,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7801358744664222,
+      "grad_norm": 2.7056854074565324,
+      "learning_rate": 4.85860593047986e-07,
+      "loss": 0.7758,
+      "num_input_tokens_seen": 137954315,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7802561173570612,
+      "grad_norm": 2.175799088930615,
+      "learning_rate": 4.853517791569617e-07,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 137976215,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7803763602477004,
+      "grad_norm": 2.070546437842019,
+      "learning_rate": 4.848431950368676e-07,
+      "loss": 0.6547,
+      "num_input_tokens_seen": 137998495,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7804966031383395,
+      "grad_norm": 0.7628411031412166,
+      "learning_rate": 4.843348407648569e-07,
+      "loss": 0.589,
+      "num_input_tokens_seen": 138059495,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7806168460289785,
+      "grad_norm": 2.1230722837911347,
+      "learning_rate": 4.838267164180457e-07,
+      "loss": 0.8174,
+      "num_input_tokens_seen": 138074885,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7807370889196176,
+      "grad_norm": 2.12148655330612,
+      "learning_rate": 4.833188220735161e-07,
+      "loss": 0.8411,
+      "num_input_tokens_seen": 138094275,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7808573318102567,
+      "grad_norm": 2.114751918650088,
+      "learning_rate": 4.828111578083147e-07,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 138110900,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7809775747008958,
+      "grad_norm": 3.0055485461883515,
+      "learning_rate": 4.823037236994549e-07,
+      "loss": 0.8025,
+      "num_input_tokens_seen": 138128785,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7810978175915348,
+      "grad_norm": 0.7794904402319447,
+      "learning_rate": 4.817965198239136e-07,
+      "loss": 0.5907,
+      "num_input_tokens_seen": 138194965,
+      "step": 6496
+    },
+    {
+      "epoch": 0.781218060482174,
+      "grad_norm": 3.105727893500193,
+      "learning_rate": 4.812895462586331e-07,
+      "loss": 0.7316,
+      "num_input_tokens_seen": 138212510,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7813383033728131,
+      "grad_norm": 3.0005052159454064,
+      "learning_rate": 4.807828030805207e-07,
+      "loss": 0.8173,
+      "num_input_tokens_seen": 138231220,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7814585462634521,
+      "grad_norm": 2.256648918402157,
+      "learning_rate": 4.802762903664495e-07,
+      "loss": 0.6722,
+      "num_input_tokens_seen": 138250120,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7815787891540913,
+      "grad_norm": 6.7909059596204955,
+      "learning_rate": 4.797700081932565e-07,
+      "loss": 0.7252,
+      "num_input_tokens_seen": 138267705,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7816990320447303,
+      "grad_norm": 2.8290493567344797,
+      "learning_rate": 4.792639566377448e-07,
+      "loss": 0.8132,
+      "num_input_tokens_seen": 138284835,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7818192749353694,
+      "grad_norm": 2.002130291538481,
+      "learning_rate": 4.78758135776681e-07,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 138304410,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7819395178260086,
+      "grad_norm": 4.537871945859692,
+      "learning_rate": 4.782525456867989e-07,
+      "loss": 0.7837,
+      "num_input_tokens_seen": 138322985,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7820597607166476,
+      "grad_norm": 1.8152758483097473,
+      "learning_rate": 4.777471864447959e-07,
+      "loss": 0.8226,
+      "num_input_tokens_seen": 138343445,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7821800036072867,
+      "grad_norm": 2.288650847117772,
+      "learning_rate": 4.772420581273344e-07,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 138360650,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7823002464979258,
+      "grad_norm": 2.4116869185501844,
+      "learning_rate": 4.7673716081104134e-07,
+      "loss": 0.7636,
+      "num_input_tokens_seen": 138380545,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7824204893885649,
+      "grad_norm": 1.8313960262361968,
+      "learning_rate": 4.762324945725102e-07,
+      "loss": 0.8379,
+      "num_input_tokens_seen": 138399710,
+      "step": 6507
+    },
+    {
+      "epoch": 0.782540732279204,
+      "grad_norm": 1.969528082994351,
+      "learning_rate": 4.7572805948829844e-07,
+      "loss": 0.7502,
+      "num_input_tokens_seen": 138419690,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7826609751698431,
+      "grad_norm": 1.8777736120112787,
+      "learning_rate": 4.7522385563492795e-07,
+      "loss": 0.7043,
+      "num_input_tokens_seen": 138439710,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7827812180604822,
+      "grad_norm": 2.557912300760641,
+      "learning_rate": 4.747198830888857e-07,
+      "loss": 0.694,
+      "num_input_tokens_seen": 138459300,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7829014609511212,
+      "grad_norm": 2.1996694476284353,
+      "learning_rate": 4.742161419266255e-07,
+      "loss": 0.6777,
+      "num_input_tokens_seen": 138478180,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7830217038417604,
+      "grad_norm": 2.5795612568071102,
+      "learning_rate": 4.7371263222456304e-07,
+      "loss": 0.6497,
+      "num_input_tokens_seen": 138495220,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7831419467323995,
+      "grad_norm": 0.8287779299546597,
+      "learning_rate": 4.732093540590807e-07,
+      "loss": 0.6469,
+      "num_input_tokens_seen": 138555810,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7832621896230385,
+      "grad_norm": 3.1322762495802983,
+      "learning_rate": 4.7270630750652475e-07,
+      "loss": 0.8185,
+      "num_input_tokens_seen": 138571485,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7833824325136777,
+      "grad_norm": 1.750614157913882,
+      "learning_rate": 4.7220349264320815e-07,
+      "loss": 0.8038,
+      "num_input_tokens_seen": 138590290,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7835026754043167,
+      "grad_norm": 0.7989667014940743,
+      "learning_rate": 4.71700909545407e-07,
+      "loss": 0.5939,
+      "num_input_tokens_seen": 138652955,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7836229182949558,
+      "grad_norm": 1.9881447994755221,
+      "learning_rate": 4.711985582893627e-07,
+      "loss": 0.7616,
+      "num_input_tokens_seen": 138671195,
+      "step": 6517
+    },
+    {
+      "epoch": 0.783743161185595,
+      "grad_norm": 2.002830156145519,
+      "learning_rate": 4.706964389512811e-07,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 138690950,
+      "step": 6518
+    },
+    {
+      "epoch": 0.783863404076234,
+      "grad_norm": 2.367181348621602,
+      "learning_rate": 4.701945516073345e-07,
+      "loss": 0.8669,
+      "num_input_tokens_seen": 138708145,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7839836469668731,
+      "grad_norm": 2.5491382204757387,
+      "learning_rate": 4.696928963336577e-07,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 138727295,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7841038898575122,
+      "grad_norm": 0.8887021278182728,
+      "learning_rate": 4.6919147320635224e-07,
+      "loss": 0.6364,
+      "num_input_tokens_seen": 138789725,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7842241327481513,
+      "grad_norm": 5.195074971803579,
+      "learning_rate": 4.6869028230148223e-07,
+      "loss": 0.7239,
+      "num_input_tokens_seen": 138807240,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7843443756387903,
+      "grad_norm": 3.9206470905707484,
+      "learning_rate": 4.6818932369507957e-07,
+      "loss": 0.6044,
+      "num_input_tokens_seen": 138826460,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7844646185294295,
+      "grad_norm": 2.5691218680073744,
+      "learning_rate": 4.676885974631386e-07,
+      "loss": 0.8762,
+      "num_input_tokens_seen": 138844540,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7845848614200686,
+      "grad_norm": 2.053408858133002,
+      "learning_rate": 4.67188103681619e-07,
+      "loss": 0.7984,
+      "num_input_tokens_seen": 138864045,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7847051043107076,
+      "grad_norm": 2.2931376213854793,
+      "learning_rate": 4.666878424264453e-07,
+      "loss": 0.6857,
+      "num_input_tokens_seen": 138883720,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7848253472013467,
+      "grad_norm": 1.8381895291923052,
+      "learning_rate": 4.661878137735069e-07,
+      "loss": 0.7316,
+      "num_input_tokens_seen": 138901630,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7849455900919858,
+      "grad_norm": 5.045677826744305,
+      "learning_rate": 4.656880177986577e-07,
+      "loss": 0.7439,
+      "num_input_tokens_seen": 138919895,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7850658329826249,
+      "grad_norm": 2.22738545499811,
+      "learning_rate": 4.6518845457771607e-07,
+      "loss": 0.8044,
+      "num_input_tokens_seen": 138938475,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7851860758732639,
+      "grad_norm": 2.391111716310641,
+      "learning_rate": 4.646891241864652e-07,
+      "loss": 0.7918,
+      "num_input_tokens_seen": 138956760,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7853063187639031,
+      "grad_norm": 2.1300471992083487,
+      "learning_rate": 4.6419002670065397e-07,
+      "loss": 0.7291,
+      "num_input_tokens_seen": 138976060,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7854265616545422,
+      "grad_norm": 2.310930386919058,
+      "learning_rate": 4.6369116219599445e-07,
+      "loss": 0.8458,
+      "num_input_tokens_seen": 138991765,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7855468045451812,
+      "grad_norm": 1.9163906904286707,
+      "learning_rate": 4.631925307481643e-07,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 139011300,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7856670474358204,
+      "grad_norm": 2.633796967505484,
+      "learning_rate": 4.6269413243280464e-07,
+      "loss": 0.7434,
+      "num_input_tokens_seen": 139030440,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7857872903264594,
+      "grad_norm": 2.9064898087318687,
+      "learning_rate": 4.621959673255236e-07,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 139046460,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7859075332170985,
+      "grad_norm": 2.5458334201319612,
+      "learning_rate": 4.6169803550189135e-07,
+      "loss": 0.8992,
+      "num_input_tokens_seen": 139061875,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7860277761077377,
+      "grad_norm": 9.78434268023342,
+      "learning_rate": 4.6120033703744424e-07,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 139080490,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7861480189983767,
+      "grad_norm": 2.2762288060598403,
+      "learning_rate": 4.6070287200768177e-07,
+      "loss": 0.778,
+      "num_input_tokens_seen": 139096890,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7862682618890158,
+      "grad_norm": 2.0139993938300944,
+      "learning_rate": 4.602056404880703e-07,
+      "loss": 0.7266,
+      "num_input_tokens_seen": 139114285,
+      "step": 6539
+    },
+    {
+      "epoch": 0.7863885047796549,
+      "grad_norm": 5.9096370463733034,
+      "learning_rate": 4.5970864255403883e-07,
+      "loss": 0.716,
+      "num_input_tokens_seen": 139135530,
+      "step": 6540
+    },
+    {
+      "epoch": 0.786508747670294,
+      "grad_norm": 2.607097106719047,
+      "learning_rate": 4.59211878280982e-07,
+      "loss": 0.8151,
+      "num_input_tokens_seen": 139154765,
+      "step": 6541
+    },
+    {
+      "epoch": 0.786628990560933,
+      "grad_norm": 2.378131763862302,
+      "learning_rate": 4.587153477442578e-07,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 139170800,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7867492334515722,
+      "grad_norm": 2.462117859171312,
+      "learning_rate": 4.582190510191899e-07,
+      "loss": 0.8153,
+      "num_input_tokens_seen": 139189180,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7868694763422113,
+      "grad_norm": 2.4289066476263925,
+      "learning_rate": 4.5772298818106625e-07,
+      "loss": 0.858,
+      "num_input_tokens_seen": 139204690,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7869897192328503,
+      "grad_norm": 4.065258907706046,
+      "learning_rate": 4.572271593051389e-07,
+      "loss": 0.7228,
+      "num_input_tokens_seen": 139221765,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7871099621234895,
+      "grad_norm": 1.684615657524531,
+      "learning_rate": 4.567315644666245e-07,
+      "loss": 0.767,
+      "num_input_tokens_seen": 139240280,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7872302050141285,
+      "grad_norm": 2.2279816301929496,
+      "learning_rate": 4.5623620374070507e-07,
+      "loss": 0.8383,
+      "num_input_tokens_seen": 139259315,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7873504479047676,
+      "grad_norm": 0.8115200289333083,
+      "learning_rate": 4.557410772025263e-07,
+      "loss": 0.6282,
+      "num_input_tokens_seen": 139320985,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7874706907954068,
+      "grad_norm": 18.92609819888793,
+      "learning_rate": 4.5524618492719803e-07,
+      "loss": 0.655,
+      "num_input_tokens_seen": 139339925,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7875909336860458,
+      "grad_norm": 1.46626843560371,
+      "learning_rate": 4.54751526989795e-07,
+      "loss": 0.7739,
+      "num_input_tokens_seen": 139361485,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7877111765766849,
+      "grad_norm": 2.56102401660103,
+      "learning_rate": 4.5425710346535706e-07,
+      "loss": 0.7876,
+      "num_input_tokens_seen": 139379150,
+      "step": 6551
+    },
+    {
+      "epoch": 0.787831419467324,
+      "grad_norm": 2.3511993593764,
+      "learning_rate": 4.537629144288877e-07,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 139396325,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7879516623579631,
+      "grad_norm": 2.761978460632437,
+      "learning_rate": 4.5326895995535477e-07,
+      "loss": 0.7415,
+      "num_input_tokens_seen": 139414945,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7880719052486022,
+      "grad_norm": 3.343338701952212,
+      "learning_rate": 4.527752401196907e-07,
+      "loss": 0.8386,
+      "num_input_tokens_seen": 139432680,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7881921481392413,
+      "grad_norm": 2.5455675086628125,
+      "learning_rate": 4.5228175499679254e-07,
+      "loss": 0.6679,
+      "num_input_tokens_seen": 139451985,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7883123910298804,
+      "grad_norm": 0.8657004409253339,
+      "learning_rate": 4.5178850466152174e-07,
+      "loss": 0.5633,
+      "num_input_tokens_seen": 139510535,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7884326339205194,
+      "grad_norm": 2.1726706978780914,
+      "learning_rate": 4.5129548918870377e-07,
+      "loss": 0.8123,
+      "num_input_tokens_seen": 139528555,
+      "step": 6557
+    },
+    {
+      "epoch": 0.7885528768111585,
+      "grad_norm": 2.6732285107988605,
+      "learning_rate": 4.5080270865312806e-07,
+      "loss": 0.832,
+      "num_input_tokens_seen": 139545470,
+      "step": 6558
+    },
+    {
+      "epoch": 0.7886731197017977,
+      "grad_norm": 3.4994461583746768,
+      "learning_rate": 4.5031016312955027e-07,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 139563505,
+      "step": 6559
+    },
+    {
+      "epoch": 0.7887933625924367,
+      "grad_norm": 2.164115294854905,
+      "learning_rate": 4.498178526926886e-07,
+      "loss": 0.7444,
+      "num_input_tokens_seen": 139584090,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7889136054830758,
+      "grad_norm": 2.5859696583032643,
+      "learning_rate": 4.4932577741722635e-07,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 139602340,
+      "step": 6561
+    },
+    {
+      "epoch": 0.7890338483737149,
+      "grad_norm": 1.7693188400056055,
+      "learning_rate": 4.4883393737780985e-07,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 139623010,
+      "step": 6562
+    },
+    {
+      "epoch": 0.789154091264354,
+      "grad_norm": 4.134475029366514,
+      "learning_rate": 4.4834233264905254e-07,
+      "loss": 0.7756,
+      "num_input_tokens_seen": 139639745,
+      "step": 6563
+    },
+    {
+      "epoch": 0.789274334154993,
+      "grad_norm": 3.8503971787461686,
+      "learning_rate": 4.478509633055294e-07,
+      "loss": 0.7165,
+      "num_input_tokens_seen": 139657175,
+      "step": 6564
+    },
+    {
+      "epoch": 0.7893945770456322,
+      "grad_norm": 3.925838407472922,
+      "learning_rate": 4.473598294217813e-07,
+      "loss": 0.8013,
+      "num_input_tokens_seen": 139672320,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7895148199362713,
+      "grad_norm": 2.168090397666778,
+      "learning_rate": 4.4686893107231196e-07,
+      "loss": 0.7197,
+      "num_input_tokens_seen": 139689855,
+      "step": 6566
+    },
+    {
+      "epoch": 0.7896350628269103,
+      "grad_norm": 2.532593047945066,
+      "learning_rate": 4.463782683315913e-07,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 139708580,
+      "step": 6567
+    },
+    {
+      "epoch": 0.7897553057175495,
+      "grad_norm": 2.233097236943236,
+      "learning_rate": 4.458878412740523e-07,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 139727080,
+      "step": 6568
+    },
+    {
+      "epoch": 0.7898755486081885,
+      "grad_norm": 3.611853276789595,
+      "learning_rate": 4.453976499740919e-07,
+      "loss": 0.7702,
+      "num_input_tokens_seen": 139744445,
+      "step": 6569
+    },
+    {
+      "epoch": 0.7899957914988276,
+      "grad_norm": 3.8846004019161797,
+      "learning_rate": 4.4490769450607215e-07,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 139761790,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7901160343894668,
+      "grad_norm": 3.077231986599135,
+      "learning_rate": 4.4441797494431845e-07,
+      "loss": 0.7345,
+      "num_input_tokens_seen": 139783315,
+      "step": 6571
+    },
+    {
+      "epoch": 0.7902362772801058,
+      "grad_norm": 4.375585990050049,
+      "learning_rate": 4.439284913631214e-07,
+      "loss": 0.7807,
+      "num_input_tokens_seen": 139800245,
+      "step": 6572
+    },
+    {
+      "epoch": 0.7903565201707449,
+      "grad_norm": 2.7973099914915642,
+      "learning_rate": 4.434392438367347e-07,
+      "loss": 0.8303,
+      "num_input_tokens_seen": 139819390,
+      "step": 6573
+    },
+    {
+      "epoch": 0.790476763061384,
+      "grad_norm": 2.166016007312907,
+      "learning_rate": 4.4295023243937677e-07,
+      "loss": 0.7337,
+      "num_input_tokens_seen": 139839315,
+      "step": 6574
+    },
+    {
+      "epoch": 0.7905970059520231,
+      "grad_norm": 1.9267302561393633,
+      "learning_rate": 4.4246145724523123e-07,
+      "loss": 0.8028,
+      "num_input_tokens_seen": 139856780,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7907172488426621,
+      "grad_norm": 2.833445551831673,
+      "learning_rate": 4.41972918328444e-07,
+      "loss": 0.7606,
+      "num_input_tokens_seen": 139873935,
+      "step": 6576
+    },
+    {
+      "epoch": 0.7908374917333013,
+      "grad_norm": 2.921887831434381,
+      "learning_rate": 4.4148461576312646e-07,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 139893320,
+      "step": 6577
+    },
+    {
+      "epoch": 0.7909577346239404,
+      "grad_norm": 1.5517627113205803,
+      "learning_rate": 4.4099654962335274e-07,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 139913490,
+      "step": 6578
+    },
+    {
+      "epoch": 0.7910779775145794,
+      "grad_norm": 1.8679197766264184,
+      "learning_rate": 4.405087199831636e-07,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 139933450,
+      "step": 6579
+    },
+    {
+      "epoch": 0.7911982204052186,
+      "grad_norm": 10.883326488497756,
+      "learning_rate": 4.400211269165619e-07,
+      "loss": 0.6714,
+      "num_input_tokens_seen": 139949625,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7913184632958576,
+      "grad_norm": 1.6273304937248816,
+      "learning_rate": 4.3953377049751463e-07,
+      "loss": 0.7656,
+      "num_input_tokens_seen": 139969770,
+      "step": 6581
+    },
+    {
+      "epoch": 0.7914387061864967,
+      "grad_norm": 2.8756865821185653,
+      "learning_rate": 4.390466507999533e-07,
+      "loss": 0.7708,
+      "num_input_tokens_seen": 139985240,
+      "step": 6582
+    },
+    {
+      "epoch": 0.7915589490771359,
+      "grad_norm": 2.4951737358532795,
+      "learning_rate": 4.385597678977744e-07,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 140003795,
+      "step": 6583
+    },
+    {
+      "epoch": 0.7916791919677749,
+      "grad_norm": 3.6922783720166055,
+      "learning_rate": 4.3807312186483726e-07,
+      "loss": 0.7416,
+      "num_input_tokens_seen": 140024235,
+      "step": 6584
+    },
+    {
+      "epoch": 0.791799434858414,
+      "grad_norm": 2.1978319902726327,
+      "learning_rate": 4.375867127749655e-07,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 140042230,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7919196777490531,
+      "grad_norm": 2.469862047310355,
+      "learning_rate": 4.3710054070194744e-07,
+      "loss": 0.6755,
+      "num_input_tokens_seen": 140061645,
+      "step": 6586
+    },
+    {
+      "epoch": 0.7920399206396922,
+      "grad_norm": 3.935687673705815,
+      "learning_rate": 4.3661460571953455e-07,
+      "loss": 0.6613,
+      "num_input_tokens_seen": 140078100,
+      "step": 6587
+    },
+    {
+      "epoch": 0.7921601635303313,
+      "grad_norm": 1.7208626897339487,
+      "learning_rate": 4.36128907901443e-07,
+      "loss": 0.6739,
+      "num_input_tokens_seen": 140097415,
+      "step": 6588
+    },
+    {
+      "epoch": 0.7922804064209703,
+      "grad_norm": 3.0157235650548704,
+      "learning_rate": 4.356434473213526e-07,
+      "loss": 0.7123,
+      "num_input_tokens_seen": 140114585,
+      "step": 6589
+    },
+    {
+      "epoch": 0.7924006493116095,
+      "grad_norm": 2.38162014564134,
+      "learning_rate": 4.351582240529068e-07,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 140135135,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7925208922022485,
+      "grad_norm": 0.7069074921835033,
+      "learning_rate": 4.346732381697149e-07,
+      "loss": 0.6098,
+      "num_input_tokens_seen": 140198985,
+      "step": 6591
+    },
+    {
+      "epoch": 0.7926411350928876,
+      "grad_norm": 1.8791145206138866,
+      "learning_rate": 4.3418848974534825e-07,
+      "loss": 0.8117,
+      "num_input_tokens_seen": 140215645,
+      "step": 6592
+    },
+    {
+      "epoch": 0.7927613779835267,
+      "grad_norm": 1.7709507319081672,
+      "learning_rate": 4.3370397885334276e-07,
+      "loss": 0.6802,
+      "num_input_tokens_seen": 140235995,
+      "step": 6593
+    },
+    {
+      "epoch": 0.7928816208741658,
+      "grad_norm": 2.191721012237967,
+      "learning_rate": 4.3321970556719777e-07,
+      "loss": 0.7479,
+      "num_input_tokens_seen": 140254010,
+      "step": 6594
+    },
+    {
+      "epoch": 0.7930018637648049,
+      "grad_norm": 3.232595447172109,
+      "learning_rate": 4.3273566996037814e-07,
+      "loss": 0.7112,
+      "num_input_tokens_seen": 140270425,
+      "step": 6595
+    },
+    {
+      "epoch": 0.793122106655444,
+      "grad_norm": 2.2419566080581093,
+      "learning_rate": 4.322518721063113e-07,
+      "loss": 0.792,
+      "num_input_tokens_seen": 140288695,
+      "step": 6596
+    },
+    {
+      "epoch": 0.7932423495460831,
+      "grad_norm": 2.0667869479864116,
+      "learning_rate": 4.3176831207838906e-07,
+      "loss": 0.6923,
+      "num_input_tokens_seen": 140311825,
+      "step": 6597
+    },
+    {
+      "epoch": 0.7933625924367221,
+      "grad_norm": 1.967228453792871,
+      "learning_rate": 4.3128498994996685e-07,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 140331020,
+      "step": 6598
+    },
+    {
+      "epoch": 0.7934828353273613,
+      "grad_norm": 2.2617734828760048,
+      "learning_rate": 4.308019057943646e-07,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 140352465,
+      "step": 6599
+    },
+    {
+      "epoch": 0.7936030782180004,
+      "grad_norm": 2.1427991752682125,
+      "learning_rate": 4.3031905968486535e-07,
+      "loss": 0.7455,
+      "num_input_tokens_seen": 140373015,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7937233211086394,
+      "grad_norm": 2.2559952354583865,
+      "learning_rate": 4.298364516947168e-07,
+      "loss": 0.6849,
+      "num_input_tokens_seen": 140389965,
+      "step": 6601
+    },
+    {
+      "epoch": 0.7938435639992786,
+      "grad_norm": 2.2881709286565264,
+      "learning_rate": 4.293540818971295e-07,
+      "loss": 0.6617,
+      "num_input_tokens_seen": 140407490,
+      "step": 6602
+    },
+    {
+      "epoch": 0.7939638068899176,
+      "grad_norm": 2.216010863510391,
+      "learning_rate": 4.2887195036527976e-07,
+      "loss": 0.7699,
+      "num_input_tokens_seen": 140426015,
+      "step": 6603
+    },
+    {
+      "epoch": 0.7940840497805567,
+      "grad_norm": 2.924610516895714,
+      "learning_rate": 4.28390057172306e-07,
+      "loss": 0.7301,
+      "num_input_tokens_seen": 140442240,
+      "step": 6604
+    },
+    {
+      "epoch": 0.7942042926711959,
+      "grad_norm": 2.9420251858647166,
+      "learning_rate": 4.279084023913111e-07,
+      "loss": 0.7234,
+      "num_input_tokens_seen": 140459835,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7943245355618349,
+      "grad_norm": 2.110265184849847,
+      "learning_rate": 4.2742698609536096e-07,
+      "loss": 0.687,
+      "num_input_tokens_seen": 140477865,
+      "step": 6606
+    },
+    {
+      "epoch": 0.794444778452474,
+      "grad_norm": 5.3948542484579685,
+      "learning_rate": 4.2694580835748706e-07,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 140497445,
+      "step": 6607
+    },
+    {
+      "epoch": 0.7945650213431131,
+      "grad_norm": 2.0683105884559136,
+      "learning_rate": 4.264648692506836e-07,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 140515955,
+      "step": 6608
+    },
+    {
+      "epoch": 0.7946852642337522,
+      "grad_norm": 1.9142453139761246,
+      "learning_rate": 4.2598416884790824e-07,
+      "loss": 0.715,
+      "num_input_tokens_seen": 140534725,
+      "step": 6609
+    },
+    {
+      "epoch": 0.7948055071243912,
+      "grad_norm": 2.7881176658288855,
+      "learning_rate": 4.255037072220824e-07,
+      "loss": 0.8062,
+      "num_input_tokens_seen": 140555815,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7949257500150304,
+      "grad_norm": 2.101215362922217,
+      "learning_rate": 4.2502348444609293e-07,
+      "loss": 0.711,
+      "num_input_tokens_seen": 140575155,
+      "step": 6611
+    },
+    {
+      "epoch": 0.7950459929056695,
+      "grad_norm": 2.4649507122318344,
+      "learning_rate": 4.2454350059278844e-07,
+      "loss": 0.6813,
+      "num_input_tokens_seen": 140595935,
+      "step": 6612
+    },
+    {
+      "epoch": 0.7951662357963085,
+      "grad_norm": 1.8590295887028097,
+      "learning_rate": 4.240637557349824e-07,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 140612870,
+      "step": 6613
+    },
+    {
+      "epoch": 0.7952864786869477,
+      "grad_norm": 2.150851352209119,
+      "learning_rate": 4.235842499454516e-07,
+      "loss": 0.6536,
+      "num_input_tokens_seen": 140632505,
+      "step": 6614
+    },
+    {
+      "epoch": 0.7954067215775867,
+      "grad_norm": 1.9800560810694914,
+      "learning_rate": 4.2310498329693687e-07,
+      "loss": 0.8183,
+      "num_input_tokens_seen": 140653125,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7955269644682258,
+      "grad_norm": 1.775172580027822,
+      "learning_rate": 4.226259558621421e-07,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 140673940,
+      "step": 6616
+    },
+    {
+      "epoch": 0.795647207358865,
+      "grad_norm": 2.1544108526428,
+      "learning_rate": 4.221471677137358e-07,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 140694475,
+      "step": 6617
+    },
+    {
+      "epoch": 0.795767450249504,
+      "grad_norm": 1.6828833693622285,
+      "learning_rate": 4.216686189243492e-07,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 140712985,
+      "step": 6618
+    },
+    {
+      "epoch": 0.7958876931401431,
+      "grad_norm": 1.7448310175183948,
+      "learning_rate": 4.211903095665785e-07,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 140732090,
+      "step": 6619
+    },
+    {
+      "epoch": 0.7960079360307821,
+      "grad_norm": 2.202566601572069,
+      "learning_rate": 4.2071223971298277e-07,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 140748995,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7961281789214213,
+      "grad_norm": 2.629815763780873,
+      "learning_rate": 4.2023440943608433e-07,
+      "loss": 0.6088,
+      "num_input_tokens_seen": 140768680,
+      "step": 6621
+    },
+    {
+      "epoch": 0.7962484218120603,
+      "grad_norm": 2.204287002448897,
+      "learning_rate": 4.1975681880836954e-07,
+      "loss": 0.7839,
+      "num_input_tokens_seen": 140788405,
+      "step": 6622
+    },
+    {
+      "epoch": 0.7963686647026994,
+      "grad_norm": 1.7594970002206367,
+      "learning_rate": 4.192794679022895e-07,
+      "loss": 0.8225,
+      "num_input_tokens_seen": 140806450,
+      "step": 6623
+    },
+    {
+      "epoch": 0.7964889075933386,
+      "grad_norm": 2.0018840328485648,
+      "learning_rate": 4.1880235679025743e-07,
+      "loss": 0.7132,
+      "num_input_tokens_seen": 140826265,
+      "step": 6624
+    },
+    {
+      "epoch": 0.7966091504839776,
+      "grad_norm": 2.8301330020851494,
+      "learning_rate": 4.1832548554465054e-07,
+      "loss": 0.6271,
+      "num_input_tokens_seen": 140844280,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7967293933746167,
+      "grad_norm": 0.7893052969496102,
+      "learning_rate": 4.1784885423780934e-07,
+      "loss": 0.6107,
+      "num_input_tokens_seen": 140901580,
+      "step": 6626
+    },
+    {
+      "epoch": 0.7968496362652558,
+      "grad_norm": 2.066843155078559,
+      "learning_rate": 4.173724629420394e-07,
+      "loss": 0.8829,
+      "num_input_tokens_seen": 140922660,
+      "step": 6627
+    },
+    {
+      "epoch": 0.7969698791558949,
+      "grad_norm": 4.259539999530233,
+      "learning_rate": 4.168963117296087e-07,
+      "loss": 0.6791,
+      "num_input_tokens_seen": 140939715,
+      "step": 6628
+    },
+    {
+      "epoch": 0.797090122046534,
+      "grad_norm": 2.479631822273694,
+      "learning_rate": 4.1642040067274876e-07,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 140959105,
+      "step": 6629
+    },
+    {
+      "epoch": 0.7972103649371731,
+      "grad_norm": 1.8333417997372683,
+      "learning_rate": 4.1594472984365493e-07,
+      "loss": 0.7173,
+      "num_input_tokens_seen": 140977510,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7973306078278122,
+      "grad_norm": 4.717198677937778,
+      "learning_rate": 4.154692993144862e-07,
+      "loss": 0.7657,
+      "num_input_tokens_seen": 140997000,
+      "step": 6631
+    },
+    {
+      "epoch": 0.7974508507184512,
+      "grad_norm": 2.2805903769213933,
+      "learning_rate": 4.1499410915736476e-07,
+      "loss": 0.7103,
+      "num_input_tokens_seen": 141015650,
+      "step": 6632
+    },
+    {
+      "epoch": 0.7975710936090904,
+      "grad_norm": 0.823809378840752,
+      "learning_rate": 4.145191594443769e-07,
+      "loss": 0.7079,
+      "num_input_tokens_seen": 141079725,
+      "step": 6633
+    },
+    {
+      "epoch": 0.7976913364997295,
+      "grad_norm": 1.9422318021280922,
+      "learning_rate": 4.140444502475713e-07,
+      "loss": 0.7054,
+      "num_input_tokens_seen": 141098995,
+      "step": 6634
+    },
+    {
+      "epoch": 0.7978115793903685,
+      "grad_norm": 2.4784895766777892,
+      "learning_rate": 4.1356998163896216e-07,
+      "loss": 0.6927,
+      "num_input_tokens_seen": 141115765,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7979318222810077,
+      "grad_norm": 2.5803522743587157,
+      "learning_rate": 4.130957536905255e-07,
+      "loss": 0.739,
+      "num_input_tokens_seen": 141133500,
+      "step": 6636
+    },
+    {
+      "epoch": 0.7980520651716467,
+      "grad_norm": 3.7813008882868284,
+      "learning_rate": 4.1262176647420134e-07,
+      "loss": 0.7084,
+      "num_input_tokens_seen": 141151385,
+      "step": 6637
+    },
+    {
+      "epoch": 0.7981723080622858,
+      "grad_norm": 18.140299554166404,
+      "learning_rate": 4.121480200618923e-07,
+      "loss": 0.7953,
+      "num_input_tokens_seen": 141170760,
+      "step": 6638
+    },
+    {
+      "epoch": 0.798292550952925,
+      "grad_norm": 2.4006875597946125,
+      "learning_rate": 4.11674514525467e-07,
+      "loss": 0.793,
+      "num_input_tokens_seen": 141190015,
+      "step": 6639
+    },
+    {
+      "epoch": 0.798412793843564,
+      "grad_norm": 0.8185702746196547,
+      "learning_rate": 4.1120124993675476e-07,
+      "loss": 0.6126,
+      "num_input_tokens_seen": 141254165,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7985330367342031,
+      "grad_norm": 2.166649431357531,
+      "learning_rate": 4.107282263675498e-07,
+      "loss": 0.6214,
+      "num_input_tokens_seen": 141271555,
+      "step": 6641
+    },
+    {
+      "epoch": 0.7986532796248422,
+      "grad_norm": 0.7493303631700791,
+      "learning_rate": 4.1025544388960907e-07,
+      "loss": 0.5432,
+      "num_input_tokens_seen": 141332315,
+      "step": 6642
+    },
+    {
+      "epoch": 0.7987735225154813,
+      "grad_norm": 2.154479093536911,
+      "learning_rate": 4.097829025746538e-07,
+      "loss": 0.7131,
+      "num_input_tokens_seen": 141353580,
+      "step": 6643
+    },
+    {
+      "epoch": 0.7988937654061203,
+      "grad_norm": 0.7051672616595179,
+      "learning_rate": 4.0931060249436757e-07,
+      "loss": 0.6227,
+      "num_input_tokens_seen": 141417140,
+      "step": 6644
+    },
+    {
+      "epoch": 0.7990140082967595,
+      "grad_norm": 3.846159148833426,
+      "learning_rate": 4.088385437203983e-07,
+      "loss": 0.6845,
+      "num_input_tokens_seen": 141433870,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7991342511873986,
+      "grad_norm": 2.736133380293792,
+      "learning_rate": 4.083667263243564e-07,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 141451935,
+      "step": 6646
+    },
+    {
+      "epoch": 0.7992544940780376,
+      "grad_norm": 2.0825976885317905,
+      "learning_rate": 4.0789515037781696e-07,
+      "loss": 0.7053,
+      "num_input_tokens_seen": 141472380,
+      "step": 6647
+    },
+    {
+      "epoch": 0.7993747369686768,
+      "grad_norm": 6.517627822534658,
+      "learning_rate": 4.0742381595231755e-07,
+      "loss": 0.8236,
+      "num_input_tokens_seen": 141488825,
+      "step": 6648
+    },
+    {
+      "epoch": 0.7994949798593158,
+      "grad_norm": 1.7369255979642453,
+      "learning_rate": 4.06952723119359e-07,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 141508420,
+      "step": 6649
+    },
+    {
+      "epoch": 0.7996152227499549,
+      "grad_norm": 3.697304491371467,
+      "learning_rate": 4.0648187195040504e-07,
+      "loss": 0.6653,
+      "num_input_tokens_seen": 141530345,
+      "step": 6650
+    },
+    {
+      "epoch": 0.799735465640594,
+      "grad_norm": 1.042984619199664,
+      "learning_rate": 4.060112625168848e-07,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 141595175,
+      "step": 6651
+    },
+    {
+      "epoch": 0.7998557085312331,
+      "grad_norm": 2.2622730295235596,
+      "learning_rate": 4.055408948901886e-07,
+      "loss": 0.7381,
+      "num_input_tokens_seen": 141616295,
+      "step": 6652
+    },
+    {
+      "epoch": 0.7999759514218722,
+      "grad_norm": 2.8461654667839715,
+      "learning_rate": 4.050707691416708e-07,
+      "loss": 0.7078,
+      "num_input_tokens_seen": 141637325,
+      "step": 6653
+    },
+    {
+      "epoch": 0.8000961943125112,
+      "grad_norm": 0.7349926385703843,
+      "learning_rate": 4.046008853426488e-07,
+      "loss": 0.6211,
+      "num_input_tokens_seen": 141700360,
+      "step": 6654
+    },
+    {
+      "epoch": 0.8002164372031504,
+      "grad_norm": 2.4138838760615267,
+      "learning_rate": 4.0413124356440464e-07,
+      "loss": 0.6276,
+      "num_input_tokens_seen": 141724125,
+      "step": 6655
+    },
+    {
+      "epoch": 0.8003366800937894,
+      "grad_norm": 2.0200897689092625,
+      "learning_rate": 4.036618438781818e-07,
+      "loss": 0.8227,
+      "num_input_tokens_seen": 141742305,
+      "step": 6656
+    },
+    {
+      "epoch": 0.8004569229844285,
+      "grad_norm": 2.061170798551594,
+      "learning_rate": 4.0319268635518797e-07,
+      "loss": 0.849,
+      "num_input_tokens_seen": 141762600,
+      "step": 6657
+    },
+    {
+      "epoch": 0.8005771658750677,
+      "grad_norm": 1.9053033050912034,
+      "learning_rate": 4.027237710665943e-07,
+      "loss": 0.7444,
+      "num_input_tokens_seen": 141780785,
+      "step": 6658
+    },
+    {
+      "epoch": 0.8006974087657067,
+      "grad_norm": 2.1208929488561132,
+      "learning_rate": 4.022550980835344e-07,
+      "loss": 0.6943,
+      "num_input_tokens_seen": 141802750,
+      "step": 6659
+    },
+    {
+      "epoch": 0.8008176516563458,
+      "grad_norm": 4.977877012078649,
+      "learning_rate": 4.017866674771058e-07,
+      "loss": 0.8022,
+      "num_input_tokens_seen": 141819955,
+      "step": 6660
+    },
+    {
+      "epoch": 0.8009378945469849,
+      "grad_norm": 2.014074460767353,
+      "learning_rate": 4.013184793183688e-07,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 141841770,
+      "step": 6661
+    },
+    {
+      "epoch": 0.801058137437624,
+      "grad_norm": 2.1151315127676122,
+      "learning_rate": 4.008505336783472e-07,
+      "loss": 0.7188,
+      "num_input_tokens_seen": 141859215,
+      "step": 6662
+    },
+    {
+      "epoch": 0.801178380328263,
+      "grad_norm": 2.0670748493270903,
+      "learning_rate": 4.003828306280284e-07,
+      "loss": 0.817,
+      "num_input_tokens_seen": 141876610,
+      "step": 6663
+    },
+    {
+      "epoch": 0.8012986232189022,
+      "grad_norm": 1.845973965206068,
+      "learning_rate": 3.999153702383626e-07,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 141894220,
+      "step": 6664
+    },
+    {
+      "epoch": 0.8014188661095413,
+      "grad_norm": 2.174803822126223,
+      "learning_rate": 3.9944815258026263e-07,
+      "loss": 0.7274,
+      "num_input_tokens_seen": 141915760,
+      "step": 6665
+    },
+    {
+      "epoch": 0.8015391090001803,
+      "grad_norm": 13.06832285281222,
+      "learning_rate": 3.9898117772460505e-07,
+      "loss": 0.8269,
+      "num_input_tokens_seen": 141935650,
+      "step": 6666
+    },
+    {
+      "epoch": 0.8016593518908195,
+      "grad_norm": 0.921291621580826,
+      "learning_rate": 3.985144457422305e-07,
+      "loss": 0.7001,
+      "num_input_tokens_seen": 141989655,
+      "step": 6667
+    },
+    {
+      "epoch": 0.8017795947814585,
+      "grad_norm": 4.0151380443713105,
+      "learning_rate": 3.9804795670394096e-07,
+      "loss": 0.754,
+      "num_input_tokens_seen": 142009500,
+      "step": 6668
+    },
+    {
+      "epoch": 0.8018998376720976,
+      "grad_norm": 2.5819981014056443,
+      "learning_rate": 3.975817106805026e-07,
+      "loss": 0.7068,
+      "num_input_tokens_seen": 142027920,
+      "step": 6669
+    },
+    {
+      "epoch": 0.8020200805627368,
+      "grad_norm": 2.212655538936499,
+      "learning_rate": 3.9711570774264433e-07,
+      "loss": 0.6394,
+      "num_input_tokens_seen": 142048315,
+      "step": 6670
+    },
+    {
+      "epoch": 0.8021403234533758,
+      "grad_norm": 2.371809147509461,
+      "learning_rate": 3.966499479610592e-07,
+      "loss": 0.6497,
+      "num_input_tokens_seen": 142066130,
+      "step": 6671
+    },
+    {
+      "epoch": 0.8022605663440149,
+      "grad_norm": 1.9087701317367574,
+      "learning_rate": 3.9618443140640225e-07,
+      "loss": 0.6502,
+      "num_input_tokens_seen": 142084760,
+      "step": 6672
+    },
+    {
+      "epoch": 0.802380809234654,
+      "grad_norm": 0.759228835808509,
+      "learning_rate": 3.957191581492918e-07,
+      "loss": 0.5507,
+      "num_input_tokens_seen": 142145240,
+      "step": 6673
+    },
+    {
+      "epoch": 0.8025010521252931,
+      "grad_norm": 5.620310477800154,
+      "learning_rate": 3.952541282603097e-07,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 142160065,
+      "step": 6674
+    },
+    {
+      "epoch": 0.8026212950159322,
+      "grad_norm": 1.865219317954989,
+      "learning_rate": 3.9478934181000013e-07,
+      "loss": 0.8288,
+      "num_input_tokens_seen": 142179810,
+      "step": 6675
+    },
+    {
+      "epoch": 0.8027415379065713,
+      "grad_norm": 6.216344304846314,
+      "learning_rate": 3.943247988688714e-07,
+      "loss": 0.8395,
+      "num_input_tokens_seen": 142198225,
+      "step": 6676
+    },
+    {
+      "epoch": 0.8028617807972104,
+      "grad_norm": 2.3350731271176586,
+      "learning_rate": 3.9386049950739377e-07,
+      "loss": 0.7117,
+      "num_input_tokens_seen": 142216415,
+      "step": 6677
+    },
+    {
+      "epoch": 0.8029820236878494,
+      "grad_norm": 3.6487187548592233,
+      "learning_rate": 3.933964437960009e-07,
+      "loss": 0.6579,
+      "num_input_tokens_seen": 142235965,
+      "step": 6678
+    },
+    {
+      "epoch": 0.8031022665784886,
+      "grad_norm": 2.7355194831009553,
+      "learning_rate": 3.929326318050907e-07,
+      "loss": 0.7095,
+      "num_input_tokens_seen": 142253355,
+      "step": 6679
+    },
+    {
+      "epoch": 0.8032225094691277,
+      "grad_norm": 2.020066783805988,
+      "learning_rate": 3.924690636050225e-07,
+      "loss": 0.7792,
+      "num_input_tokens_seen": 142270485,
+      "step": 6680
+    },
+    {
+      "epoch": 0.8033427523597667,
+      "grad_norm": 3.2068520478362346,
+      "learning_rate": 3.9200573926611915e-07,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 142291620,
+      "step": 6681
+    },
+    {
+      "epoch": 0.8034629952504058,
+      "grad_norm": 2.10061308169592,
+      "learning_rate": 3.9154265885866613e-07,
+      "loss": 0.7193,
+      "num_input_tokens_seen": 142310650,
+      "step": 6682
+    },
+    {
+      "epoch": 0.8035832381410449,
+      "grad_norm": 3.317079256855228,
+      "learning_rate": 3.910798224529135e-07,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 142328495,
+      "step": 6683
+    },
+    {
+      "epoch": 0.803703481031684,
+      "grad_norm": 5.584756791538455,
+      "learning_rate": 3.9061723011907245e-07,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 142347570,
+      "step": 6684
+    },
+    {
+      "epoch": 0.803823723922323,
+      "grad_norm": 2.4808570259105167,
+      "learning_rate": 3.901548819273179e-07,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 142367305,
+      "step": 6685
+    },
+    {
+      "epoch": 0.8039439668129622,
+      "grad_norm": 2.038494087397355,
+      "learning_rate": 3.896927779477881e-07,
+      "loss": 0.6844,
+      "num_input_tokens_seen": 142386285,
+      "step": 6686
+    },
+    {
+      "epoch": 0.8040642097036013,
+      "grad_norm": 2.8535043409312704,
+      "learning_rate": 3.892309182505833e-07,
+      "loss": 0.6729,
+      "num_input_tokens_seen": 142403820,
+      "step": 6687
+    },
+    {
+      "epoch": 0.8041844525942403,
+      "grad_norm": 4.054940566723818,
+      "learning_rate": 3.887693029057675e-07,
+      "loss": 0.8568,
+      "num_input_tokens_seen": 142423050,
+      "step": 6688
+    },
+    {
+      "epoch": 0.8043046954848795,
+      "grad_norm": 2.1832544728071235,
+      "learning_rate": 3.8830793198336753e-07,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 142442360,
+      "step": 6689
+    },
+    {
+      "epoch": 0.8044249383755185,
+      "grad_norm": 4.666573500188218,
+      "learning_rate": 3.878468055533721e-07,
+      "loss": 0.7005,
+      "num_input_tokens_seen": 142464620,
+      "step": 6690
+    },
+    {
+      "epoch": 0.8045451812661576,
+      "grad_norm": 3.0258409366219223,
+      "learning_rate": 3.8738592368573507e-07,
+      "loss": 0.8473,
+      "num_input_tokens_seen": 142481895,
+      "step": 6691
+    },
+    {
+      "epoch": 0.8046654241567968,
+      "grad_norm": 2.1204446880215864,
+      "learning_rate": 3.8692528645037137e-07,
+      "loss": 0.8707,
+      "num_input_tokens_seen": 142500795,
+      "step": 6692
+    },
+    {
+      "epoch": 0.8047856670474358,
+      "grad_norm": 2.5912381251580676,
+      "learning_rate": 3.8646489391715907e-07,
+      "loss": 0.7707,
+      "num_input_tokens_seen": 142514810,
+      "step": 6693
+    },
+    {
+      "epoch": 0.8049059099380749,
+      "grad_norm": 2.5851729898860936,
+      "learning_rate": 3.8600474615593903e-07,
+      "loss": 0.8777,
+      "num_input_tokens_seen": 142529145,
+      "step": 6694
+    },
+    {
+      "epoch": 0.805026152828714,
+      "grad_norm": 0.9188288201349036,
+      "learning_rate": 3.8554484323651605e-07,
+      "loss": 0.6557,
+      "num_input_tokens_seen": 142590735,
+      "step": 6695
+    },
+    {
+      "epoch": 0.8051463957193531,
+      "grad_norm": 1.9903972938415635,
+      "learning_rate": 3.85085185228657e-07,
+      "loss": 0.7848,
+      "num_input_tokens_seen": 142609425,
+      "step": 6696
+    },
+    {
+      "epoch": 0.8052666386099921,
+      "grad_norm": 3.2140251742275936,
+      "learning_rate": 3.8462577220209114e-07,
+      "loss": 0.7364,
+      "num_input_tokens_seen": 142629520,
+      "step": 6697
+    },
+    {
+      "epoch": 0.8053868815006313,
+      "grad_norm": 0.7113146974068315,
+      "learning_rate": 3.841666042265106e-07,
+      "loss": 0.6071,
+      "num_input_tokens_seen": 142698890,
+      "step": 6698
+    },
+    {
+      "epoch": 0.8055071243912704,
+      "grad_norm": 2.6706431701741256,
+      "learning_rate": 3.837076813715723e-07,
+      "loss": 0.6825,
+      "num_input_tokens_seen": 142718495,
+      "step": 6699
+    },
+    {
+      "epoch": 0.8056273672819094,
+      "grad_norm": 2.148499868023375,
+      "learning_rate": 3.832490037068934e-07,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 142737005,
+      "step": 6700
+    },
+    {
+      "epoch": 0.8057476101725486,
+      "grad_norm": 2.1737305236238154,
+      "learning_rate": 3.827905713020554e-07,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 142754370,
+      "step": 6701
+    },
+    {
+      "epoch": 0.8058678530631876,
+      "grad_norm": 2.365473207320476,
+      "learning_rate": 3.823323842266017e-07,
+      "loss": 0.6735,
+      "num_input_tokens_seen": 142773485,
+      "step": 6702
+    },
+    {
+      "epoch": 0.8059880959538267,
+      "grad_norm": 2.6862700972810596,
+      "learning_rate": 3.818744425500393e-07,
+      "loss": 0.733,
+      "num_input_tokens_seen": 142791220,
+      "step": 6703
+    },
+    {
+      "epoch": 0.8061083388444659,
+      "grad_norm": 2.241068931896184,
+      "learning_rate": 3.814167463418372e-07,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 142809970,
+      "step": 6704
+    },
+    {
+      "epoch": 0.8062285817351049,
+      "grad_norm": 9.429142067582834,
+      "learning_rate": 3.809592956714278e-07,
+      "loss": 0.6463,
+      "num_input_tokens_seen": 142832925,
+      "step": 6705
+    },
+    {
+      "epoch": 0.806348824625744,
+      "grad_norm": 2.0593662389606386,
+      "learning_rate": 3.805020906082057e-07,
+      "loss": 0.736,
+      "num_input_tokens_seen": 142851220,
+      "step": 6706
+    },
+    {
+      "epoch": 0.8064690675163831,
+      "grad_norm": 2.638522533626239,
+      "learning_rate": 3.8004513122152917e-07,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 142869250,
+      "step": 6707
+    },
+    {
+      "epoch": 0.8065893104070222,
+      "grad_norm": 1.902348412841278,
+      "learning_rate": 3.79588417580718e-07,
+      "loss": 0.6648,
+      "num_input_tokens_seen": 142887080,
+      "step": 6708
+    },
+    {
+      "epoch": 0.8067095532976613,
+      "grad_norm": 2.0590983231847724,
+      "learning_rate": 3.791319497550558e-07,
+      "loss": 0.7595,
+      "num_input_tokens_seen": 142904630,
+      "step": 6709
+    },
+    {
+      "epoch": 0.8068297961883004,
+      "grad_norm": 2.219228161743034,
+      "learning_rate": 3.7867572781378755e-07,
+      "loss": 0.7104,
+      "num_input_tokens_seen": 142921915,
+      "step": 6710
+    },
+    {
+      "epoch": 0.8069500390789395,
+      "grad_norm": 1.6897252461476986,
+      "learning_rate": 3.782197518261225e-07,
+      "loss": 0.7326,
+      "num_input_tokens_seen": 142941075,
+      "step": 6711
+    },
+    {
+      "epoch": 0.8070702819695785,
+      "grad_norm": 2.480708620891473,
+      "learning_rate": 3.777640218612319e-07,
+      "loss": 0.9579,
+      "num_input_tokens_seen": 142958780,
+      "step": 6712
+    },
+    {
+      "epoch": 0.8071905248602176,
+      "grad_norm": 2.7879108565817585,
+      "learning_rate": 3.7730853798824945e-07,
+      "loss": 0.7068,
+      "num_input_tokens_seen": 142977555,
+      "step": 6713
+    },
+    {
+      "epoch": 0.8073107677508568,
+      "grad_norm": 2.42795345404584,
+      "learning_rate": 3.768533002762708e-07,
+      "loss": 0.7598,
+      "num_input_tokens_seen": 143000810,
+      "step": 6714
+    },
+    {
+      "epoch": 0.8074310106414958,
+      "grad_norm": 2.0449347474493926,
+      "learning_rate": 3.763983087943567e-07,
+      "loss": 0.7563,
+      "num_input_tokens_seen": 143019920,
+      "step": 6715
+    },
+    {
+      "epoch": 0.8075512535321349,
+      "grad_norm": 1.7563104276740333,
+      "learning_rate": 3.759435636115282e-07,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 143040425,
+      "step": 6716
+    },
+    {
+      "epoch": 0.807671496422774,
+      "grad_norm": 1.832554866791834,
+      "learning_rate": 3.7548906479676967e-07,
+      "loss": 0.738,
+      "num_input_tokens_seen": 143059740,
+      "step": 6717
+    },
+    {
+      "epoch": 0.8077917393134131,
+      "grad_norm": 2.50729886812423,
+      "learning_rate": 3.7503481241902855e-07,
+      "loss": 0.7072,
+      "num_input_tokens_seen": 143079435,
+      "step": 6718
+    },
+    {
+      "epoch": 0.8079119822040521,
+      "grad_norm": 4.582345259405457,
+      "learning_rate": 3.745808065472145e-07,
+      "loss": 0.7975,
+      "num_input_tokens_seen": 143096450,
+      "step": 6719
+    },
+    {
+      "epoch": 0.8080322250946913,
+      "grad_norm": 1.7226459971937955,
+      "learning_rate": 3.741270472501994e-07,
+      "loss": 0.7556,
+      "num_input_tokens_seen": 143116810,
+      "step": 6720
+    },
+    {
+      "epoch": 0.8081524679853304,
+      "grad_norm": 1.975239213208688,
+      "learning_rate": 3.736735345968187e-07,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 143136140,
+      "step": 6721
+    },
+    {
+      "epoch": 0.8082727108759694,
+      "grad_norm": 8.697357206168736,
+      "learning_rate": 3.732202686558692e-07,
+      "loss": 0.7869,
+      "num_input_tokens_seen": 143154895,
+      "step": 6722
+    },
+    {
+      "epoch": 0.8083929537666086,
+      "grad_norm": 2.0376427754196493,
+      "learning_rate": 3.7276724949611206e-07,
+      "loss": 0.7223,
+      "num_input_tokens_seen": 143174725,
+      "step": 6723
+    },
+    {
+      "epoch": 0.8085131966572476,
+      "grad_norm": 2.352813015761238,
+      "learning_rate": 3.723144771862694e-07,
+      "loss": 0.7377,
+      "num_input_tokens_seen": 143195085,
+      "step": 6724
+    },
+    {
+      "epoch": 0.8086334395478867,
+      "grad_norm": 1.6784396475362502,
+      "learning_rate": 3.718619517950263e-07,
+      "loss": 0.7625,
+      "num_input_tokens_seen": 143215400,
+      "step": 6725
+    },
+    {
+      "epoch": 0.8087536824385259,
+      "grad_norm": 2.2099908650210236,
+      "learning_rate": 3.714096733910301e-07,
+      "loss": 0.765,
+      "num_input_tokens_seen": 143232645,
+      "step": 6726
+    },
+    {
+      "epoch": 0.8088739253291649,
+      "grad_norm": 3.883394270349313,
+      "learning_rate": 3.7095764204289216e-07,
+      "loss": 0.6982,
+      "num_input_tokens_seen": 143253165,
+      "step": 6727
+    },
+    {
+      "epoch": 0.808994168219804,
+      "grad_norm": 2.56278938015523,
+      "learning_rate": 3.7050585781918463e-07,
+      "loss": 0.7315,
+      "num_input_tokens_seen": 143273185,
+      "step": 6728
+    },
+    {
+      "epoch": 0.8091144111104431,
+      "grad_norm": 2.2712109289236997,
+      "learning_rate": 3.700543207884428e-07,
+      "loss": 0.6861,
+      "num_input_tokens_seen": 143289815,
+      "step": 6729
+    },
+    {
+      "epoch": 0.8092346540010822,
+      "grad_norm": 2.66601702341341,
+      "learning_rate": 3.6960303101916466e-07,
+      "loss": 0.7029,
+      "num_input_tokens_seen": 143309450,
+      "step": 6730
+    },
+    {
+      "epoch": 0.8093548968917212,
+      "grad_norm": 0.8214335467969754,
+      "learning_rate": 3.6915198857981047e-07,
+      "loss": 0.5876,
+      "num_input_tokens_seen": 143374370,
+      "step": 6731
+    },
+    {
+      "epoch": 0.8094751397823604,
+      "grad_norm": 1.9391259163086911,
+      "learning_rate": 3.687011935388027e-07,
+      "loss": 0.6796,
+      "num_input_tokens_seen": 143396985,
+      "step": 6732
+    },
+    {
+      "epoch": 0.8095953826729995,
+      "grad_norm": 2.4245265938286367,
+      "learning_rate": 3.6825064596452715e-07,
+      "loss": 0.7216,
+      "num_input_tokens_seen": 143417050,
+      "step": 6733
+    },
+    {
+      "epoch": 0.8097156255636385,
+      "grad_norm": 1.8143374295696109,
+      "learning_rate": 3.678003459253305e-07,
+      "loss": 0.7034,
+      "num_input_tokens_seen": 143437620,
+      "step": 6734
+    },
+    {
+      "epoch": 0.8098358684542777,
+      "grad_norm": 2.1423278422042458,
+      "learning_rate": 3.673502934895241e-07,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 143456845,
+      "step": 6735
+    },
+    {
+      "epoch": 0.8099561113449167,
+      "grad_norm": 0.7168994638594164,
+      "learning_rate": 3.669004887253802e-07,
+      "loss": 0.5935,
+      "num_input_tokens_seen": 143522855,
+      "step": 6736
+    },
+    {
+      "epoch": 0.8100763542355558,
+      "grad_norm": 1.7156257068197367,
+      "learning_rate": 3.664509317011335e-07,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 143542910,
+      "step": 6737
+    },
+    {
+      "epoch": 0.810196597126195,
+      "grad_norm": 1.895946684944535,
+      "learning_rate": 3.6600162248498134e-07,
+      "loss": 0.7267,
+      "num_input_tokens_seen": 143566260,
+      "step": 6738
+    },
+    {
+      "epoch": 0.810316840016834,
+      "grad_norm": 1.9241933921021666,
+      "learning_rate": 3.6555256114508426e-07,
+      "loss": 0.7593,
+      "num_input_tokens_seen": 143585775,
+      "step": 6739
+    },
+    {
+      "epoch": 0.8104370829074731,
+      "grad_norm": 1.9645312809474171,
+      "learning_rate": 3.651037477495642e-07,
+      "loss": 0.7224,
+      "num_input_tokens_seen": 143606945,
+      "step": 6740
+    },
+    {
+      "epoch": 0.8105573257981122,
+      "grad_norm": 2.7667999849248037,
+      "learning_rate": 3.6465518236650584e-07,
+      "loss": 0.6747,
+      "num_input_tokens_seen": 143626810,
+      "step": 6741
+    },
+    {
+      "epoch": 0.8106775686887513,
+      "grad_norm": 1.7810624871962215,
+      "learning_rate": 3.642068650639554e-07,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 143646275,
+      "step": 6742
+    },
+    {
+      "epoch": 0.8107978115793903,
+      "grad_norm": 2.61235974471993,
+      "learning_rate": 3.6375879590992334e-07,
+      "loss": 0.6424,
+      "num_input_tokens_seen": 143666340,
+      "step": 6743
+    },
+    {
+      "epoch": 0.8109180544700295,
+      "grad_norm": 2.2109984483867255,
+      "learning_rate": 3.6331097497238104e-07,
+      "loss": 0.7997,
+      "num_input_tokens_seen": 143685505,
+      "step": 6744
+    },
+    {
+      "epoch": 0.8110382973606686,
+      "grad_norm": 3.9619369668625457,
+      "learning_rate": 3.628634023192627e-07,
+      "loss": 0.7908,
+      "num_input_tokens_seen": 143705470,
+      "step": 6745
+    },
+    {
+      "epoch": 0.8111585402513076,
+      "grad_norm": 2.6808505791682546,
+      "learning_rate": 3.624160780184644e-07,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 143722405,
+      "step": 6746
+    },
+    {
+      "epoch": 0.8112787831419467,
+      "grad_norm": 2.3236594670981168,
+      "learning_rate": 3.6196900213784496e-07,
+      "loss": 0.7431,
+      "num_input_tokens_seen": 143741440,
+      "step": 6747
+    },
+    {
+      "epoch": 0.8113990260325858,
+      "grad_norm": 2.187873408270472,
+      "learning_rate": 3.6152217474522575e-07,
+      "loss": 0.858,
+      "num_input_tokens_seen": 143757975,
+      "step": 6748
+    },
+    {
+      "epoch": 0.8115192689232249,
+      "grad_norm": 1.7824908365274366,
+      "learning_rate": 3.6107559590838975e-07,
+      "loss": 0.7219,
+      "num_input_tokens_seen": 143776680,
+      "step": 6749
+    },
+    {
+      "epoch": 0.811639511813864,
+      "grad_norm": 4.215220747446123,
+      "learning_rate": 3.606292656950822e-07,
+      "loss": 0.6579,
+      "num_input_tokens_seen": 143794810,
+      "step": 6750
+    },
+    {
+      "epoch": 0.8117597547045031,
+      "grad_norm": 2.2951263267094584,
+      "learning_rate": 3.601831841730121e-07,
+      "loss": 0.8562,
+      "num_input_tokens_seen": 143812450,
+      "step": 6751
+    },
+    {
+      "epoch": 0.8118799975951422,
+      "grad_norm": 2.0775748802028513,
+      "learning_rate": 3.5973735140984916e-07,
+      "loss": 0.7325,
+      "num_input_tokens_seen": 143832340,
+      "step": 6752
+    },
+    {
+      "epoch": 0.8120002404857812,
+      "grad_norm": 2.4919784637661975,
+      "learning_rate": 3.5929176747322607e-07,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 143851165,
+      "step": 6753
+    },
+    {
+      "epoch": 0.8121204833764204,
+      "grad_norm": 0.839488909640401,
+      "learning_rate": 3.588464324307365e-07,
+      "loss": 0.5734,
+      "num_input_tokens_seen": 143914510,
+      "step": 6754
+    },
+    {
+      "epoch": 0.8122407262670595,
+      "grad_norm": 2.225475316344406,
+      "learning_rate": 3.584013463499391e-07,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 143932850,
+      "step": 6755
+    },
+    {
+      "epoch": 0.8123609691576985,
+      "grad_norm": 0.7295797115039838,
+      "learning_rate": 3.579565092983521e-07,
+      "loss": 0.6505,
+      "num_input_tokens_seen": 143993690,
+      "step": 6756
+    },
+    {
+      "epoch": 0.8124812120483377,
+      "grad_norm": 3.08029310634769,
+      "learning_rate": 3.57511921343457e-07,
+      "loss": 0.8313,
+      "num_input_tokens_seen": 144011925,
+      "step": 6757
+    },
+    {
+      "epoch": 0.8126014549389767,
+      "grad_norm": 2.0872777291460576,
+      "learning_rate": 3.5706758255269696e-07,
+      "loss": 0.8095,
+      "num_input_tokens_seen": 144030100,
+      "step": 6758
+    },
+    {
+      "epoch": 0.8127216978296158,
+      "grad_norm": 1.8481565537196298,
+      "learning_rate": 3.5662349299347906e-07,
+      "loss": 0.6939,
+      "num_input_tokens_seen": 144049020,
+      "step": 6759
+    },
+    {
+      "epoch": 0.812841940720255,
+      "grad_norm": 1.5939232972372117,
+      "learning_rate": 3.561796527331706e-07,
+      "loss": 0.716,
+      "num_input_tokens_seen": 144070415,
+      "step": 6760
+    },
+    {
+      "epoch": 0.812962183610894,
+      "grad_norm": 1.8820097197013812,
+      "learning_rate": 3.557360618391023e-07,
+      "loss": 0.7662,
+      "num_input_tokens_seen": 144090140,
+      "step": 6761
+    },
+    {
+      "epoch": 0.8130824265015331,
+      "grad_norm": 2.142511248968733,
+      "learning_rate": 3.5529272037856493e-07,
+      "loss": 0.7764,
+      "num_input_tokens_seen": 144108075,
+      "step": 6762
+    },
+    {
+      "epoch": 0.8132026693921722,
+      "grad_norm": 0.8245767328882193,
+      "learning_rate": 3.548496284188149e-07,
+      "loss": 0.5792,
+      "num_input_tokens_seen": 144168000,
+      "step": 6763
+    },
+    {
+      "epoch": 0.8133229122828113,
+      "grad_norm": 1.9470619800707958,
+      "learning_rate": 3.544067860270681e-07,
+      "loss": 0.7835,
+      "num_input_tokens_seen": 144185295,
+      "step": 6764
+    },
+    {
+      "epoch": 0.8134431551734503,
+      "grad_norm": 2.7052414425551303,
+      "learning_rate": 3.539641932705035e-07,
+      "loss": 0.7065,
+      "num_input_tokens_seen": 144203495,
+      "step": 6765
+    },
+    {
+      "epoch": 0.8135633980640895,
+      "grad_norm": 2.740916746181709,
+      "learning_rate": 3.535218502162614e-07,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 144222785,
+      "step": 6766
+    },
+    {
+      "epoch": 0.8136836409547286,
+      "grad_norm": 2.006321378693079,
+      "learning_rate": 3.530797569314461e-07,
+      "loss": 0.7662,
+      "num_input_tokens_seen": 144241530,
+      "step": 6767
+    },
+    {
+      "epoch": 0.8138038838453676,
+      "grad_norm": 3.5750600330424738,
+      "learning_rate": 3.5263791348312235e-07,
+      "loss": 0.7726,
+      "num_input_tokens_seen": 144260445,
+      "step": 6768
+    },
+    {
+      "epoch": 0.8139241267360068,
+      "grad_norm": 2.3435695748073626,
+      "learning_rate": 3.521963199383171e-07,
+      "loss": 0.7013,
+      "num_input_tokens_seen": 144283120,
+      "step": 6769
+    },
+    {
+      "epoch": 0.8140443696266458,
+      "grad_norm": 3.762055943219793,
+      "learning_rate": 3.517549763640197e-07,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 144300480,
+      "step": 6770
+    },
+    {
+      "epoch": 0.8141646125172849,
+      "grad_norm": 2.553451781079344,
+      "learning_rate": 3.5131388282718224e-07,
+      "loss": 0.7102,
+      "num_input_tokens_seen": 144320070,
+      "step": 6771
+    },
+    {
+      "epoch": 0.8142848554079241,
+      "grad_norm": 3.4638544394923065,
+      "learning_rate": 3.508730393947179e-07,
+      "loss": 0.6972,
+      "num_input_tokens_seen": 144343045,
+      "step": 6772
+    },
+    {
+      "epoch": 0.8144050982985631,
+      "grad_norm": 1.8236744885837812,
+      "learning_rate": 3.504324461335024e-07,
+      "loss": 0.7118,
+      "num_input_tokens_seen": 144362875,
+      "step": 6773
+    },
+    {
+      "epoch": 0.8145253411892022,
+      "grad_norm": 3.0073805622307894,
+      "learning_rate": 3.499921031103732e-07,
+      "loss": 0.8685,
+      "num_input_tokens_seen": 144383365,
+      "step": 6774
+    },
+    {
+      "epoch": 0.8146455840798413,
+      "grad_norm": 1.956143532793424,
+      "learning_rate": 3.4955201039212987e-07,
+      "loss": 0.7733,
+      "num_input_tokens_seen": 144404005,
+      "step": 6775
+    },
+    {
+      "epoch": 0.8147658269704804,
+      "grad_norm": 2.6089722193477565,
+      "learning_rate": 3.4911216804553465e-07,
+      "loss": 0.641,
+      "num_input_tokens_seen": 144422625,
+      "step": 6776
+    },
+    {
+      "epoch": 0.8148860698611194,
+      "grad_norm": 2.291035423071957,
+      "learning_rate": 3.486725761373106e-07,
+      "loss": 0.707,
+      "num_input_tokens_seen": 144441540,
+      "step": 6777
+    },
+    {
+      "epoch": 0.8150063127517585,
+      "grad_norm": 1.6912839146547407,
+      "learning_rate": 3.4823323473414343e-07,
+      "loss": 0.8455,
+      "num_input_tokens_seen": 144460780,
+      "step": 6778
+    },
+    {
+      "epoch": 0.8151265556423977,
+      "grad_norm": 2.4085970245324115,
+      "learning_rate": 3.477941439026819e-07,
+      "loss": 0.7547,
+      "num_input_tokens_seen": 144478720,
+      "step": 6779
+    },
+    {
+      "epoch": 0.8152467985330367,
+      "grad_norm": 2.2644124108576675,
+      "learning_rate": 3.473553037095349e-07,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 144497465,
+      "step": 6780
+    },
+    {
+      "epoch": 0.8153670414236758,
+      "grad_norm": 2.0315189680606407,
+      "learning_rate": 3.469167142212743e-07,
+      "loss": 0.8208,
+      "num_input_tokens_seen": 144519030,
+      "step": 6781
+    },
+    {
+      "epoch": 0.8154872843143149,
+      "grad_norm": 4.655670528311071,
+      "learning_rate": 3.4647837550443337e-07,
+      "loss": 0.6272,
+      "num_input_tokens_seen": 144537315,
+      "step": 6782
+    },
+    {
+      "epoch": 0.815607527204954,
+      "grad_norm": 1.896099603273935,
+      "learning_rate": 3.460402876255086e-07,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 144554425,
+      "step": 6783
+    },
+    {
+      "epoch": 0.815727770095593,
+      "grad_norm": 4.07434327339687,
+      "learning_rate": 3.456024506509574e-07,
+      "loss": 0.7128,
+      "num_input_tokens_seen": 144575065,
+      "step": 6784
+    },
+    {
+      "epoch": 0.8158480129862322,
+      "grad_norm": 2.1230070914579504,
+      "learning_rate": 3.4516486464719873e-07,
+      "loss": 0.737,
+      "num_input_tokens_seen": 144594175,
+      "step": 6785
+    },
+    {
+      "epoch": 0.8159682558768713,
+      "grad_norm": 1.7757839812508975,
+      "learning_rate": 3.4472752968061403e-07,
+      "loss": 0.6219,
+      "num_input_tokens_seen": 144618325,
+      "step": 6786
+    },
+    {
+      "epoch": 0.8160884987675103,
+      "grad_norm": 2.407705812064104,
+      "learning_rate": 3.442904458175475e-07,
+      "loss": 0.7313,
+      "num_input_tokens_seen": 144635365,
+      "step": 6787
+    },
+    {
+      "epoch": 0.8162087416581495,
+      "grad_norm": 1.5783639154366238,
+      "learning_rate": 3.438536131243037e-07,
+      "loss": 0.7516,
+      "num_input_tokens_seen": 144656245,
+      "step": 6788
+    },
+    {
+      "epoch": 0.8163289845487885,
+      "grad_norm": 2.245599028765818,
+      "learning_rate": 3.434170316671503e-07,
+      "loss": 0.6054,
+      "num_input_tokens_seen": 144680995,
+      "step": 6789
+    },
+    {
+      "epoch": 0.8164492274394276,
+      "grad_norm": 2.385280338446686,
+      "learning_rate": 3.4298070151231583e-07,
+      "loss": 0.8965,
+      "num_input_tokens_seen": 144696115,
+      "step": 6790
+    },
+    {
+      "epoch": 0.8165694703300668,
+      "grad_norm": 2.8780795058389823,
+      "learning_rate": 3.425446227259916e-07,
+      "loss": 0.594,
+      "num_input_tokens_seen": 144716800,
+      "step": 6791
+    },
+    {
+      "epoch": 0.8166897132207058,
+      "grad_norm": 2.4027732823822645,
+      "learning_rate": 3.4210879537433023e-07,
+      "loss": 0.8221,
+      "num_input_tokens_seen": 144736285,
+      "step": 6792
+    },
+    {
+      "epoch": 0.8168099561113449,
+      "grad_norm": 2.516120194959324,
+      "learning_rate": 3.416732195234464e-07,
+      "loss": 0.797,
+      "num_input_tokens_seen": 144756060,
+      "step": 6793
+    },
+    {
+      "epoch": 0.816930199001984,
+      "grad_norm": 1.5330840700177697,
+      "learning_rate": 3.4123789523941613e-07,
+      "loss": 0.7862,
+      "num_input_tokens_seen": 144775605,
+      "step": 6794
+    },
+    {
+      "epoch": 0.8170504418926231,
+      "grad_norm": 1.9323610795302726,
+      "learning_rate": 3.4080282258827884e-07,
+      "loss": 0.6327,
+      "num_input_tokens_seen": 144793700,
+      "step": 6795
+    },
+    {
+      "epoch": 0.8171706847832622,
+      "grad_norm": 3.2559633552140026,
+      "learning_rate": 3.403680016360342e-07,
+      "loss": 0.7135,
+      "num_input_tokens_seen": 144812025,
+      "step": 6796
+    },
+    {
+      "epoch": 0.8172909276739013,
+      "grad_norm": 3.4548089502689283,
+      "learning_rate": 3.3993343244864403e-07,
+      "loss": 0.6709,
+      "num_input_tokens_seen": 144831335,
+      "step": 6797
+    },
+    {
+      "epoch": 0.8174111705645404,
+      "grad_norm": 1.817604040064371,
+      "learning_rate": 3.3949911509203167e-07,
+      "loss": 0.718,
+      "num_input_tokens_seen": 144854175,
+      "step": 6798
+    },
+    {
+      "epoch": 0.8175314134551794,
+      "grad_norm": 2.3568813145078793,
+      "learning_rate": 3.3906504963208396e-07,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 144870590,
+      "step": 6799
+    },
+    {
+      "epoch": 0.8176516563458186,
+      "grad_norm": 2.953986018636422,
+      "learning_rate": 3.3863123613464774e-07,
+      "loss": 0.651,
+      "num_input_tokens_seen": 144889210,
+      "step": 6800
+    },
+    {
+      "epoch": 0.8177718992364577,
+      "grad_norm": 2.5861307961050017,
+      "learning_rate": 3.381976746655317e-07,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 144908685,
+      "step": 6801
+    },
+    {
+      "epoch": 0.8178921421270967,
+      "grad_norm": 2.160117599437323,
+      "learning_rate": 3.3776436529050687e-07,
+      "loss": 0.6738,
+      "num_input_tokens_seen": 144927955,
+      "step": 6802
+    },
+    {
+      "epoch": 0.8180123850177359,
+      "grad_norm": 2.8496104303781524,
+      "learning_rate": 3.3733130807530684e-07,
+      "loss": 0.7155,
+      "num_input_tokens_seen": 144951735,
+      "step": 6803
+    },
+    {
+      "epoch": 0.8181326279083749,
+      "grad_norm": 2.573090701006446,
+      "learning_rate": 3.3689850308562574e-07,
+      "loss": 0.7705,
+      "num_input_tokens_seen": 144971900,
+      "step": 6804
+    },
+    {
+      "epoch": 0.818252870799014,
+      "grad_norm": 1.9294438025692988,
+      "learning_rate": 3.364659503871188e-07,
+      "loss": 0.7709,
+      "num_input_tokens_seen": 144989555,
+      "step": 6805
+    },
+    {
+      "epoch": 0.8183731136896532,
+      "grad_norm": 2.045228053642145,
+      "learning_rate": 3.3603365004540417e-07,
+      "loss": 0.8286,
+      "num_input_tokens_seen": 145007570,
+      "step": 6806
+    },
+    {
+      "epoch": 0.8184933565802922,
+      "grad_norm": 2.4923176712314157,
+      "learning_rate": 3.356016021260624e-07,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 145027620,
+      "step": 6807
+    },
+    {
+      "epoch": 0.8186135994709313,
+      "grad_norm": 3.066955069842093,
+      "learning_rate": 3.35169806694634e-07,
+      "loss": 0.641,
+      "num_input_tokens_seen": 145045590,
+      "step": 6808
+    },
+    {
+      "epoch": 0.8187338423615703,
+      "grad_norm": 0.7377240889855464,
+      "learning_rate": 3.347382638166223e-07,
+      "loss": 0.6213,
+      "num_input_tokens_seen": 145116450,
+      "step": 6809
+    },
+    {
+      "epoch": 0.8188540852522095,
+      "grad_norm": 2.1274000878465413,
+      "learning_rate": 3.343069735574917e-07,
+      "loss": 0.81,
+      "num_input_tokens_seen": 145133860,
+      "step": 6810
+    },
+    {
+      "epoch": 0.8189743281428485,
+      "grad_norm": 3.7813880538337137,
+      "learning_rate": 3.3387593598266907e-07,
+      "loss": 0.7389,
+      "num_input_tokens_seen": 145150190,
+      "step": 6811
+    },
+    {
+      "epoch": 0.8190945710334876,
+      "grad_norm": 1.992086818031112,
+      "learning_rate": 3.3344515115754225e-07,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 145168890,
+      "step": 6812
+    },
+    {
+      "epoch": 0.8192148139241268,
+      "grad_norm": 2.9252191007406734,
+      "learning_rate": 3.33014619147461e-07,
+      "loss": 0.7869,
+      "num_input_tokens_seen": 145186635,
+      "step": 6813
+    },
+    {
+      "epoch": 0.8193350568147658,
+      "grad_norm": 2.4391270049403926,
+      "learning_rate": 3.325843400177362e-07,
+      "loss": 0.712,
+      "num_input_tokens_seen": 145207695,
+      "step": 6814
+    },
+    {
+      "epoch": 0.8194552997054049,
+      "grad_norm": 2.1512663405099866,
+      "learning_rate": 3.3215431383364156e-07,
+      "loss": 0.7268,
+      "num_input_tokens_seen": 145227570,
+      "step": 6815
+    },
+    {
+      "epoch": 0.819575542596044,
+      "grad_norm": 2.362265162117775,
+      "learning_rate": 3.3172454066041164e-07,
+      "loss": 0.6034,
+      "num_input_tokens_seen": 145246795,
+      "step": 6816
+    },
+    {
+      "epoch": 0.8196957854866831,
+      "grad_norm": 2.530884814567089,
+      "learning_rate": 3.3129502056324234e-07,
+      "loss": 0.7545,
+      "num_input_tokens_seen": 145267880,
+      "step": 6817
+    },
+    {
+      "epoch": 0.8198160283773221,
+      "grad_norm": 0.8186426484697743,
+      "learning_rate": 3.3086575360729165e-07,
+      "loss": 0.6282,
+      "num_input_tokens_seen": 145325135,
+      "step": 6818
+    },
+    {
+      "epoch": 0.8199362712679613,
+      "grad_norm": 1.9541116357457324,
+      "learning_rate": 3.3043673985767906e-07,
+      "loss": 0.7085,
+      "num_input_tokens_seen": 145343920,
+      "step": 6819
+    },
+    {
+      "epoch": 0.8200565141586004,
+      "grad_norm": 1.995718698350831,
+      "learning_rate": 3.3000797937948564e-07,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 145361935,
+      "step": 6820
+    },
+    {
+      "epoch": 0.8201767570492394,
+      "grad_norm": 0.9576204151446579,
+      "learning_rate": 3.2957947223775384e-07,
+      "loss": 0.6772,
+      "num_input_tokens_seen": 145425260,
+      "step": 6821
+    },
+    {
+      "epoch": 0.8202969999398786,
+      "grad_norm": 1.9824797352389605,
+      "learning_rate": 3.291512184974876e-07,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 145445370,
+      "step": 6822
+    },
+    {
+      "epoch": 0.8204172428305176,
+      "grad_norm": 2.0286719449556863,
+      "learning_rate": 3.2872321822365346e-07,
+      "loss": 0.6679,
+      "num_input_tokens_seen": 145465305,
+      "step": 6823
+    },
+    {
+      "epoch": 0.8205374857211567,
+      "grad_norm": 2.0360447715407815,
+      "learning_rate": 3.282954714811783e-07,
+      "loss": 0.7324,
+      "num_input_tokens_seen": 145483930,
+      "step": 6824
+    },
+    {
+      "epoch": 0.8206577286117959,
+      "grad_norm": 2.4382427972761773,
+      "learning_rate": 3.2786797833495093e-07,
+      "loss": 0.702,
+      "num_input_tokens_seen": 145499005,
+      "step": 6825
+    },
+    {
+      "epoch": 0.8207779715024349,
+      "grad_norm": 2.2596869139181464,
+      "learning_rate": 3.274407388498213e-07,
+      "loss": 0.7199,
+      "num_input_tokens_seen": 145516855,
+      "step": 6826
+    },
+    {
+      "epoch": 0.820898214393074,
+      "grad_norm": 3.26530753405961,
+      "learning_rate": 3.270137530906021e-07,
+      "loss": 0.7368,
+      "num_input_tokens_seen": 145535810,
+      "step": 6827
+    },
+    {
+      "epoch": 0.8210184572837131,
+      "grad_norm": 2.1316511086561487,
+      "learning_rate": 3.265870211220665e-07,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 145553365,
+      "step": 6828
+    },
+    {
+      "epoch": 0.8211387001743522,
+      "grad_norm": 3.775283647238993,
+      "learning_rate": 3.2616054300894934e-07,
+      "loss": 0.8091,
+      "num_input_tokens_seen": 145572535,
+      "step": 6829
+    },
+    {
+      "epoch": 0.8212589430649913,
+      "grad_norm": 2.1094774104931644,
+      "learning_rate": 3.257343188159465e-07,
+      "loss": 0.841,
+      "num_input_tokens_seen": 145591800,
+      "step": 6830
+    },
+    {
+      "epoch": 0.8213791859556304,
+      "grad_norm": 2.5451732459221934,
+      "learning_rate": 3.2530834860771663e-07,
+      "loss": 0.6527,
+      "num_input_tokens_seen": 145610900,
+      "step": 6831
+    },
+    {
+      "epoch": 0.8214994288462695,
+      "grad_norm": 3.1006880998731217,
+      "learning_rate": 3.248826324488789e-07,
+      "loss": 0.7307,
+      "num_input_tokens_seen": 145627915,
+      "step": 6832
+    },
+    {
+      "epoch": 0.8216196717369085,
+      "grad_norm": 2.857064873872957,
+      "learning_rate": 3.244571704040138e-07,
+      "loss": 0.8705,
+      "num_input_tokens_seen": 145647795,
+      "step": 6833
+    },
+    {
+      "epoch": 0.8217399146275477,
+      "grad_norm": 2.197935822779527,
+      "learning_rate": 3.2403196253766374e-07,
+      "loss": 0.7306,
+      "num_input_tokens_seen": 145666595,
+      "step": 6834
+    },
+    {
+      "epoch": 0.8218601575181868,
+      "grad_norm": 2.4457567958888453,
+      "learning_rate": 3.2360700891433254e-07,
+      "loss": 0.7832,
+      "num_input_tokens_seen": 145685340,
+      "step": 6835
+    },
+    {
+      "epoch": 0.8219804004088258,
+      "grad_norm": 0.8258212032887101,
+      "learning_rate": 3.2318230959848513e-07,
+      "loss": 0.5951,
+      "num_input_tokens_seen": 145739700,
+      "step": 6836
+    },
+    {
+      "epoch": 0.822100643299465,
+      "grad_norm": 2.9037616186578186,
+      "learning_rate": 3.2275786465454814e-07,
+      "loss": 0.7471,
+      "num_input_tokens_seen": 145756070,
+      "step": 6837
+    },
+    {
+      "epoch": 0.822220886190104,
+      "grad_norm": 2.4724078103708926,
+      "learning_rate": 3.2233367414690917e-07,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 145777980,
+      "step": 6838
+    },
+    {
+      "epoch": 0.8223411290807431,
+      "grad_norm": 2.2402347064016834,
+      "learning_rate": 3.219097381399183e-07,
+      "loss": 0.8355,
+      "num_input_tokens_seen": 145794875,
+      "step": 6839
+    },
+    {
+      "epoch": 0.8224613719713821,
+      "grad_norm": 1.8410913141495147,
+      "learning_rate": 3.2148605669788584e-07,
+      "loss": 0.7968,
+      "num_input_tokens_seen": 145814485,
+      "step": 6840
+    },
+    {
+      "epoch": 0.8225816148620213,
+      "grad_norm": 20.45698117660499,
+      "learning_rate": 3.2106262988508405e-07,
+      "loss": 0.7653,
+      "num_input_tokens_seen": 145832255,
+      "step": 6841
+    },
+    {
+      "epoch": 0.8227018577526604,
+      "grad_norm": 3.438138768387695,
+      "learning_rate": 3.206394577657461e-07,
+      "loss": 0.7331,
+      "num_input_tokens_seen": 145849755,
+      "step": 6842
+    },
+    {
+      "epoch": 0.8228221006432994,
+      "grad_norm": 3.1820139605565383,
+      "learning_rate": 3.202165404040675e-07,
+      "loss": 0.712,
+      "num_input_tokens_seen": 145867395,
+      "step": 6843
+    },
+    {
+      "epoch": 0.8229423435339386,
+      "grad_norm": 3.006719561106939,
+      "learning_rate": 3.1979387786420396e-07,
+      "loss": 0.7427,
+      "num_input_tokens_seen": 145887355,
+      "step": 6844
+    },
+    {
+      "epoch": 0.8230625864245776,
+      "grad_norm": 2.439902438076832,
+      "learning_rate": 3.1937147021027346e-07,
+      "loss": 0.818,
+      "num_input_tokens_seen": 145905530,
+      "step": 6845
+    },
+    {
+      "epoch": 0.8231828293152167,
+      "grad_norm": 9.404427719808812,
+      "learning_rate": 3.189493175063542e-07,
+      "loss": 0.7611,
+      "num_input_tokens_seen": 145922485,
+      "step": 6846
+    },
+    {
+      "epoch": 0.8233030722058559,
+      "grad_norm": 2.2164797040193887,
+      "learning_rate": 3.1852741981648776e-07,
+      "loss": 0.6694,
+      "num_input_tokens_seen": 145940855,
+      "step": 6847
+    },
+    {
+      "epoch": 0.8234233150964949,
+      "grad_norm": 2.614384598470918,
+      "learning_rate": 3.1810577720467446e-07,
+      "loss": 0.6945,
+      "num_input_tokens_seen": 145962305,
+      "step": 6848
+    },
+    {
+      "epoch": 0.823543557987134,
+      "grad_norm": 1.6186250296599518,
+      "learning_rate": 3.176843897348773e-07,
+      "loss": 0.564,
+      "num_input_tokens_seen": 145985220,
+      "step": 6849
+    },
+    {
+      "epoch": 0.8236638008777731,
+      "grad_norm": 6.218079243556821,
+      "learning_rate": 3.1726325747102034e-07,
+      "loss": 0.7526,
+      "num_input_tokens_seen": 146003315,
+      "step": 6850
+    },
+    {
+      "epoch": 0.8237840437684122,
+      "grad_norm": 1.555671543908552,
+      "learning_rate": 3.1684238047698974e-07,
+      "loss": 0.6364,
+      "num_input_tokens_seen": 146031305,
+      "step": 6851
+    },
+    {
+      "epoch": 0.8239042866590512,
+      "grad_norm": 2.7926815256262767,
+      "learning_rate": 3.1642175881663155e-07,
+      "loss": 0.5268,
+      "num_input_tokens_seen": 146050755,
+      "step": 6852
+    },
+    {
+      "epoch": 0.8240245295496904,
+      "grad_norm": 2.084756321585468,
+      "learning_rate": 3.1600139255375413e-07,
+      "loss": 0.8388,
+      "num_input_tokens_seen": 146071310,
+      "step": 6853
+    },
+    {
+      "epoch": 0.8241447724403295,
+      "grad_norm": 7.215655114401023,
+      "learning_rate": 3.1558128175212615e-07,
+      "loss": 0.7508,
+      "num_input_tokens_seen": 146091405,
+      "step": 6854
+    },
+    {
+      "epoch": 0.8242650153309685,
+      "grad_norm": 2.3119603509379276,
+      "learning_rate": 3.151614264754787e-07,
+      "loss": 0.7701,
+      "num_input_tokens_seen": 146109070,
+      "step": 6855
+    },
+    {
+      "epoch": 0.8243852582216077,
+      "grad_norm": 2.440012365661594,
+      "learning_rate": 3.147418267875035e-07,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 146126920,
+      "step": 6856
+    },
+    {
+      "epoch": 0.8245055011122467,
+      "grad_norm": 3.1584048500821402,
+      "learning_rate": 3.1432248275185315e-07,
+      "loss": 0.6533,
+      "num_input_tokens_seen": 146147150,
+      "step": 6857
+    },
+    {
+      "epoch": 0.8246257440028858,
+      "grad_norm": 2.470926391087847,
+      "learning_rate": 3.139033944321412e-07,
+      "loss": 0.7737,
+      "num_input_tokens_seen": 146164230,
+      "step": 6858
+    },
+    {
+      "epoch": 0.824745986893525,
+      "grad_norm": 2.1079992570074575,
+      "learning_rate": 3.134845618919444e-07,
+      "loss": 0.784,
+      "num_input_tokens_seen": 146184410,
+      "step": 6859
+    },
+    {
+      "epoch": 0.824866229784164,
+      "grad_norm": 1.7752250146199624,
+      "learning_rate": 3.1306598519479876e-07,
+      "loss": 0.8232,
+      "num_input_tokens_seen": 146203950,
+      "step": 6860
+    },
+    {
+      "epoch": 0.8249864726748031,
+      "grad_norm": 1.7049616475349059,
+      "learning_rate": 3.1264766440420177e-07,
+      "loss": 0.7793,
+      "num_input_tokens_seen": 146226140,
+      "step": 6861
+    },
+    {
+      "epoch": 0.8251067155654422,
+      "grad_norm": 2.177157480033531,
+      "learning_rate": 3.122295995836124e-07,
+      "loss": 0.672,
+      "num_input_tokens_seen": 146245730,
+      "step": 6862
+    },
+    {
+      "epoch": 0.8252269584560813,
+      "grad_norm": 2.7597382259785492,
+      "learning_rate": 3.118117907964508e-07,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 146267395,
+      "step": 6863
+    },
+    {
+      "epoch": 0.8253472013467203,
+      "grad_norm": 2.3260151631472286,
+      "learning_rate": 3.1139423810609856e-07,
+      "loss": 0.801,
+      "num_input_tokens_seen": 146283810,
+      "step": 6864
+    },
+    {
+      "epoch": 0.8254674442373595,
+      "grad_norm": 3.1430570449193564,
+      "learning_rate": 3.109769415758976e-07,
+      "loss": 0.753,
+      "num_input_tokens_seen": 146303415,
+      "step": 6865
+    },
+    {
+      "epoch": 0.8255876871279986,
+      "grad_norm": 5.230790794251164,
+      "learning_rate": 3.105599012691511e-07,
+      "loss": 0.7574,
+      "num_input_tokens_seen": 146321565,
+      "step": 6866
+    },
+    {
+      "epoch": 0.8257079300186376,
+      "grad_norm": 1.6259356643546288,
+      "learning_rate": 3.101431172491249e-07,
+      "loss": 0.8168,
+      "num_input_tokens_seen": 146342830,
+      "step": 6867
+    },
+    {
+      "epoch": 0.8258281729092768,
+      "grad_norm": 2.649303264881949,
+      "learning_rate": 3.097265895790444e-07,
+      "loss": 0.7146,
+      "num_input_tokens_seen": 146360760,
+      "step": 6868
+    },
+    {
+      "epoch": 0.8259484157999158,
+      "grad_norm": 2.2360640517528574,
+      "learning_rate": 3.093103183220962e-07,
+      "loss": 0.8294,
+      "num_input_tokens_seen": 146380525,
+      "step": 6869
+    },
+    {
+      "epoch": 0.8260686586905549,
+      "grad_norm": 0.9414163219170635,
+      "learning_rate": 3.0889430354142796e-07,
+      "loss": 0.6524,
+      "num_input_tokens_seen": 146441755,
+      "step": 6870
+    },
+    {
+      "epoch": 0.826188901581194,
+      "grad_norm": 4.690599122356733,
+      "learning_rate": 3.084785453001497e-07,
+      "loss": 0.6928,
+      "num_input_tokens_seen": 146462390,
+      "step": 6871
+    },
+    {
+      "epoch": 0.8263091444718331,
+      "grad_norm": 2.4681720492792527,
+      "learning_rate": 3.080630436613314e-07,
+      "loss": 0.809,
+      "num_input_tokens_seen": 146479880,
+      "step": 6872
+    },
+    {
+      "epoch": 0.8264293873624722,
+      "grad_norm": 2.6533995089520195,
+      "learning_rate": 3.076477986880039e-07,
+      "loss": 0.8509,
+      "num_input_tokens_seen": 146497395,
+      "step": 6873
+    },
+    {
+      "epoch": 0.8265496302531112,
+      "grad_norm": 2.0374102606125555,
+      "learning_rate": 3.072328104431594e-07,
+      "loss": 0.6846,
+      "num_input_tokens_seen": 146519070,
+      "step": 6874
+    },
+    {
+      "epoch": 0.8266698731437504,
+      "grad_norm": 2.310733761696155,
+      "learning_rate": 3.068180789897521e-07,
+      "loss": 0.7606,
+      "num_input_tokens_seen": 146537200,
+      "step": 6875
+    },
+    {
+      "epoch": 0.8267901160343895,
+      "grad_norm": 1.6052174509603416,
+      "learning_rate": 3.064036043906961e-07,
+      "loss": 0.8095,
+      "num_input_tokens_seen": 146560360,
+      "step": 6876
+    },
+    {
+      "epoch": 0.8269103589250285,
+      "grad_norm": 2.173445408744171,
+      "learning_rate": 3.059893867088668e-07,
+      "loss": 0.6718,
+      "num_input_tokens_seen": 146584225,
+      "step": 6877
+    },
+    {
+      "epoch": 0.8270306018156677,
+      "grad_norm": 2.170178064455718,
+      "learning_rate": 3.055754260071004e-07,
+      "loss": 0.6628,
+      "num_input_tokens_seen": 146606240,
+      "step": 6878
+    },
+    {
+      "epoch": 0.8271508447063067,
+      "grad_norm": 2.031145438682488,
+      "learning_rate": 3.051617223481948e-07,
+      "loss": 0.7335,
+      "num_input_tokens_seen": 146627280,
+      "step": 6879
+    },
+    {
+      "epoch": 0.8272710875969458,
+      "grad_norm": 2.6585038131449275,
+      "learning_rate": 3.0474827579490825e-07,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 146644630,
+      "step": 6880
+    },
+    {
+      "epoch": 0.827391330487585,
+      "grad_norm": 1.9611839149006123,
+      "learning_rate": 3.043350864099605e-07,
+      "loss": 0.8422,
+      "num_input_tokens_seen": 146662910,
+      "step": 6881
+    },
+    {
+      "epoch": 0.827511573378224,
+      "grad_norm": 4.792019379462137,
+      "learning_rate": 3.039221542560315e-07,
+      "loss": 0.805,
+      "num_input_tokens_seen": 146679195,
+      "step": 6882
+    },
+    {
+      "epoch": 0.8276318162688631,
+      "grad_norm": 2.2181429408365365,
+      "learning_rate": 3.0350947939576356e-07,
+      "loss": 0.742,
+      "num_input_tokens_seen": 146698070,
+      "step": 6883
+    },
+    {
+      "epoch": 0.8277520591595022,
+      "grad_norm": 1.8573367472872755,
+      "learning_rate": 3.0309706189175876e-07,
+      "loss": 0.7212,
+      "num_input_tokens_seen": 146717625,
+      "step": 6884
+    },
+    {
+      "epoch": 0.8278723020501413,
+      "grad_norm": 0.8388458037637345,
+      "learning_rate": 3.0268490180658045e-07,
+      "loss": 0.5944,
+      "num_input_tokens_seen": 146780125,
+      "step": 6885
+    },
+    {
+      "epoch": 0.8279925449407803,
+      "grad_norm": 2.239293896097615,
+      "learning_rate": 3.0227299920275263e-07,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 146796160,
+      "step": 6886
+    },
+    {
+      "epoch": 0.8281127878314195,
+      "grad_norm": 3.2239081546695596,
+      "learning_rate": 3.018613541427613e-07,
+      "loss": 0.8492,
+      "num_input_tokens_seen": 146815400,
+      "step": 6887
+    },
+    {
+      "epoch": 0.8282330307220586,
+      "grad_norm": 31.248229298434207,
+      "learning_rate": 3.0144996668905243e-07,
+      "loss": 0.7338,
+      "num_input_tokens_seen": 146832500,
+      "step": 6888
+    },
+    {
+      "epoch": 0.8283532736126976,
+      "grad_norm": 3.4196753874493604,
+      "learning_rate": 3.010388369040331e-07,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 146850880,
+      "step": 6889
+    },
+    {
+      "epoch": 0.8284735165033368,
+      "grad_norm": 2.144420070344818,
+      "learning_rate": 3.006279648500709e-07,
+      "loss": 0.8215,
+      "num_input_tokens_seen": 146871540,
+      "step": 6890
+    },
+    {
+      "epoch": 0.8285937593939758,
+      "grad_norm": 3.698852984919603,
+      "learning_rate": 3.002173505894965e-07,
+      "loss": 0.6444,
+      "num_input_tokens_seen": 146890410,
+      "step": 6891
+    },
+    {
+      "epoch": 0.8287140022846149,
+      "grad_norm": 3.25734162111977,
+      "learning_rate": 2.9980699418459774e-07,
+      "loss": 0.6195,
+      "num_input_tokens_seen": 146909200,
+      "step": 6892
+    },
+    {
+      "epoch": 0.8288342451752541,
+      "grad_norm": 4.654758041637784,
+      "learning_rate": 2.993968956976263e-07,
+      "loss": 0.6173,
+      "num_input_tokens_seen": 146976665,
+      "step": 6893
+    },
+    {
+      "epoch": 0.8289544880658931,
+      "grad_norm": 3.3273314773725917,
+      "learning_rate": 2.9898705519079313e-07,
+      "loss": 0.6898,
+      "num_input_tokens_seen": 146995490,
+      "step": 6894
+    },
+    {
+      "epoch": 0.8290747309565322,
+      "grad_norm": 1.9948104495682495,
+      "learning_rate": 2.985774727262715e-07,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 147014055,
+      "step": 6895
+    },
+    {
+      "epoch": 0.8291949738471713,
+      "grad_norm": 2.10140458208863,
+      "learning_rate": 2.981681483661949e-07,
+      "loss": 0.8057,
+      "num_input_tokens_seen": 147033360,
+      "step": 6896
+    },
+    {
+      "epoch": 0.8293152167378104,
+      "grad_norm": 1.6378127293717482,
+      "learning_rate": 2.9775908217265675e-07,
+      "loss": 0.7027,
+      "num_input_tokens_seen": 147058315,
+      "step": 6897
+    },
+    {
+      "epoch": 0.8294354596284494,
+      "grad_norm": 0.8402322842744447,
+      "learning_rate": 2.973502742077121e-07,
+      "loss": 0.524,
+      "num_input_tokens_seen": 147118370,
+      "step": 6898
+    },
+    {
+      "epoch": 0.8295557025190886,
+      "grad_norm": 1.9221327873232499,
+      "learning_rate": 2.969417245333774e-07,
+      "loss": 0.7092,
+      "num_input_tokens_seen": 147137470,
+      "step": 6899
+    },
+    {
+      "epoch": 0.8296759454097277,
+      "grad_norm": 4.40503988085388,
+      "learning_rate": 2.9653343321162915e-07,
+      "loss": 0.7789,
+      "num_input_tokens_seen": 147156700,
+      "step": 6900
+    },
+    {
+      "epoch": 0.8297961883003667,
+      "grad_norm": 2.268618319649715,
+      "learning_rate": 2.9612540030440446e-07,
+      "loss": 0.6381,
+      "num_input_tokens_seen": 147176965,
+      "step": 6901
+    },
+    {
+      "epoch": 0.8299164311910058,
+      "grad_norm": 0.8806013932452963,
+      "learning_rate": 2.957176258736016e-07,
+      "loss": 0.6473,
+      "num_input_tokens_seen": 147233070,
+      "step": 6902
+    },
+    {
+      "epoch": 0.8300366740816449,
+      "grad_norm": 1.7415016845063431,
+      "learning_rate": 2.953101099810802e-07,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 147252395,
+      "step": 6903
+    },
+    {
+      "epoch": 0.830156916972284,
+      "grad_norm": 6.696914305963923,
+      "learning_rate": 2.9490285268865965e-07,
+      "loss": 0.8357,
+      "num_input_tokens_seen": 147269605,
+      "step": 6904
+    },
+    {
+      "epoch": 0.830277159862923,
+      "grad_norm": 2.3844055815975507,
+      "learning_rate": 2.9449585405812085e-07,
+      "loss": 0.7999,
+      "num_input_tokens_seen": 147286705,
+      "step": 6905
+    },
+    {
+      "epoch": 0.8303974027535622,
+      "grad_norm": 2.1866853018681685,
+      "learning_rate": 2.940891141512047e-07,
+      "loss": 0.7401,
+      "num_input_tokens_seen": 147304445,
+      "step": 6906
+    },
+    {
+      "epoch": 0.8305176456442013,
+      "grad_norm": 2.519838971183083,
+      "learning_rate": 2.9368263302961385e-07,
+      "loss": 0.7092,
+      "num_input_tokens_seen": 147322865,
+      "step": 6907
+    },
+    {
+      "epoch": 0.8306378885348403,
+      "grad_norm": 2.8178309203404464,
+      "learning_rate": 2.9327641075501075e-07,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 147341575,
+      "step": 6908
+    },
+    {
+      "epoch": 0.8307581314254795,
+      "grad_norm": 3.795649059275563,
+      "learning_rate": 2.9287044738901913e-07,
+      "loss": 0.6518,
+      "num_input_tokens_seen": 147359280,
+      "step": 6909
+    },
+    {
+      "epoch": 0.8308783743161186,
+      "grad_norm": 9.297190464101524,
+      "learning_rate": 2.9246474299322274e-07,
+      "loss": 0.9062,
+      "num_input_tokens_seen": 147374560,
+      "step": 6910
+    },
+    {
+      "epoch": 0.8309986172067576,
+      "grad_norm": 0.9338561477490201,
+      "learning_rate": 2.920592976291678e-07,
+      "loss": 0.6629,
+      "num_input_tokens_seen": 147431610,
+      "step": 6911
+    },
+    {
+      "epoch": 0.8311188600973968,
+      "grad_norm": 2.24408649308667,
+      "learning_rate": 2.916541113583595e-07,
+      "loss": 0.8054,
+      "num_input_tokens_seen": 147449830,
+      "step": 6912
+    },
+    {
+      "epoch": 0.8312391029880358,
+      "grad_norm": 3.7041626483805925,
+      "learning_rate": 2.912491842422642e-07,
+      "loss": 0.6528,
+      "num_input_tokens_seen": 147467255,
+      "step": 6913
+    },
+    {
+      "epoch": 0.8313593458786749,
+      "grad_norm": 1.734805722474592,
+      "learning_rate": 2.9084451634230857e-07,
+      "loss": 0.6973,
+      "num_input_tokens_seen": 147486275,
+      "step": 6914
+    },
+    {
+      "epoch": 0.831479588769314,
+      "grad_norm": 2.259316725953428,
+      "learning_rate": 2.9044010771988125e-07,
+      "loss": 0.7114,
+      "num_input_tokens_seen": 147505810,
+      "step": 6915
+    },
+    {
+      "epoch": 0.8315998316599531,
+      "grad_norm": 1.9088973247391254,
+      "learning_rate": 2.900359584363303e-07,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 147528635,
+      "step": 6916
+    },
+    {
+      "epoch": 0.8317200745505922,
+      "grad_norm": 2.821963192012367,
+      "learning_rate": 2.8963206855296494e-07,
+      "loss": 0.8317,
+      "num_input_tokens_seen": 147544595,
+      "step": 6917
+    },
+    {
+      "epoch": 0.8318403174412313,
+      "grad_norm": 1.9870771864690584,
+      "learning_rate": 2.8922843813105437e-07,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 147565730,
+      "step": 6918
+    },
+    {
+      "epoch": 0.8319605603318704,
+      "grad_norm": 2.9016051919697614,
+      "learning_rate": 2.888250672318302e-07,
+      "loss": 0.707,
+      "num_input_tokens_seen": 147582850,
+      "step": 6919
+    },
+    {
+      "epoch": 0.8320808032225094,
+      "grad_norm": 4.27053055477984,
+      "learning_rate": 2.8842195591648243e-07,
+      "loss": 0.6867,
+      "num_input_tokens_seen": 147605715,
+      "step": 6920
+    },
+    {
+      "epoch": 0.8322010461131486,
+      "grad_norm": 2.0286754341757165,
+      "learning_rate": 2.880191042461635e-07,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 147621375,
+      "step": 6921
+    },
+    {
+      "epoch": 0.8323212890037877,
+      "grad_norm": 2.398742455431569,
+      "learning_rate": 2.876165122819849e-07,
+      "loss": 0.7981,
+      "num_input_tokens_seen": 147639075,
+      "step": 6922
+    },
+    {
+      "epoch": 0.8324415318944267,
+      "grad_norm": 1.903549917531399,
+      "learning_rate": 2.872141800850201e-07,
+      "loss": 0.7856,
+      "num_input_tokens_seen": 147655970,
+      "step": 6923
+    },
+    {
+      "epoch": 0.8325617747850659,
+      "grad_norm": 1.8428676634987566,
+      "learning_rate": 2.868121077163024e-07,
+      "loss": 0.7261,
+      "num_input_tokens_seen": 147675245,
+      "step": 6924
+    },
+    {
+      "epoch": 0.8326820176757049,
+      "grad_norm": 1.8544383886565183,
+      "learning_rate": 2.864102952368257e-07,
+      "loss": 0.7153,
+      "num_input_tokens_seen": 147692890,
+      "step": 6925
+    },
+    {
+      "epoch": 0.832802260566344,
+      "grad_norm": 1.4953249974212524,
+      "learning_rate": 2.860087427075444e-07,
+      "loss": 0.5917,
+      "num_input_tokens_seen": 147716860,
+      "step": 6926
+    },
+    {
+      "epoch": 0.8329225034569832,
+      "grad_norm": 2.702330756249681,
+      "learning_rate": 2.856074501893744e-07,
+      "loss": 0.8583,
+      "num_input_tokens_seen": 147731780,
+      "step": 6927
+    },
+    {
+      "epoch": 0.8330427463476222,
+      "grad_norm": 1.8647738549899766,
+      "learning_rate": 2.8520641774319097e-07,
+      "loss": 0.8111,
+      "num_input_tokens_seen": 147749590,
+      "step": 6928
+    },
+    {
+      "epoch": 0.8331629892382613,
+      "grad_norm": 2.2931058763702103,
+      "learning_rate": 2.848056454298309e-07,
+      "loss": 0.7469,
+      "num_input_tokens_seen": 147766635,
+      "step": 6929
+    },
+    {
+      "epoch": 0.8332832321289004,
+      "grad_norm": 4.024077278275758,
+      "learning_rate": 2.844051333100901e-07,
+      "loss": 0.652,
+      "num_input_tokens_seen": 147783900,
+      "step": 6930
+    },
+    {
+      "epoch": 0.8334034750195395,
+      "grad_norm": 2.398267877039075,
+      "learning_rate": 2.840048814447269e-07,
+      "loss": 0.8361,
+      "num_input_tokens_seen": 147801785,
+      "step": 6931
+    },
+    {
+      "epoch": 0.8335237179101785,
+      "grad_norm": 2.6760525134882385,
+      "learning_rate": 2.836048898944587e-07,
+      "loss": 0.7302,
+      "num_input_tokens_seen": 147819930,
+      "step": 6932
+    },
+    {
+      "epoch": 0.8336439608008177,
+      "grad_norm": 3.639357120635123,
+      "learning_rate": 2.832051587199642e-07,
+      "loss": 0.7155,
+      "num_input_tokens_seen": 147836905,
+      "step": 6933
+    },
+    {
+      "epoch": 0.8337642036914568,
+      "grad_norm": 0.8003977797803753,
+      "learning_rate": 2.828056879818821e-07,
+      "loss": 0.6068,
+      "num_input_tokens_seen": 147895700,
+      "step": 6934
+    },
+    {
+      "epoch": 0.8338844465820958,
+      "grad_norm": 2.6400155919357906,
+      "learning_rate": 2.824064777408117e-07,
+      "loss": 0.8272,
+      "num_input_tokens_seen": 147915210,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8340046894727349,
+      "grad_norm": 2.7805958423433967,
+      "learning_rate": 2.820075280573131e-07,
+      "loss": 0.7531,
+      "num_input_tokens_seen": 147937920,
+      "step": 6936
+    },
+    {
+      "epoch": 0.834124932363374,
+      "grad_norm": 2.284618252767419,
+      "learning_rate": 2.8160883899190667e-07,
+      "loss": 0.8012,
+      "num_input_tokens_seen": 147960910,
+      "step": 6937
+    },
+    {
+      "epoch": 0.8342451752540131,
+      "grad_norm": 2.429735895223449,
+      "learning_rate": 2.8121041060507234e-07,
+      "loss": 0.736,
+      "num_input_tokens_seen": 147979660,
+      "step": 6938
+    },
+    {
+      "epoch": 0.8343654181446521,
+      "grad_norm": 1.703920002955549,
+      "learning_rate": 2.808122429572528e-07,
+      "loss": 0.7079,
+      "num_input_tokens_seen": 147999585,
+      "step": 6939
+    },
+    {
+      "epoch": 0.8344856610352913,
+      "grad_norm": 11.686142348938322,
+      "learning_rate": 2.804143361088489e-07,
+      "loss": 0.7478,
+      "num_input_tokens_seen": 148018485,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8346059039259304,
+      "grad_norm": 2.320912860524297,
+      "learning_rate": 2.800166901202232e-07,
+      "loss": 0.7622,
+      "num_input_tokens_seen": 148036175,
+      "step": 6941
+    },
+    {
+      "epoch": 0.8347261468165694,
+      "grad_norm": 1.823295875907801,
+      "learning_rate": 2.796193050516975e-07,
+      "loss": 0.6926,
+      "num_input_tokens_seen": 148060140,
+      "step": 6942
+    },
+    {
+      "epoch": 0.8348463897072086,
+      "grad_norm": 4.279533183587005,
+      "learning_rate": 2.792221809635558e-07,
+      "loss": 0.7553,
+      "num_input_tokens_seen": 148080490,
+      "step": 6943
+    },
+    {
+      "epoch": 0.8349666325978476,
+      "grad_norm": 2.5403554696079134,
+      "learning_rate": 2.788253179160411e-07,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 148101370,
+      "step": 6944
+    },
+    {
+      "epoch": 0.8350868754884867,
+      "grad_norm": 2.103267811249552,
+      "learning_rate": 2.7842871596935725e-07,
+      "loss": 0.6489,
+      "num_input_tokens_seen": 148119605,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8352071183791259,
+      "grad_norm": 1.7235735954889084,
+      "learning_rate": 2.780323751836677e-07,
+      "loss": 0.6837,
+      "num_input_tokens_seen": 148140540,
+      "step": 6946
+    },
+    {
+      "epoch": 0.8353273612697649,
+      "grad_norm": 1.8385923208278383,
+      "learning_rate": 2.776362956190983e-07,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 148161090,
+      "step": 6947
+    },
+    {
+      "epoch": 0.835447604160404,
+      "grad_norm": 2.1937308038126324,
+      "learning_rate": 2.772404773357335e-07,
+      "loss": 0.7618,
+      "num_input_tokens_seen": 148180215,
+      "step": 6948
+    },
+    {
+      "epoch": 0.8355678470510431,
+      "grad_norm": 6.144809029585824,
+      "learning_rate": 2.7684492039361853e-07,
+      "loss": 0.7701,
+      "num_input_tokens_seen": 148199160,
+      "step": 6949
+    },
+    {
+      "epoch": 0.8356880899416822,
+      "grad_norm": 2.101532973335265,
+      "learning_rate": 2.76449624852759e-07,
+      "loss": 0.832,
+      "num_input_tokens_seen": 148217855,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8358083328323213,
+      "grad_norm": 3.5760020255416882,
+      "learning_rate": 2.760545907731211e-07,
+      "loss": 0.776,
+      "num_input_tokens_seen": 148238150,
+      "step": 6951
+    },
+    {
+      "epoch": 0.8359285757229604,
+      "grad_norm": 2.441141431187933,
+      "learning_rate": 2.75659818214631e-07,
+      "loss": 0.6717,
+      "num_input_tokens_seen": 148258975,
+      "step": 6952
+    },
+    {
+      "epoch": 0.8360488186135995,
+      "grad_norm": 2.063119426133086,
+      "learning_rate": 2.7526530723717534e-07,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 148278130,
+      "step": 6953
+    },
+    {
+      "epoch": 0.8361690615042385,
+      "grad_norm": 2.2532227426526137,
+      "learning_rate": 2.7487105790060105e-07,
+      "loss": 0.7393,
+      "num_input_tokens_seen": 148297260,
+      "step": 6954
+    },
+    {
+      "epoch": 0.8362893043948777,
+      "grad_norm": 2.2361013395692475,
+      "learning_rate": 2.7447707026471587e-07,
+      "loss": 0.6872,
+      "num_input_tokens_seen": 148319955,
+      "step": 6955
+    },
+    {
+      "epoch": 0.8364095472855168,
+      "grad_norm": 1.9918158716260648,
+      "learning_rate": 2.740833443892874e-07,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 148337845,
+      "step": 6956
+    },
+    {
+      "epoch": 0.8365297901761558,
+      "grad_norm": 1.9507490717495841,
+      "learning_rate": 2.7368988033404327e-07,
+      "loss": 0.7852,
+      "num_input_tokens_seen": 148355080,
+      "step": 6957
+    },
+    {
+      "epoch": 0.836650033066795,
+      "grad_norm": 1.5961202174944367,
+      "learning_rate": 2.732966781586712e-07,
+      "loss": 0.8453,
+      "num_input_tokens_seen": 148374545,
+      "step": 6958
+    },
+    {
+      "epoch": 0.836770275957434,
+      "grad_norm": 1.9101106254508742,
+      "learning_rate": 2.729037379228205e-07,
+      "loss": 0.6685,
+      "num_input_tokens_seen": 148394450,
+      "step": 6959
+    },
+    {
+      "epoch": 0.8368905188480731,
+      "grad_norm": 1.7039840853833346,
+      "learning_rate": 2.725110596860998e-07,
+      "loss": 0.7996,
+      "num_input_tokens_seen": 148414850,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8370107617387123,
+      "grad_norm": 2.502650459534539,
+      "learning_rate": 2.7211864350807776e-07,
+      "loss": 0.6962,
+      "num_input_tokens_seen": 148432770,
+      "step": 6961
+    },
+    {
+      "epoch": 0.8371310046293513,
+      "grad_norm": 2.026286214006468,
+      "learning_rate": 2.717264894482831e-07,
+      "loss": 0.7325,
+      "num_input_tokens_seen": 148452830,
+      "step": 6962
+    },
+    {
+      "epoch": 0.8372512475199904,
+      "grad_norm": 4.324551923176521,
+      "learning_rate": 2.7133459756620646e-07,
+      "loss": 0.8035,
+      "num_input_tokens_seen": 148469745,
+      "step": 6963
+    },
+    {
+      "epoch": 0.8373714904106295,
+      "grad_norm": 1.9730274315620218,
+      "learning_rate": 2.709429679212969e-07,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 148489065,
+      "step": 6964
+    },
+    {
+      "epoch": 0.8374917333012686,
+      "grad_norm": 1.9750490967147278,
+      "learning_rate": 2.7055160057296424e-07,
+      "loss": 0.7488,
+      "num_input_tokens_seen": 148506025,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8376119761919076,
+      "grad_norm": 2.0732451084698806,
+      "learning_rate": 2.7016049558057896e-07,
+      "loss": 0.7181,
+      "num_input_tokens_seen": 148527705,
+      "step": 6966
+    },
+    {
+      "epoch": 0.8377322190825467,
+      "grad_norm": 1.9453842065679938,
+      "learning_rate": 2.6976965300347074e-07,
+      "loss": 0.7105,
+      "num_input_tokens_seen": 148550035,
+      "step": 6967
+    },
+    {
+      "epoch": 0.8378524619731859,
+      "grad_norm": 3.396684710335882,
+      "learning_rate": 2.693790729009309e-07,
+      "loss": 0.688,
+      "num_input_tokens_seen": 148571365,
+      "step": 6968
+    },
+    {
+      "epoch": 0.8379727048638249,
+      "grad_norm": 4.80072991963761,
+      "learning_rate": 2.6898875533220946e-07,
+      "loss": 0.8736,
+      "num_input_tokens_seen": 148590390,
+      "step": 6969
+    },
+    {
+      "epoch": 0.838092947754464,
+      "grad_norm": 1.9100846336874302,
+      "learning_rate": 2.685987003565171e-07,
+      "loss": 0.8083,
+      "num_input_tokens_seen": 148608150,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8382131906451031,
+      "grad_norm": 2.5960785425030175,
+      "learning_rate": 2.6820890803302566e-07,
+      "loss": 0.7613,
+      "num_input_tokens_seen": 148623395,
+      "step": 6971
+    },
+    {
+      "epoch": 0.8383334335357422,
+      "grad_norm": 2.146671274393012,
+      "learning_rate": 2.67819378420866e-07,
+      "loss": 0.8112,
+      "num_input_tokens_seen": 148641905,
+      "step": 6972
+    },
+    {
+      "epoch": 0.8384536764263812,
+      "grad_norm": 3.5919793867400522,
+      "learning_rate": 2.6743011157912933e-07,
+      "loss": 0.6808,
+      "num_input_tokens_seen": 148661345,
+      "step": 6973
+    },
+    {
+      "epoch": 0.8385739193170204,
+      "grad_norm": 2.3003579604372875,
+      "learning_rate": 2.6704110756686683e-07,
+      "loss": 0.6504,
+      "num_input_tokens_seen": 148681890,
+      "step": 6974
+    },
+    {
+      "epoch": 0.8386941622076595,
+      "grad_norm": 2.2560659493105275,
+      "learning_rate": 2.6665236644309085e-07,
+      "loss": 0.8339,
+      "num_input_tokens_seen": 148701920,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8388144050982985,
+      "grad_norm": 2.183155857114762,
+      "learning_rate": 2.662638882667727e-07,
+      "loss": 0.7898,
+      "num_input_tokens_seen": 148720580,
+      "step": 6976
+    },
+    {
+      "epoch": 0.8389346479889377,
+      "grad_norm": 2.1207740955438683,
+      "learning_rate": 2.658756730968443e-07,
+      "loss": 0.721,
+      "num_input_tokens_seen": 148738765,
+      "step": 6977
+    },
+    {
+      "epoch": 0.8390548908795767,
+      "grad_norm": 2.620582790305174,
+      "learning_rate": 2.654877209921975e-07,
+      "loss": 0.8787,
+      "num_input_tokens_seen": 148756020,
+      "step": 6978
+    },
+    {
+      "epoch": 0.8391751337702158,
+      "grad_norm": 2.7957844633839644,
+      "learning_rate": 2.651000320116843e-07,
+      "loss": 0.6283,
+      "num_input_tokens_seen": 148776625,
+      "step": 6979
+    },
+    {
+      "epoch": 0.839295376660855,
+      "grad_norm": 2.2994686652372653,
+      "learning_rate": 2.647126062141167e-07,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 148795420,
+      "step": 6980
+    },
+    {
+      "epoch": 0.839415619551494,
+      "grad_norm": 2.185893216462019,
+      "learning_rate": 2.643254436582674e-07,
+      "loss": 0.8266,
+      "num_input_tokens_seen": 148814630,
+      "step": 6981
+    },
+    {
+      "epoch": 0.8395358624421331,
+      "grad_norm": 2.1147213452862674,
+      "learning_rate": 2.6393854440286743e-07,
+      "loss": 0.8202,
+      "num_input_tokens_seen": 148833520,
+      "step": 6982
+    },
+    {
+      "epoch": 0.8396561053327722,
+      "grad_norm": 2.2342781966876424,
+      "learning_rate": 2.6355190850661045e-07,
+      "loss": 0.7008,
+      "num_input_tokens_seen": 148850075,
+      "step": 6983
+    },
+    {
+      "epoch": 0.8397763482234113,
+      "grad_norm": 2.4874857158277583,
+      "learning_rate": 2.631655360281486e-07,
+      "loss": 0.8566,
+      "num_input_tokens_seen": 148869470,
+      "step": 6984
+    },
+    {
+      "epoch": 0.8398965911140504,
+      "grad_norm": 2.7039312951341965,
+      "learning_rate": 2.6277942702609366e-07,
+      "loss": 0.6566,
+      "num_input_tokens_seen": 148888670,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8400168340046895,
+      "grad_norm": 3.1963195216116365,
+      "learning_rate": 2.6239358155901816e-07,
+      "loss": 0.8676,
+      "num_input_tokens_seen": 148906770,
+      "step": 6986
+    },
+    {
+      "epoch": 0.8401370768953286,
+      "grad_norm": 3.1957382170888673,
+      "learning_rate": 2.6200799968545516e-07,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 148926785,
+      "step": 6987
+    },
+    {
+      "epoch": 0.8402573197859676,
+      "grad_norm": 0.8141094580973,
+      "learning_rate": 2.616226814638969e-07,
+      "loss": 0.593,
+      "num_input_tokens_seen": 148991610,
+      "step": 6988
+    },
+    {
+      "epoch": 0.8403775626766068,
+      "grad_norm": 2.0061494102156794,
+      "learning_rate": 2.612376269527954e-07,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 149011035,
+      "step": 6989
+    },
+    {
+      "epoch": 0.8404978055672458,
+      "grad_norm": 24.261070551092935,
+      "learning_rate": 2.608528362105631e-07,
+      "loss": 0.6775,
+      "num_input_tokens_seen": 149030125,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8406180484578849,
+      "grad_norm": 2.46070628071618,
+      "learning_rate": 2.6046830929557327e-07,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 149049495,
+      "step": 6991
+    },
+    {
+      "epoch": 0.8407382913485241,
+      "grad_norm": 2.344915412832524,
+      "learning_rate": 2.6008404626615776e-07,
+      "loss": 0.8426,
+      "num_input_tokens_seen": 149067715,
+      "step": 6992
+    },
+    {
+      "epoch": 0.8408585342391631,
+      "grad_norm": 2.754917772900143,
+      "learning_rate": 2.597000471806092e-07,
+      "loss": 0.7293,
+      "num_input_tokens_seen": 149084000,
+      "step": 6993
+    },
+    {
+      "epoch": 0.8409787771298022,
+      "grad_norm": 2.5364283389798,
+      "learning_rate": 2.5931631209717976e-07,
+      "loss": 0.7272,
+      "num_input_tokens_seen": 149102585,
+      "step": 6994
+    },
+    {
+      "epoch": 0.8410990200204413,
+      "grad_norm": 2.0493255185944728,
+      "learning_rate": 2.5893284107408165e-07,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 149119675,
+      "step": 6995
+    },
+    {
+      "epoch": 0.8412192629110804,
+      "grad_norm": 2.2518907794419074,
+      "learning_rate": 2.5854963416948726e-07,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 149141660,
+      "step": 6996
+    },
+    {
+      "epoch": 0.8413395058017195,
+      "grad_norm": 1.7050513470355861,
+      "learning_rate": 2.581666914415286e-07,
+      "loss": 0.6865,
+      "num_input_tokens_seen": 149162560,
+      "step": 6997
+    },
+    {
+      "epoch": 0.8414597486923585,
+      "grad_norm": 0.9404238787506684,
+      "learning_rate": 2.5778401294829777e-07,
+      "loss": 0.73,
+      "num_input_tokens_seen": 149221020,
+      "step": 6998
+    },
+    {
+      "epoch": 0.8415799915829977,
+      "grad_norm": 2.7058926233499103,
+      "learning_rate": 2.574015987478473e-07,
+      "loss": 0.644,
+      "num_input_tokens_seen": 149238870,
+      "step": 6999
+    },
+    {
+      "epoch": 0.8417002344736367,
+      "grad_norm": 7.603321120843654,
+      "learning_rate": 2.570194488981887e-07,
+      "loss": 0.8555,
+      "num_input_tokens_seen": 149255135,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8418204773642758,
+      "grad_norm": 0.9231476941463659,
+      "learning_rate": 2.566375634572939e-07,
+      "loss": 0.6408,
+      "num_input_tokens_seen": 149315495,
+      "step": 7001
+    },
+    {
+      "epoch": 0.841940720254915,
+      "grad_norm": 2.2160526855752862,
+      "learning_rate": 2.562559424830943e-07,
+      "loss": 0.7522,
+      "num_input_tokens_seen": 149333175,
+      "step": 7002
+    },
+    {
+      "epoch": 0.842060963145554,
+      "grad_norm": 3.7530695055914203,
+      "learning_rate": 2.558745860334821e-07,
+      "loss": 0.6983,
+      "num_input_tokens_seen": 149350185,
+      "step": 7003
+    },
+    {
+      "epoch": 0.8421812060361931,
+      "grad_norm": 2.4648088736051403,
+      "learning_rate": 2.554934941663085e-07,
+      "loss": 0.8362,
+      "num_input_tokens_seen": 149367440,
+      "step": 7004
+    },
+    {
+      "epoch": 0.8423014489268322,
+      "grad_norm": 8.561751198289116,
+      "learning_rate": 2.5511266693938484e-07,
+      "loss": 0.7233,
+      "num_input_tokens_seen": 149385620,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8424216918174713,
+      "grad_norm": 1.5788227851948324,
+      "learning_rate": 2.5473210441048176e-07,
+      "loss": 0.7632,
+      "num_input_tokens_seen": 149406835,
+      "step": 7006
+    },
+    {
+      "epoch": 0.8425419347081103,
+      "grad_norm": 2.11376861019725,
+      "learning_rate": 2.5435180663733113e-07,
+      "loss": 0.7667,
+      "num_input_tokens_seen": 149426855,
+      "step": 7007
+    },
+    {
+      "epoch": 0.8426621775987495,
+      "grad_norm": 2.8285857509830254,
+      "learning_rate": 2.539717736776237e-07,
+      "loss": 0.7099,
+      "num_input_tokens_seen": 149442800,
+      "step": 7008
+    },
+    {
+      "epoch": 0.8427824204893886,
+      "grad_norm": 3.4524010749935616,
+      "learning_rate": 2.535920055890097e-07,
+      "loss": 0.7529,
+      "num_input_tokens_seen": 149463815,
+      "step": 7009
+    },
+    {
+      "epoch": 0.8429026633800276,
+      "grad_norm": 2.4412854795045345,
+      "learning_rate": 2.5321250242910006e-07,
+      "loss": 0.6423,
+      "num_input_tokens_seen": 149481450,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8430229062706668,
+      "grad_norm": 2.969574117544732,
+      "learning_rate": 2.5283326425546493e-07,
+      "loss": 0.8536,
+      "num_input_tokens_seen": 149500280,
+      "step": 7011
+    },
+    {
+      "epoch": 0.8431431491613058,
+      "grad_norm": 2.700931634059641,
+      "learning_rate": 2.5245429112563443e-07,
+      "loss": 0.6875,
+      "num_input_tokens_seen": 149520675,
+      "step": 7012
+    },
+    {
+      "epoch": 0.8432633920519449,
+      "grad_norm": 2.0078023355510495,
+      "learning_rate": 2.5207558309709865e-07,
+      "loss": 0.8188,
+      "num_input_tokens_seen": 149540130,
+      "step": 7013
+    },
+    {
+      "epoch": 0.8433836349425841,
+      "grad_norm": 0.7264068538051787,
+      "learning_rate": 2.516971402273065e-07,
+      "loss": 0.585,
+      "num_input_tokens_seen": 149605915,
+      "step": 7014
+    },
+    {
+      "epoch": 0.8435038778332231,
+      "grad_norm": 1.8956640360411938,
+      "learning_rate": 2.513189625736687e-07,
+      "loss": 0.6632,
+      "num_input_tokens_seen": 149622530,
+      "step": 7015
+    },
+    {
+      "epoch": 0.8436241207238622,
+      "grad_norm": 2.2422708545220043,
+      "learning_rate": 2.5094105019355385e-07,
+      "loss": 0.7011,
+      "num_input_tokens_seen": 149637885,
+      "step": 7016
+    },
+    {
+      "epoch": 0.8437443636145013,
+      "grad_norm": 3.5569036399022353,
+      "learning_rate": 2.5056340314429116e-07,
+      "loss": 0.749,
+      "num_input_tokens_seen": 149655070,
+      "step": 7017
+    },
+    {
+      "epoch": 0.8438646065051404,
+      "grad_norm": 2.364431627049494,
+      "learning_rate": 2.5018602148316857e-07,
+      "loss": 0.8016,
+      "num_input_tokens_seen": 149670825,
+      "step": 7018
+    },
+    {
+      "epoch": 0.8439848493957794,
+      "grad_norm": 1.7690261723177203,
+      "learning_rate": 2.498089052674359e-07,
+      "loss": 0.7918,
+      "num_input_tokens_seen": 149688520,
+      "step": 7019
+    },
+    {
+      "epoch": 0.8441050922864186,
+      "grad_norm": 2.140767648819049,
+      "learning_rate": 2.494320545543007e-07,
+      "loss": 0.7533,
+      "num_input_tokens_seen": 149707810,
+      "step": 7020
+    },
+    {
+      "epoch": 0.8442253351770577,
+      "grad_norm": 2.2221936890966374,
+      "learning_rate": 2.490554694009308e-07,
+      "loss": 0.6698,
+      "num_input_tokens_seen": 149728395,
+      "step": 7021
+    },
+    {
+      "epoch": 0.8443455780676967,
+      "grad_norm": 1.5737884257485262,
+      "learning_rate": 2.4867914986445426e-07,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 149750505,
+      "step": 7022
+    },
+    {
+      "epoch": 0.8444658209583359,
+      "grad_norm": 3.302121640607582,
+      "learning_rate": 2.483030960019581e-07,
+      "loss": 0.7002,
+      "num_input_tokens_seen": 149774155,
+      "step": 7023
+    },
+    {
+      "epoch": 0.8445860638489749,
+      "grad_norm": 0.7568110950777192,
+      "learning_rate": 2.4792730787048956e-07,
+      "loss": 0.5684,
+      "num_input_tokens_seen": 149827240,
+      "step": 7024
+    },
+    {
+      "epoch": 0.844706306739614,
+      "grad_norm": 0.8060351232851412,
+      "learning_rate": 2.475517855270552e-07,
+      "loss": 0.6647,
+      "num_input_tokens_seen": 149887040,
+      "step": 7025
+    },
+    {
+      "epoch": 0.8448265496302532,
+      "grad_norm": 1.896761020819705,
+      "learning_rate": 2.4717652902862143e-07,
+      "loss": 0.7253,
+      "num_input_tokens_seen": 149905735,
+      "step": 7026
+    },
+    {
+      "epoch": 0.8449467925208922,
+      "grad_norm": 1.8544717123263836,
+      "learning_rate": 2.4680153843211495e-07,
+      "loss": 0.813,
+      "num_input_tokens_seen": 149925385,
+      "step": 7027
+    },
+    {
+      "epoch": 0.8450670354115313,
+      "grad_norm": 3.0420278636111466,
+      "learning_rate": 2.464268137944212e-07,
+      "loss": 0.7144,
+      "num_input_tokens_seen": 149946400,
+      "step": 7028
+    },
+    {
+      "epoch": 0.8451872783021703,
+      "grad_norm": 1.9624355658659327,
+      "learning_rate": 2.460523551723854e-07,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 149964160,
+      "step": 7029
+    },
+    {
+      "epoch": 0.8453075211928095,
+      "grad_norm": 1.9029728300252038,
+      "learning_rate": 2.456781626228124e-07,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 149983385,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8454277640834486,
+      "grad_norm": 1.0699732595512892,
+      "learning_rate": 2.453042362024675e-07,
+      "loss": 0.7477,
+      "num_input_tokens_seen": 150036350,
+      "step": 7031
+    },
+    {
+      "epoch": 0.8455480069740876,
+      "grad_norm": 1.8955352264829122,
+      "learning_rate": 2.449305759680751e-07,
+      "loss": 0.728,
+      "num_input_tokens_seen": 150057395,
+      "step": 7032
+    },
+    {
+      "epoch": 0.8456682498647268,
+      "grad_norm": 1.8172100038432952,
+      "learning_rate": 2.445571819763188e-07,
+      "loss": 0.7454,
+      "num_input_tokens_seen": 150079415,
+      "step": 7033
+    },
+    {
+      "epoch": 0.8457884927553658,
+      "grad_norm": 2.699841256259373,
+      "learning_rate": 2.441840542838418e-07,
+      "loss": 0.5799,
+      "num_input_tokens_seen": 150099345,
+      "step": 7034
+    },
+    {
+      "epoch": 0.8459087356460049,
+      "grad_norm": 3.118410657928793,
+      "learning_rate": 2.4381119294724816e-07,
+      "loss": 0.7179,
+      "num_input_tokens_seen": 150116510,
+      "step": 7035
+    },
+    {
+      "epoch": 0.846028978536644,
+      "grad_norm": 3.316587881044612,
+      "learning_rate": 2.434385980231004e-07,
+      "loss": 0.546,
+      "num_input_tokens_seen": 150135070,
+      "step": 7036
+    },
+    {
+      "epoch": 0.8461492214272831,
+      "grad_norm": 1.8737851121679632,
+      "learning_rate": 2.4306626956792043e-07,
+      "loss": 0.6469,
+      "num_input_tokens_seen": 150159735,
+      "step": 7037
+    },
+    {
+      "epoch": 0.8462694643179222,
+      "grad_norm": 4.876373211076824,
+      "learning_rate": 2.426942076381906e-07,
+      "loss": 0.7547,
+      "num_input_tokens_seen": 150177500,
+      "step": 7038
+    },
+    {
+      "epoch": 0.8463897072085613,
+      "grad_norm": 4.263012352849394,
+      "learning_rate": 2.4232241229035223e-07,
+      "loss": 0.826,
+      "num_input_tokens_seen": 150194975,
+      "step": 7039
+    },
+    {
+      "epoch": 0.8465099500992004,
+      "grad_norm": 0.8406978610807823,
+      "learning_rate": 2.419508835808064e-07,
+      "loss": 0.5975,
+      "num_input_tokens_seen": 150251250,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8466301929898394,
+      "grad_norm": 17.58985235175416,
+      "learning_rate": 2.415796215659136e-07,
+      "loss": 0.6236,
+      "num_input_tokens_seen": 150267675,
+      "step": 7041
+    },
+    {
+      "epoch": 0.8467504358804786,
+      "grad_norm": 5.951262392195484,
+      "learning_rate": 2.412086263019939e-07,
+      "loss": 0.7728,
+      "num_input_tokens_seen": 150285420,
+      "step": 7042
+    },
+    {
+      "epoch": 0.8468706787711177,
+      "grad_norm": 1.8287172526399715,
+      "learning_rate": 2.408378978453276e-07,
+      "loss": 0.797,
+      "num_input_tokens_seen": 150305260,
+      "step": 7043
+    },
+    {
+      "epoch": 0.8469909216617567,
+      "grad_norm": 0.8153187237949509,
+      "learning_rate": 2.404674362521533e-07,
+      "loss": 0.6629,
+      "num_input_tokens_seen": 150363475,
+      "step": 7044
+    },
+    {
+      "epoch": 0.8471111645523959,
+      "grad_norm": 2.6547907029001907,
+      "learning_rate": 2.4009724157866997e-07,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 150380255,
+      "step": 7045
+    },
+    {
+      "epoch": 0.8472314074430349,
+      "grad_norm": 2.07571613263205,
+      "learning_rate": 2.3972731388103564e-07,
+      "loss": 0.7596,
+      "num_input_tokens_seen": 150398455,
+      "step": 7046
+    },
+    {
+      "epoch": 0.847351650333674,
+      "grad_norm": 0.8263367032748078,
+      "learning_rate": 2.393576532153683e-07,
+      "loss": 0.6476,
+      "num_input_tokens_seen": 150461960,
+      "step": 7047
+    },
+    {
+      "epoch": 0.8474718932243132,
+      "grad_norm": 0.9807950941046557,
+      "learning_rate": 2.389882596377453e-07,
+      "loss": 0.6267,
+      "num_input_tokens_seen": 150515945,
+      "step": 7048
+    },
+    {
+      "epoch": 0.8475921361149522,
+      "grad_norm": 1.9082715675593804,
+      "learning_rate": 2.386191332042031e-07,
+      "loss": 0.758,
+      "num_input_tokens_seen": 150537560,
+      "step": 7049
+    },
+    {
+      "epoch": 0.8477123790055913,
+      "grad_norm": 1.8911575917975172,
+      "learning_rate": 2.382502739707375e-07,
+      "loss": 0.7269,
+      "num_input_tokens_seen": 150557755,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8478326218962304,
+      "grad_norm": 2.1422948224332776,
+      "learning_rate": 2.3788168199330515e-07,
+      "loss": 0.6628,
+      "num_input_tokens_seen": 150579035,
+      "step": 7051
+    },
+    {
+      "epoch": 0.8479528647868695,
+      "grad_norm": 2.3746617085126958,
+      "learning_rate": 2.375133573278205e-07,
+      "loss": 0.7253,
+      "num_input_tokens_seen": 150600015,
+      "step": 7052
+    },
+    {
+      "epoch": 0.8480731076775085,
+      "grad_norm": 2.2291697052645327,
+      "learning_rate": 2.371453000301582e-07,
+      "loss": 0.7887,
+      "num_input_tokens_seen": 150618420,
+      "step": 7053
+    },
+    {
+      "epoch": 0.8481933505681477,
+      "grad_norm": 1.8937414140644757,
+      "learning_rate": 2.3677751015615222e-07,
+      "loss": 0.7347,
+      "num_input_tokens_seen": 150640215,
+      "step": 7054
+    },
+    {
+      "epoch": 0.8483135934587868,
+      "grad_norm": 1.9748290387350074,
+      "learning_rate": 2.3640998776159593e-07,
+      "loss": 0.8482,
+      "num_input_tokens_seen": 150657440,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8484338363494258,
+      "grad_norm": 1.741265989049003,
+      "learning_rate": 2.3604273290224253e-07,
+      "loss": 0.8049,
+      "num_input_tokens_seen": 150677875,
+      "step": 7056
+    },
+    {
+      "epoch": 0.848554079240065,
+      "grad_norm": 2.685245561921207,
+      "learning_rate": 2.356757456338039e-07,
+      "loss": 0.7406,
+      "num_input_tokens_seen": 150695080,
+      "step": 7057
+    },
+    {
+      "epoch": 0.848674322130704,
+      "grad_norm": 0.853456328368847,
+      "learning_rate": 2.3530902601195147e-07,
+      "loss": 0.6532,
+      "num_input_tokens_seen": 150763290,
+      "step": 7058
+    },
+    {
+      "epoch": 0.8487945650213431,
+      "grad_norm": 2.510092769480272,
+      "learning_rate": 2.34942574092317e-07,
+      "loss": 0.7783,
+      "num_input_tokens_seen": 150778260,
+      "step": 7059
+    },
+    {
+      "epoch": 0.8489148079119821,
+      "grad_norm": 4.460202404886829,
+      "learning_rate": 2.345763899304909e-07,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 150795970,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8490350508026213,
+      "grad_norm": 2.4355810019979374,
+      "learning_rate": 2.3421047358202252e-07,
+      "loss": 0.6425,
+      "num_input_tokens_seen": 150814540,
+      "step": 7061
+    },
+    {
+      "epoch": 0.8491552936932604,
+      "grad_norm": 3.0332326048770257,
+      "learning_rate": 2.33844825102421e-07,
+      "loss": 0.8206,
+      "num_input_tokens_seen": 150832120,
+      "step": 7062
+    },
+    {
+      "epoch": 0.8492755365838994,
+      "grad_norm": 3.188543223196614,
+      "learning_rate": 2.3347944454715575e-07,
+      "loss": 0.7682,
+      "num_input_tokens_seen": 150848230,
+      "step": 7063
+    },
+    {
+      "epoch": 0.8493957794745386,
+      "grad_norm": 2.3233737441554725,
+      "learning_rate": 2.331143319716542e-07,
+      "loss": 0.6634,
+      "num_input_tokens_seen": 150867480,
+      "step": 7064
+    },
+    {
+      "epoch": 0.8495160223651776,
+      "grad_norm": 4.462323509616469,
+      "learning_rate": 2.3274948743130363e-07,
+      "loss": 0.6467,
+      "num_input_tokens_seen": 150887035,
+      "step": 7065
+    },
+    {
+      "epoch": 0.8496362652558167,
+      "grad_norm": 1.6797241525802804,
+      "learning_rate": 2.3238491098145085e-07,
+      "loss": 0.7897,
+      "num_input_tokens_seen": 150906285,
+      "step": 7066
+    },
+    {
+      "epoch": 0.8497565081464559,
+      "grad_norm": 2.4242288837038406,
+      "learning_rate": 2.3202060267740141e-07,
+      "loss": 0.7303,
+      "num_input_tokens_seen": 150923530,
+      "step": 7067
+    },
+    {
+      "epoch": 0.8498767510370949,
+      "grad_norm": 2.4316967109254835,
+      "learning_rate": 2.316565625744209e-07,
+      "loss": 0.7703,
+      "num_input_tokens_seen": 150941770,
+      "step": 7068
+    },
+    {
+      "epoch": 0.849996993927734,
+      "grad_norm": 4.575553863754309,
+      "learning_rate": 2.31292790727734e-07,
+      "loss": 0.8961,
+      "num_input_tokens_seen": 150959055,
+      "step": 7069
+    },
+    {
+      "epoch": 0.8501172368183731,
+      "grad_norm": 2.842109170699587,
+      "learning_rate": 2.3092928719252392e-07,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 150977175,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8502374797090122,
+      "grad_norm": 2.0921256149427565,
+      "learning_rate": 2.3056605202393475e-07,
+      "loss": 0.7757,
+      "num_input_tokens_seen": 150994455,
+      "step": 7071
+    },
+    {
+      "epoch": 0.8503577225996513,
+      "grad_norm": 2.4569211708733656,
+      "learning_rate": 2.3020308527706888e-07,
+      "loss": 0.6662,
+      "num_input_tokens_seen": 151013590,
+      "step": 7072
+    },
+    {
+      "epoch": 0.8504779654902904,
+      "grad_norm": 1.872827926695392,
+      "learning_rate": 2.298403870069876e-07,
+      "loss": 0.8809,
+      "num_input_tokens_seen": 151032620,
+      "step": 7073
+    },
+    {
+      "epoch": 0.8505982083809295,
+      "grad_norm": 1.9564084635512284,
+      "learning_rate": 2.2947795726871177e-07,
+      "loss": 0.7853,
+      "num_input_tokens_seen": 151053365,
+      "step": 7074
+    },
+    {
+      "epoch": 0.8507184512715685,
+      "grad_norm": 1.7907204789757285,
+      "learning_rate": 2.2911579611722253e-07,
+      "loss": 0.848,
+      "num_input_tokens_seen": 151072230,
+      "step": 7075
+    },
+    {
+      "epoch": 0.8508386941622077,
+      "grad_norm": 1.9292736461567053,
+      "learning_rate": 2.2875390360745905e-07,
+      "loss": 0.8688,
+      "num_input_tokens_seen": 151091355,
+      "step": 7076
+    },
+    {
+      "epoch": 0.8509589370528468,
+      "grad_norm": 3.558941402147183,
+      "learning_rate": 2.2839227979432008e-07,
+      "loss": 0.7716,
+      "num_input_tokens_seen": 151108725,
+      "step": 7077
+    },
+    {
+      "epoch": 0.8510791799434858,
+      "grad_norm": 2.0825181234485863,
+      "learning_rate": 2.2803092473266328e-07,
+      "loss": 0.8398,
+      "num_input_tokens_seen": 151125970,
+      "step": 7078
+    },
+    {
+      "epoch": 0.851199422834125,
+      "grad_norm": 2.678802518549154,
+      "learning_rate": 2.2766983847730682e-07,
+      "loss": 0.8593,
+      "num_input_tokens_seen": 151145360,
+      "step": 7079
+    },
+    {
+      "epoch": 0.851319665724764,
+      "grad_norm": 2.1370111705562023,
+      "learning_rate": 2.2730902108302663e-07,
+      "loss": 0.6653,
+      "num_input_tokens_seen": 151161995,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8514399086154031,
+      "grad_norm": 1.6101808947683822,
+      "learning_rate": 2.269484726045583e-07,
+      "loss": 0.6863,
+      "num_input_tokens_seen": 151180630,
+      "step": 7081
+    },
+    {
+      "epoch": 0.8515601515060423,
+      "grad_norm": 1.932912260384305,
+      "learning_rate": 2.2658819309659715e-07,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 151200550,
+      "step": 7082
+    },
+    {
+      "epoch": 0.8516803943966813,
+      "grad_norm": 2.139634794704658,
+      "learning_rate": 2.2622818261379706e-07,
+      "loss": 0.8365,
+      "num_input_tokens_seen": 151217290,
+      "step": 7083
+    },
+    {
+      "epoch": 0.8518006372873204,
+      "grad_norm": 4.433274764354832,
+      "learning_rate": 2.2586844121077142e-07,
+      "loss": 0.7463,
+      "num_input_tokens_seen": 151235520,
+      "step": 7084
+    },
+    {
+      "epoch": 0.8519208801779595,
+      "grad_norm": 2.2363025787205655,
+      "learning_rate": 2.255089689420926e-07,
+      "loss": 0.7127,
+      "num_input_tokens_seen": 151254755,
+      "step": 7085
+    },
+    {
+      "epoch": 0.8520411230685986,
+      "grad_norm": 0.6980516998447036,
+      "learning_rate": 2.2514976586229184e-07,
+      "loss": 0.59,
+      "num_input_tokens_seen": 151322420,
+      "step": 7086
+    },
+    {
+      "epoch": 0.8521613659592376,
+      "grad_norm": 0.9274846236901506,
+      "learning_rate": 2.247908320258609e-07,
+      "loss": 0.6288,
+      "num_input_tokens_seen": 151382230,
+      "step": 7087
+    },
+    {
+      "epoch": 0.8522816088498768,
+      "grad_norm": 2.5354770140202394,
+      "learning_rate": 2.2443216748724914e-07,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 151402660,
+      "step": 7088
+    },
+    {
+      "epoch": 0.8524018517405159,
+      "grad_norm": 2.3561222562200204,
+      "learning_rate": 2.2407377230086588e-07,
+      "loss": 0.7432,
+      "num_input_tokens_seen": 151424735,
+      "step": 7089
+    },
+    {
+      "epoch": 0.8525220946311549,
+      "grad_norm": 2.5433032758430936,
+      "learning_rate": 2.23715646521079e-07,
+      "loss": 0.8287,
+      "num_input_tokens_seen": 151441975,
+      "step": 7090
+    },
+    {
+      "epoch": 0.852642337521794,
+      "grad_norm": 2.3356782846759097,
+      "learning_rate": 2.233577902022168e-07,
+      "loss": 0.831,
+      "num_input_tokens_seen": 151458315,
+      "step": 7091
+    },
+    {
+      "epoch": 0.8527625804124331,
+      "grad_norm": 0.8857943101204383,
+      "learning_rate": 2.2300020339856497e-07,
+      "loss": 0.6239,
+      "num_input_tokens_seen": 151520720,
+      "step": 7092
+    },
+    {
+      "epoch": 0.8528828233030722,
+      "grad_norm": 3.4988663917784457,
+      "learning_rate": 2.2264288616436966e-07,
+      "loss": 0.7704,
+      "num_input_tokens_seen": 151540695,
+      "step": 7093
+    },
+    {
+      "epoch": 0.8530030661937112,
+      "grad_norm": 2.1399794375545587,
+      "learning_rate": 2.2228583855383464e-07,
+      "loss": 0.7326,
+      "num_input_tokens_seen": 151557215,
+      "step": 7094
+    },
+    {
+      "epoch": 0.8531233090843504,
+      "grad_norm": 1.8371527139517003,
+      "learning_rate": 2.2192906062112527e-07,
+      "loss": 0.6724,
+      "num_input_tokens_seen": 151576810,
+      "step": 7095
+    },
+    {
+      "epoch": 0.8532435519749895,
+      "grad_norm": 1.553842364359297,
+      "learning_rate": 2.2157255242036333e-07,
+      "loss": 0.7,
+      "num_input_tokens_seen": 151600195,
+      "step": 7096
+    },
+    {
+      "epoch": 0.8533637948656285,
+      "grad_norm": 2.1987024811894735,
+      "learning_rate": 2.2121631400563178e-07,
+      "loss": 0.7394,
+      "num_input_tokens_seen": 151619745,
+      "step": 7097
+    },
+    {
+      "epoch": 0.8534840377562677,
+      "grad_norm": 0.8445379874529323,
+      "learning_rate": 2.208603454309701e-07,
+      "loss": 0.6057,
+      "num_input_tokens_seen": 151677555,
+      "step": 7098
+    },
+    {
+      "epoch": 0.8536042806469067,
+      "grad_norm": 2.920393185815172,
+      "learning_rate": 2.2050464675037994e-07,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 151695900,
+      "step": 7099
+    },
+    {
+      "epoch": 0.8537245235375458,
+      "grad_norm": 2.7386850105403147,
+      "learning_rate": 2.2014921801782016e-07,
+      "loss": 0.7269,
+      "num_input_tokens_seen": 151715110,
+      "step": 7100
+    },
+    {
+      "epoch": 0.853844766428185,
+      "grad_norm": 2.1101372242820666,
+      "learning_rate": 2.1979405928720872e-07,
+      "loss": 0.732,
+      "num_input_tokens_seen": 151734485,
+      "step": 7101
+    },
+    {
+      "epoch": 0.853965009318824,
+      "grad_norm": 2.6447096604926554,
+      "learning_rate": 2.1943917061242257e-07,
+      "loss": 0.7855,
+      "num_input_tokens_seen": 151754060,
+      "step": 7102
+    },
+    {
+      "epoch": 0.8540852522094631,
+      "grad_norm": 2.191214286158684,
+      "learning_rate": 2.1908455204729903e-07,
+      "loss": 0.6596,
+      "num_input_tokens_seen": 151772930,
+      "step": 7103
+    },
+    {
+      "epoch": 0.8542054951001022,
+      "grad_norm": 2.096211094482049,
+      "learning_rate": 2.187302036456331e-07,
+      "loss": 0.7732,
+      "num_input_tokens_seen": 151791715,
+      "step": 7104
+    },
+    {
+      "epoch": 0.8543257379907413,
+      "grad_norm": 3.504360039067773,
+      "learning_rate": 2.183761254611789e-07,
+      "loss": 0.7576,
+      "num_input_tokens_seen": 151811760,
+      "step": 7105
+    },
+    {
+      "epoch": 0.8544459808813804,
+      "grad_norm": 2.0663234275142783,
+      "learning_rate": 2.1802231754764945e-07,
+      "loss": 0.7039,
+      "num_input_tokens_seen": 151836920,
+      "step": 7106
+    },
+    {
+      "epoch": 0.8545662237720195,
+      "grad_norm": 2.0362714465032403,
+      "learning_rate": 2.17668779958718e-07,
+      "loss": 0.7565,
+      "num_input_tokens_seen": 151859220,
+      "step": 7107
+    },
+    {
+      "epoch": 0.8546864666626586,
+      "grad_norm": 2.362050105227234,
+      "learning_rate": 2.1731551274801553e-07,
+      "loss": 0.7973,
+      "num_input_tokens_seen": 151875380,
+      "step": 7108
+    },
+    {
+      "epoch": 0.8548067095532976,
+      "grad_norm": 2.0477178248243044,
+      "learning_rate": 2.169625159691324e-07,
+      "loss": 0.604,
+      "num_input_tokens_seen": 151894975,
+      "step": 7109
+    },
+    {
+      "epoch": 0.8549269524439368,
+      "grad_norm": 2.8749073584890485,
+      "learning_rate": 2.1660978967561784e-07,
+      "loss": 0.7452,
+      "num_input_tokens_seen": 151914030,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8550471953345758,
+      "grad_norm": 2.6451303424907833,
+      "learning_rate": 2.1625733392098035e-07,
+      "loss": 0.7835,
+      "num_input_tokens_seen": 151929360,
+      "step": 7111
+    },
+    {
+      "epoch": 0.8551674382252149,
+      "grad_norm": 2.225476234905436,
+      "learning_rate": 2.1590514875868692e-07,
+      "loss": 0.7907,
+      "num_input_tokens_seen": 151949210,
+      "step": 7112
+    },
+    {
+      "epoch": 0.8552876811158541,
+      "grad_norm": 2.924170868868411,
+      "learning_rate": 2.155532342421642e-07,
+      "loss": 0.7225,
+      "num_input_tokens_seen": 151966930,
+      "step": 7113
+    },
+    {
+      "epoch": 0.8554079240064931,
+      "grad_norm": 2.0468407863242066,
+      "learning_rate": 2.1520159042479636e-07,
+      "loss": 0.7794,
+      "num_input_tokens_seen": 151984940,
+      "step": 7114
+    },
+    {
+      "epoch": 0.8555281668971322,
+      "grad_norm": 2.6096756245830206,
+      "learning_rate": 2.148502173599287e-07,
+      "loss": 0.7035,
+      "num_input_tokens_seen": 152002800,
+      "step": 7115
+    },
+    {
+      "epoch": 0.8556484097877713,
+      "grad_norm": 4.887049003488496,
+      "learning_rate": 2.1449911510086372e-07,
+      "loss": 0.6491,
+      "num_input_tokens_seen": 152021990,
+      "step": 7116
+    },
+    {
+      "epoch": 0.8557686526784104,
+      "grad_norm": 2.0839366022197363,
+      "learning_rate": 2.1414828370086324e-07,
+      "loss": 0.7665,
+      "num_input_tokens_seen": 152042250,
+      "step": 7117
+    },
+    {
+      "epoch": 0.8558888955690495,
+      "grad_norm": 1.831528610812624,
+      "learning_rate": 2.1379772321314782e-07,
+      "loss": 0.7146,
+      "num_input_tokens_seen": 152060015,
+      "step": 7118
+    },
+    {
+      "epoch": 0.8560091384596886,
+      "grad_norm": 2.3503315951839143,
+      "learning_rate": 2.1344743369089802e-07,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 152075515,
+      "step": 7119
+    },
+    {
+      "epoch": 0.8561293813503277,
+      "grad_norm": 2.050192104853889,
+      "learning_rate": 2.130974151872522e-07,
+      "loss": 0.8123,
+      "num_input_tokens_seen": 152095570,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8562496242409667,
+      "grad_norm": 2.102795715196339,
+      "learning_rate": 2.1274766775530773e-07,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 152115155,
+      "step": 7121
+    },
+    {
+      "epoch": 0.8563698671316058,
+      "grad_norm": 1.9510248557037422,
+      "learning_rate": 2.1239819144812056e-07,
+      "loss": 0.7877,
+      "num_input_tokens_seen": 152129335,
+      "step": 7122
+    },
+    {
+      "epoch": 0.856490110022245,
+      "grad_norm": 1.9979452963810218,
+      "learning_rate": 2.120489863187067e-07,
+      "loss": 0.6888,
+      "num_input_tokens_seen": 152153945,
+      "step": 7123
+    },
+    {
+      "epoch": 0.856610352912884,
+      "grad_norm": 1.9357833165382563,
+      "learning_rate": 2.1170005242004006e-07,
+      "loss": 0.7623,
+      "num_input_tokens_seen": 152175015,
+      "step": 7124
+    },
+    {
+      "epoch": 0.8567305958035231,
+      "grad_norm": 2.386084644553693,
+      "learning_rate": 2.1135138980505384e-07,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 152195405,
+      "step": 7125
+    },
+    {
+      "epoch": 0.8568508386941622,
+      "grad_norm": 1.8508099347400397,
+      "learning_rate": 2.110029985266395e-07,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 152214830,
+      "step": 7126
+    },
+    {
+      "epoch": 0.8569710815848013,
+      "grad_norm": 1.747903818307868,
+      "learning_rate": 2.1065487863764787e-07,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 152232895,
+      "step": 7127
+    },
+    {
+      "epoch": 0.8570913244754403,
+      "grad_norm": 1.6592736947031055,
+      "learning_rate": 2.1030703019088846e-07,
+      "loss": 0.8503,
+      "num_input_tokens_seen": 152253245,
+      "step": 7128
+    },
+    {
+      "epoch": 0.8572115673660795,
+      "grad_norm": 2.2555919744515083,
+      "learning_rate": 2.0995945323912956e-07,
+      "loss": 0.7051,
+      "num_input_tokens_seen": 152271650,
+      "step": 7129
+    },
+    {
+      "epoch": 0.8573318102567186,
+      "grad_norm": 1.6952760273826992,
+      "learning_rate": 2.0961214783509806e-07,
+      "loss": 0.78,
+      "num_input_tokens_seen": 152294250,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8574520531473576,
+      "grad_norm": 3.3150091970414177,
+      "learning_rate": 2.0926511403148051e-07,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 152312935,
+      "step": 7131
+    },
+    {
+      "epoch": 0.8575722960379968,
+      "grad_norm": 2.320398321897453,
+      "learning_rate": 2.0891835188092143e-07,
+      "loss": 0.7567,
+      "num_input_tokens_seen": 152329655,
+      "step": 7132
+    },
+    {
+      "epoch": 0.8576925389286358,
+      "grad_norm": 1.908287808946306,
+      "learning_rate": 2.0857186143602434e-07,
+      "loss": 0.8096,
+      "num_input_tokens_seen": 152348020,
+      "step": 7133
+    },
+    {
+      "epoch": 0.8578127818192749,
+      "grad_norm": 2.0352865886651688,
+      "learning_rate": 2.0822564274935094e-07,
+      "loss": 0.6746,
+      "num_input_tokens_seen": 152367165,
+      "step": 7134
+    },
+    {
+      "epoch": 0.8579330247099141,
+      "grad_norm": 2.9696702151476138,
+      "learning_rate": 2.0787969587342346e-07,
+      "loss": 0.6672,
+      "num_input_tokens_seen": 152389605,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8580532676005531,
+      "grad_norm": 2.166406478957432,
+      "learning_rate": 2.0753402086072124e-07,
+      "loss": 0.739,
+      "num_input_tokens_seen": 152407955,
+      "step": 7136
+    },
+    {
+      "epoch": 0.8581735104911922,
+      "grad_norm": 2.709845163384798,
+      "learning_rate": 2.071886177636828e-07,
+      "loss": 0.7557,
+      "num_input_tokens_seen": 152424460,
+      "step": 7137
+    },
+    {
+      "epoch": 0.8582937533818313,
+      "grad_norm": 2.033111600699077,
+      "learning_rate": 2.068434866347053e-07,
+      "loss": 0.8177,
+      "num_input_tokens_seen": 152444360,
+      "step": 7138
+    },
+    {
+      "epoch": 0.8584139962724704,
+      "grad_norm": 2.0407494257726317,
+      "learning_rate": 2.0649862752614555e-07,
+      "loss": 0.6129,
+      "num_input_tokens_seen": 152462790,
+      "step": 7139
+    },
+    {
+      "epoch": 0.8585342391631094,
+      "grad_norm": 0.8049465740917714,
+      "learning_rate": 2.0615404049031838e-07,
+      "loss": 0.5945,
+      "num_input_tokens_seen": 152519480,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8586544820537486,
+      "grad_norm": 3.2630454824190163,
+      "learning_rate": 2.058097255794966e-07,
+      "loss": 0.7785,
+      "num_input_tokens_seen": 152534290,
+      "step": 7141
+    },
+    {
+      "epoch": 0.8587747249443877,
+      "grad_norm": 0.8115580668202989,
+      "learning_rate": 2.054656828459125e-07,
+      "loss": 0.5502,
+      "num_input_tokens_seen": 152598120,
+      "step": 7142
+    },
+    {
+      "epoch": 0.8588949678350267,
+      "grad_norm": 4.92193482901055,
+      "learning_rate": 2.051219123417578e-07,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 152617900,
+      "step": 7143
+    },
+    {
+      "epoch": 0.8590152107256659,
+      "grad_norm": 2.9717207988956065,
+      "learning_rate": 2.0477841411918196e-07,
+      "loss": 0.5957,
+      "num_input_tokens_seen": 152637145,
+      "step": 7144
+    },
+    {
+      "epoch": 0.859135453616305,
+      "grad_norm": 2.216565312955087,
+      "learning_rate": 2.0443518823029326e-07,
+      "loss": 0.7448,
+      "num_input_tokens_seen": 152657405,
+      "step": 7145
+    },
+    {
+      "epoch": 0.859255696506944,
+      "grad_norm": 2.791394646109027,
+      "learning_rate": 2.0409223472715854e-07,
+      "loss": 0.7626,
+      "num_input_tokens_seen": 152674270,
+      "step": 7146
+    },
+    {
+      "epoch": 0.8593759393975832,
+      "grad_norm": 1.9852725640866236,
+      "learning_rate": 2.0374955366180434e-07,
+      "loss": 0.7413,
+      "num_input_tokens_seen": 152691630,
+      "step": 7147
+    },
+    {
+      "epoch": 0.8594961822882222,
+      "grad_norm": 1.8941941624099565,
+      "learning_rate": 2.034071450862147e-07,
+      "loss": 0.7219,
+      "num_input_tokens_seen": 152708820,
+      "step": 7148
+    },
+    {
+      "epoch": 0.8596164251788613,
+      "grad_norm": 2.006034916351908,
+      "learning_rate": 2.030650090523327e-07,
+      "loss": 0.7632,
+      "num_input_tokens_seen": 152727730,
+      "step": 7149
+    },
+    {
+      "epoch": 0.8597366680695004,
+      "grad_norm": 2.1913441206873965,
+      "learning_rate": 2.027231456120595e-07,
+      "loss": 0.5931,
+      "num_input_tokens_seen": 152747845,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8598569109601395,
+      "grad_norm": 1.8126852530663988,
+      "learning_rate": 2.023815548172567e-07,
+      "loss": 0.7231,
+      "num_input_tokens_seen": 152767635,
+      "step": 7151
+    },
+    {
+      "epoch": 0.8599771538507786,
+      "grad_norm": 2.3543047215394055,
+      "learning_rate": 2.0204023671974267e-07,
+      "loss": 0.6587,
+      "num_input_tokens_seen": 152786740,
+      "step": 7152
+    },
+    {
+      "epoch": 0.8600973967414177,
+      "grad_norm": 2.266877085644117,
+      "learning_rate": 2.0169919137129532e-07,
+      "loss": 0.8086,
+      "num_input_tokens_seen": 152804900,
+      "step": 7153
+    },
+    {
+      "epoch": 0.8602176396320568,
+      "grad_norm": 2.24249984969273,
+      "learning_rate": 2.013584188236508e-07,
+      "loss": 0.6874,
+      "num_input_tokens_seen": 152822525,
+      "step": 7154
+    },
+    {
+      "epoch": 0.8603378825226958,
+      "grad_norm": 2.1326073804878236,
+      "learning_rate": 2.0101791912850396e-07,
+      "loss": 0.7917,
+      "num_input_tokens_seen": 152841785,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8604581254133349,
+      "grad_norm": 1.933624357412785,
+      "learning_rate": 2.0067769233750842e-07,
+      "loss": 0.6266,
+      "num_input_tokens_seen": 152863160,
+      "step": 7156
+    },
+    {
+      "epoch": 0.860578368303974,
+      "grad_norm": 2.383261236466245,
+      "learning_rate": 2.003377385022764e-07,
+      "loss": 0.6947,
+      "num_input_tokens_seen": 152881705,
+      "step": 7157
+    },
+    {
+      "epoch": 0.8606986111946131,
+      "grad_norm": 2.2464755370106944,
+      "learning_rate": 1.9999805767437826e-07,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 152900315,
+      "step": 7158
+    },
+    {
+      "epoch": 0.8608188540852522,
+      "grad_norm": 1.9152654052759959,
+      "learning_rate": 1.9965864990534386e-07,
+      "loss": 0.7124,
+      "num_input_tokens_seen": 152920560,
+      "step": 7159
+    },
+    {
+      "epoch": 0.8609390969758913,
+      "grad_norm": 2.0099014811244,
+      "learning_rate": 1.9931951524666092e-07,
+      "loss": 0.7726,
+      "num_input_tokens_seen": 152941370,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8610593398665304,
+      "grad_norm": 1.730678366576472,
+      "learning_rate": 1.989806537497758e-07,
+      "loss": 0.8021,
+      "num_input_tokens_seen": 152961295,
+      "step": 7161
+    },
+    {
+      "epoch": 0.8611795827571694,
+      "grad_norm": 2.3178195240729074,
+      "learning_rate": 1.9864206546609297e-07,
+      "loss": 0.7217,
+      "num_input_tokens_seen": 152979855,
+      "step": 7162
+    },
+    {
+      "epoch": 0.8612998256478086,
+      "grad_norm": 2.0057397641487835,
+      "learning_rate": 1.983037504469771e-07,
+      "loss": 0.8386,
+      "num_input_tokens_seen": 152998285,
+      "step": 7163
+    },
+    {
+      "epoch": 0.8614200685384477,
+      "grad_norm": 2.4475923873306864,
+      "learning_rate": 1.9796570874374984e-07,
+      "loss": 0.6637,
+      "num_input_tokens_seen": 153018110,
+      "step": 7164
+    },
+    {
+      "epoch": 0.8615403114290867,
+      "grad_norm": 2.3754437351453275,
+      "learning_rate": 1.976279404076917e-07,
+      "loss": 0.7671,
+      "num_input_tokens_seen": 153037230,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8616605543197259,
+      "grad_norm": 3.32312976187681,
+      "learning_rate": 1.9729044549004148e-07,
+      "loss": 0.7548,
+      "num_input_tokens_seen": 153058335,
+      "step": 7166
+    },
+    {
+      "epoch": 0.8617807972103649,
+      "grad_norm": 2.1710292128460456,
+      "learning_rate": 1.9695322404199798e-07,
+      "loss": 0.6986,
+      "num_input_tokens_seen": 153080100,
+      "step": 7167
+    },
+    {
+      "epoch": 0.861901040101004,
+      "grad_norm": 2.0820525311060445,
+      "learning_rate": 1.9661627611471654e-07,
+      "loss": 0.8197,
+      "num_input_tokens_seen": 153099615,
+      "step": 7168
+    },
+    {
+      "epoch": 0.8620212829916432,
+      "grad_norm": 2.05609631521287,
+      "learning_rate": 1.9627960175931246e-07,
+      "loss": 0.7029,
+      "num_input_tokens_seen": 153124035,
+      "step": 7169
+    },
+    {
+      "epoch": 0.8621415258822822,
+      "grad_norm": 2.926873636571819,
+      "learning_rate": 1.9594320102685847e-07,
+      "loss": 0.7405,
+      "num_input_tokens_seen": 153143025,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8622617687729213,
+      "grad_norm": 2.5005060277344726,
+      "learning_rate": 1.956070739683864e-07,
+      "loss": 0.6348,
+      "num_input_tokens_seen": 153162080,
+      "step": 7171
+    },
+    {
+      "epoch": 0.8623820116635604,
+      "grad_norm": 1.819274772706039,
+      "learning_rate": 1.9527122063488678e-07,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 153182915,
+      "step": 7172
+    },
+    {
+      "epoch": 0.8625022545541995,
+      "grad_norm": 1.8947983161705129,
+      "learning_rate": 1.9493564107730797e-07,
+      "loss": 0.7972,
+      "num_input_tokens_seen": 153202635,
+      "step": 7173
+    },
+    {
+      "epoch": 0.8626224974448385,
+      "grad_norm": 2.3719586764524063,
+      "learning_rate": 1.9460033534655684e-07,
+      "loss": 0.6112,
+      "num_input_tokens_seen": 153221715,
+      "step": 7174
+    },
+    {
+      "epoch": 0.8627427403354777,
+      "grad_norm": 3.153760040971219,
+      "learning_rate": 1.9426530349349978e-07,
+      "loss": 0.8304,
+      "num_input_tokens_seen": 153241885,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8628629832261168,
+      "grad_norm": 2.0797077003325457,
+      "learning_rate": 1.9393054556896038e-07,
+      "loss": 0.6505,
+      "num_input_tokens_seen": 153259305,
+      "step": 7176
+    },
+    {
+      "epoch": 0.8629832261167558,
+      "grad_norm": 2.473491592330941,
+      "learning_rate": 1.9359606162372133e-07,
+      "loss": 0.6877,
+      "num_input_tokens_seen": 153280630,
+      "step": 7177
+    },
+    {
+      "epoch": 0.863103469007395,
+      "grad_norm": 2.0471905583743077,
+      "learning_rate": 1.9326185170852293e-07,
+      "loss": 0.7069,
+      "num_input_tokens_seen": 153299315,
+      "step": 7178
+    },
+    {
+      "epoch": 0.863223711898034,
+      "grad_norm": 2.7662760660244743,
+      "learning_rate": 1.9292791587406553e-07,
+      "loss": 0.712,
+      "num_input_tokens_seen": 153317895,
+      "step": 7179
+    },
+    {
+      "epoch": 0.8633439547886731,
+      "grad_norm": 2.234619315226899,
+      "learning_rate": 1.9259425417100661e-07,
+      "loss": 0.8585,
+      "num_input_tokens_seen": 153333730,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8634641976793123,
+      "grad_norm": 3.7684167119657914,
+      "learning_rate": 1.9226086664996234e-07,
+      "loss": 0.7386,
+      "num_input_tokens_seen": 153351695,
+      "step": 7181
+    },
+    {
+      "epoch": 0.8635844405699513,
+      "grad_norm": 2.3184627211060347,
+      "learning_rate": 1.9192775336150667e-07,
+      "loss": 0.7411,
+      "num_input_tokens_seen": 153371715,
+      "step": 7182
+    },
+    {
+      "epoch": 0.8637046834605904,
+      "grad_norm": 0.7894201711477891,
+      "learning_rate": 1.9159491435617415e-07,
+      "loss": 0.5742,
+      "num_input_tokens_seen": 153426110,
+      "step": 7183
+    },
+    {
+      "epoch": 0.8638249263512295,
+      "grad_norm": 1.9464572640896318,
+      "learning_rate": 1.9126234968445498e-07,
+      "loss": 0.7678,
+      "num_input_tokens_seen": 153445520,
+      "step": 7184
+    },
+    {
+      "epoch": 0.8639451692418686,
+      "grad_norm": 2.533004740365005,
+      "learning_rate": 1.9093005939679907e-07,
+      "loss": 0.6631,
+      "num_input_tokens_seen": 153467195,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8640654121325076,
+      "grad_norm": 2.101528304251122,
+      "learning_rate": 1.9059804354361452e-07,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 153484690,
+      "step": 7186
+    },
+    {
+      "epoch": 0.8641856550231467,
+      "grad_norm": 2.021698217452282,
+      "learning_rate": 1.902663021752684e-07,
+      "loss": 0.7045,
+      "num_input_tokens_seen": 153505840,
+      "step": 7187
+    },
+    {
+      "epoch": 0.8643058979137859,
+      "grad_norm": 2.5000420485830075,
+      "learning_rate": 1.8993483534208556e-07,
+      "loss": 0.811,
+      "num_input_tokens_seen": 153524470,
+      "step": 7188
+    },
+    {
+      "epoch": 0.8644261408044249,
+      "grad_norm": 2.6667501661910396,
+      "learning_rate": 1.8960364309434884e-07,
+      "loss": 0.7369,
+      "num_input_tokens_seen": 153541685,
+      "step": 7189
+    },
+    {
+      "epoch": 0.864546383695064,
+      "grad_norm": 1.8983065177083056,
+      "learning_rate": 1.8927272548229967e-07,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 153561095,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8646666265857031,
+      "grad_norm": 1.7037925639123406,
+      "learning_rate": 1.8894208255613876e-07,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 153580130,
+      "step": 7191
+    },
+    {
+      "epoch": 0.8647868694763422,
+      "grad_norm": 4.952810886922987,
+      "learning_rate": 1.8861171436602397e-07,
+      "loss": 0.7662,
+      "num_input_tokens_seen": 153596965,
+      "step": 7192
+    },
+    {
+      "epoch": 0.8649071123669813,
+      "grad_norm": 2.6416261396699854,
+      "learning_rate": 1.882816209620719e-07,
+      "loss": 0.8028,
+      "num_input_tokens_seen": 153613395,
+      "step": 7193
+    },
+    {
+      "epoch": 0.8650273552576204,
+      "grad_norm": 1.861466738641778,
+      "learning_rate": 1.8795180239435693e-07,
+      "loss": 0.7631,
+      "num_input_tokens_seen": 153631970,
+      "step": 7194
+    },
+    {
+      "epoch": 0.8651475981482595,
+      "grad_norm": 2.9463203968605125,
+      "learning_rate": 1.8762225871291348e-07,
+      "loss": 0.7543,
+      "num_input_tokens_seen": 153647565,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8652678410388985,
+      "grad_norm": 1.8297474270851621,
+      "learning_rate": 1.8729298996773201e-07,
+      "loss": 0.7981,
+      "num_input_tokens_seen": 153666035,
+      "step": 7196
+    },
+    {
+      "epoch": 0.8653880839295377,
+      "grad_norm": 0.8650190380246979,
+      "learning_rate": 1.8696399620876301e-07,
+      "loss": 0.6366,
+      "num_input_tokens_seen": 153722785,
+      "step": 7197
+    },
+    {
+      "epoch": 0.8655083268201768,
+      "grad_norm": 2.7651777171582173,
+      "learning_rate": 1.866352774859141e-07,
+      "loss": 0.7877,
+      "num_input_tokens_seen": 153737730,
+      "step": 7198
+    },
+    {
+      "epoch": 0.8656285697108158,
+      "grad_norm": 3.871772522618278,
+      "learning_rate": 1.8630683384905188e-07,
+      "loss": 0.6964,
+      "num_input_tokens_seen": 153756780,
+      "step": 7199
+    },
+    {
+      "epoch": 0.865748812601455,
+      "grad_norm": 2.0663543380812,
+      "learning_rate": 1.859786653480009e-07,
+      "loss": 0.8864,
+      "num_input_tokens_seen": 153771615,
+      "step": 7200
+    },
+    {
+      "epoch": 0.865869055492094,
+      "grad_norm": 2.0056519617050177,
+      "learning_rate": 1.8565077203254398e-07,
+      "loss": 0.7388,
+      "num_input_tokens_seen": 153796795,
+      "step": 7201
+    },
+    {
+      "epoch": 0.8659892983827331,
+      "grad_norm": 3.8256225415782117,
+      "learning_rate": 1.8532315395242203e-07,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 153812965,
+      "step": 7202
+    },
+    {
+      "epoch": 0.8661095412733723,
+      "grad_norm": 2.364187639422691,
+      "learning_rate": 1.849958111573353e-07,
+      "loss": 0.7145,
+      "num_input_tokens_seen": 153831290,
+      "step": 7203
+    },
+    {
+      "epoch": 0.8662297841640113,
+      "grad_norm": 1.9510397215728377,
+      "learning_rate": 1.8466874369694074e-07,
+      "loss": 0.6328,
+      "num_input_tokens_seen": 153848705,
+      "step": 7204
+    },
+    {
+      "epoch": 0.8663500270546504,
+      "grad_norm": 2.417359695290982,
+      "learning_rate": 1.8434195162085443e-07,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 153865350,
+      "step": 7205
+    },
+    {
+      "epoch": 0.8664702699452895,
+      "grad_norm": 3.1851356552454932,
+      "learning_rate": 1.8401543497865023e-07,
+      "loss": 0.7861,
+      "num_input_tokens_seen": 153883070,
+      "step": 7206
+    },
+    {
+      "epoch": 0.8665905128359286,
+      "grad_norm": 12.605545248843363,
+      "learning_rate": 1.836891938198608e-07,
+      "loss": 0.6415,
+      "num_input_tokens_seen": 153903215,
+      "step": 7207
+    },
+    {
+      "epoch": 0.8667107557265676,
+      "grad_norm": 3.5488315507197026,
+      "learning_rate": 1.8336322819397677e-07,
+      "loss": 0.7152,
+      "num_input_tokens_seen": 153920470,
+      "step": 7208
+    },
+    {
+      "epoch": 0.8668309986172068,
+      "grad_norm": 2.4241916386651905,
+      "learning_rate": 1.8303753815044654e-07,
+      "loss": 0.6199,
+      "num_input_tokens_seen": 153939495,
+      "step": 7209
+    },
+    {
+      "epoch": 0.8669512415078459,
+      "grad_norm": 2.429468190799404,
+      "learning_rate": 1.8271212373867684e-07,
+      "loss": 0.7005,
+      "num_input_tokens_seen": 153956660,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8670714843984849,
+      "grad_norm": 3.7225790900409974,
+      "learning_rate": 1.823869850080333e-07,
+      "loss": 0.7489,
+      "num_input_tokens_seen": 153969145,
+      "step": 7211
+    },
+    {
+      "epoch": 0.8671917272891241,
+      "grad_norm": 0.8766079965587327,
+      "learning_rate": 1.820621220078391e-07,
+      "loss": 0.6246,
+      "num_input_tokens_seen": 154032775,
+      "step": 7212
+    },
+    {
+      "epoch": 0.8673119701797631,
+      "grad_norm": 2.625262119250147,
+      "learning_rate": 1.8173753478737553e-07,
+      "loss": 0.669,
+      "num_input_tokens_seen": 154052930,
+      "step": 7213
+    },
+    {
+      "epoch": 0.8674322130704022,
+      "grad_norm": 3.478789691036733,
+      "learning_rate": 1.8141322339588205e-07,
+      "loss": 0.7898,
+      "num_input_tokens_seen": 154069990,
+      "step": 7214
+    },
+    {
+      "epoch": 0.8675524559610414,
+      "grad_norm": 2.5694640397871886,
+      "learning_rate": 1.810891878825569e-07,
+      "loss": 0.6895,
+      "num_input_tokens_seen": 154089685,
+      "step": 7215
+    },
+    {
+      "epoch": 0.8676726988516804,
+      "grad_norm": 2.6537659561055045,
+      "learning_rate": 1.8076542829655561e-07,
+      "loss": 0.7122,
+      "num_input_tokens_seen": 154108210,
+      "step": 7216
+    },
+    {
+      "epoch": 0.8677929417423195,
+      "grad_norm": 2.521058556490146,
+      "learning_rate": 1.8044194468699248e-07,
+      "loss": 0.7989,
+      "num_input_tokens_seen": 154125240,
+      "step": 7217
+    },
+    {
+      "epoch": 0.8679131846329585,
+      "grad_norm": 2.6363711273523633,
+      "learning_rate": 1.8011873710293912e-07,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 154143465,
+      "step": 7218
+    },
+    {
+      "epoch": 0.8680334275235977,
+      "grad_norm": 3.260746694393361,
+      "learning_rate": 1.7979580559342677e-07,
+      "loss": 0.6885,
+      "num_input_tokens_seen": 154163915,
+      "step": 7219
+    },
+    {
+      "epoch": 0.8681536704142367,
+      "grad_norm": 1.9330496047579002,
+      "learning_rate": 1.7947315020744358e-07,
+      "loss": 0.6584,
+      "num_input_tokens_seen": 154184730,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8682739133048758,
+      "grad_norm": 1.880700272046773,
+      "learning_rate": 1.7915077099393594e-07,
+      "loss": 0.7866,
+      "num_input_tokens_seen": 154201050,
+      "step": 7221
+    },
+    {
+      "epoch": 0.868394156195515,
+      "grad_norm": 2.3584848776473586,
+      "learning_rate": 1.788286680018083e-07,
+      "loss": 0.7369,
+      "num_input_tokens_seen": 154219480,
+      "step": 7222
+    },
+    {
+      "epoch": 0.868514399086154,
+      "grad_norm": 2.3088326095588028,
+      "learning_rate": 1.7850684127992398e-07,
+      "loss": 0.7179,
+      "num_input_tokens_seen": 154238945,
+      "step": 7223
+    },
+    {
+      "epoch": 0.8686346419767931,
+      "grad_norm": 1.9794273381585965,
+      "learning_rate": 1.7818529087710378e-07,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 154259020,
+      "step": 7224
+    },
+    {
+      "epoch": 0.8687548848674322,
+      "grad_norm": 1.9152934112428484,
+      "learning_rate": 1.7786401684212637e-07,
+      "loss": 0.8426,
+      "num_input_tokens_seen": 154277570,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8688751277580713,
+      "grad_norm": 0.7503385091815515,
+      "learning_rate": 1.7754301922372838e-07,
+      "loss": 0.5712,
+      "num_input_tokens_seen": 154326935,
+      "step": 7226
+    },
+    {
+      "epoch": 0.8689953706487104,
+      "grad_norm": 2.053629416093874,
+      "learning_rate": 1.7722229807060617e-07,
+      "loss": 0.8031,
+      "num_input_tokens_seen": 154345235,
+      "step": 7227
+    },
+    {
+      "epoch": 0.8691156135393495,
+      "grad_norm": 3.3843710108863205,
+      "learning_rate": 1.7690185343141172e-07,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 154364870,
+      "step": 7228
+    },
+    {
+      "epoch": 0.8692358564299886,
+      "grad_norm": 2.1988372194241936,
+      "learning_rate": 1.7658168535475636e-07,
+      "loss": 0.7065,
+      "num_input_tokens_seen": 154382375,
+      "step": 7229
+    },
+    {
+      "epoch": 0.8693560993206276,
+      "grad_norm": 1.7677365937711709,
+      "learning_rate": 1.7626179388920948e-07,
+      "loss": 0.642,
+      "num_input_tokens_seen": 154403375,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8694763422112668,
+      "grad_norm": 2.157889907551522,
+      "learning_rate": 1.7594217908329866e-07,
+      "loss": 0.8004,
+      "num_input_tokens_seen": 154425280,
+      "step": 7231
+    },
+    {
+      "epoch": 0.8695965851019059,
+      "grad_norm": 2.5372761874371497,
+      "learning_rate": 1.7562284098550895e-07,
+      "loss": 0.73,
+      "num_input_tokens_seen": 154444710,
+      "step": 7232
+    },
+    {
+      "epoch": 0.8697168279925449,
+      "grad_norm": 0.8987151687924567,
+      "learning_rate": 1.753037796442838e-07,
+      "loss": 0.6638,
+      "num_input_tokens_seen": 154503870,
+      "step": 7233
+    },
+    {
+      "epoch": 0.8698370708831841,
+      "grad_norm": 2.707181722592954,
+      "learning_rate": 1.74984995108024e-07,
+      "loss": 0.7423,
+      "num_input_tokens_seen": 154521520,
+      "step": 7234
+    },
+    {
+      "epoch": 0.8699573137738231,
+      "grad_norm": 2.193273155476818,
+      "learning_rate": 1.7466648742508981e-07,
+      "loss": 0.8238,
+      "num_input_tokens_seen": 154537425,
+      "step": 7235
+    },
+    {
+      "epoch": 0.8700775566644622,
+      "grad_norm": 2.1199136904213507,
+      "learning_rate": 1.7434825664379837e-07,
+      "loss": 0.832,
+      "num_input_tokens_seen": 154555650,
+      "step": 7236
+    },
+    {
+      "epoch": 0.8701977995551013,
+      "grad_norm": 3.637289849030688,
+      "learning_rate": 1.740303028124246e-07,
+      "loss": 0.8603,
+      "num_input_tokens_seen": 154571430,
+      "step": 7237
+    },
+    {
+      "epoch": 0.8703180424457404,
+      "grad_norm": 2.0837906820959997,
+      "learning_rate": 1.7371262597920188e-07,
+      "loss": 0.753,
+      "num_input_tokens_seen": 154593210,
+      "step": 7238
+    },
+    {
+      "epoch": 0.8704382853363795,
+      "grad_norm": 1.5109766343947568,
+      "learning_rate": 1.7339522619232195e-07,
+      "loss": 0.7523,
+      "num_input_tokens_seen": 154611310,
+      "step": 7239
+    },
+    {
+      "epoch": 0.8705585282270186,
+      "grad_norm": 4.104608892162011,
+      "learning_rate": 1.730781034999338e-07,
+      "loss": 0.7466,
+      "num_input_tokens_seen": 154632610,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8706787711176577,
+      "grad_norm": 2.2682240701066587,
+      "learning_rate": 1.7276125795014497e-07,
+      "loss": 0.7365,
+      "num_input_tokens_seen": 154650780,
+      "step": 7241
+    },
+    {
+      "epoch": 0.8707990140082967,
+      "grad_norm": 2.187120399509654,
+      "learning_rate": 1.7244468959102054e-07,
+      "loss": 0.6667,
+      "num_input_tokens_seen": 154667555,
+      "step": 7242
+    },
+    {
+      "epoch": 0.8709192568989359,
+      "grad_norm": 2.3911697489612855,
+      "learning_rate": 1.7212839847058348e-07,
+      "loss": 0.8442,
+      "num_input_tokens_seen": 154682405,
+      "step": 7243
+    },
+    {
+      "epoch": 0.871039499789575,
+      "grad_norm": 2.362377550081417,
+      "learning_rate": 1.7181238463681514e-07,
+      "loss": 0.7363,
+      "num_input_tokens_seen": 154701170,
+      "step": 7244
+    },
+    {
+      "epoch": 0.871159742680214,
+      "grad_norm": 2.0016769998469623,
+      "learning_rate": 1.714966481376543e-07,
+      "loss": 0.7155,
+      "num_input_tokens_seen": 154717570,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8712799855708532,
+      "grad_norm": 1.9892272558102755,
+      "learning_rate": 1.7118118902099797e-07,
+      "loss": 0.8197,
+      "num_input_tokens_seen": 154735375,
+      "step": 7246
+    },
+    {
+      "epoch": 0.8714002284614922,
+      "grad_norm": 2.009736885711811,
+      "learning_rate": 1.7086600733470146e-07,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 154755765,
+      "step": 7247
+    },
+    {
+      "epoch": 0.8715204713521313,
+      "grad_norm": 1.8908046441173751,
+      "learning_rate": 1.7055110312657738e-07,
+      "loss": 0.7585,
+      "num_input_tokens_seen": 154774980,
+      "step": 7248
+    },
+    {
+      "epoch": 0.8716407142427703,
+      "grad_norm": 2.503460605775502,
+      "learning_rate": 1.702364764443962e-07,
+      "loss": 0.7369,
+      "num_input_tokens_seen": 154793775,
+      "step": 7249
+    },
+    {
+      "epoch": 0.8717609571334095,
+      "grad_norm": 2.9099414206939205,
+      "learning_rate": 1.699221273358864e-07,
+      "loss": 0.7205,
+      "num_input_tokens_seen": 154813160,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8718812000240486,
+      "grad_norm": 2.199683174645047,
+      "learning_rate": 1.6960805584873538e-07,
+      "loss": 0.7473,
+      "num_input_tokens_seen": 154830880,
+      "step": 7251
+    },
+    {
+      "epoch": 0.8720014429146876,
+      "grad_norm": 2.371451260425662,
+      "learning_rate": 1.6929426203058684e-07,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 154851025,
+      "step": 7252
+    },
+    {
+      "epoch": 0.8721216858053268,
+      "grad_norm": 2.5995821005727184,
+      "learning_rate": 1.689807459290431e-07,
+      "loss": 0.8036,
+      "num_input_tokens_seen": 154869400,
+      "step": 7253
+    },
+    {
+      "epoch": 0.8722419286959658,
+      "grad_norm": 2.348181247702717,
+      "learning_rate": 1.6866750759166392e-07,
+      "loss": 0.6999,
+      "num_input_tokens_seen": 154889100,
+      "step": 7254
+    },
+    {
+      "epoch": 0.8723621715866049,
+      "grad_norm": 2.724670660934635,
+      "learning_rate": 1.683545470659684e-07,
+      "loss": 0.7675,
+      "num_input_tokens_seen": 154906650,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8724824144772441,
+      "grad_norm": 2.2240112864833455,
+      "learning_rate": 1.680418643994317e-07,
+      "loss": 0.7382,
+      "num_input_tokens_seen": 154924940,
+      "step": 7256
+    },
+    {
+      "epoch": 0.8726026573678831,
+      "grad_norm": 0.9865027062594387,
+      "learning_rate": 1.6772945963948738e-07,
+      "loss": 0.6932,
+      "num_input_tokens_seen": 154982825,
+      "step": 7257
+    },
+    {
+      "epoch": 0.8727229002585222,
+      "grad_norm": 2.5453450184035367,
+      "learning_rate": 1.6741733283352733e-07,
+      "loss": 0.7684,
+      "num_input_tokens_seen": 155000150,
+      "step": 7258
+    },
+    {
+      "epoch": 0.8728431431491613,
+      "grad_norm": 2.8346749231499566,
+      "learning_rate": 1.6710548402890102e-07,
+      "loss": 0.8298,
+      "num_input_tokens_seen": 155020395,
+      "step": 7259
+    },
+    {
+      "epoch": 0.8729633860398004,
+      "grad_norm": 1.997145367734175,
+      "learning_rate": 1.6679391327291527e-07,
+      "loss": 0.6625,
+      "num_input_tokens_seen": 155041320,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8730836289304394,
+      "grad_norm": 3.165372802917187,
+      "learning_rate": 1.6648262061283535e-07,
+      "loss": 0.6792,
+      "num_input_tokens_seen": 155056340,
+      "step": 7261
+    },
+    {
+      "epoch": 0.8732038718210786,
+      "grad_norm": 2.0410575566125853,
+      "learning_rate": 1.6617160609588353e-07,
+      "loss": 0.7278,
+      "num_input_tokens_seen": 155075235,
+      "step": 7262
+    },
+    {
+      "epoch": 0.8733241147117177,
+      "grad_norm": 2.535664053738104,
+      "learning_rate": 1.6586086976924163e-07,
+      "loss": 0.7046,
+      "num_input_tokens_seen": 155090455,
+      "step": 7263
+    },
+    {
+      "epoch": 0.8734443576023567,
+      "grad_norm": 2.1026735246017703,
+      "learning_rate": 1.6555041168004747e-07,
+      "loss": 0.786,
+      "num_input_tokens_seen": 155109495,
+      "step": 7264
+    },
+    {
+      "epoch": 0.8735646004929959,
+      "grad_norm": 2.1188259079447227,
+      "learning_rate": 1.6524023187539715e-07,
+      "loss": 0.6854,
+      "num_input_tokens_seen": 155127500,
+      "step": 7265
+    },
+    {
+      "epoch": 0.873684843383635,
+      "grad_norm": 2.612217863088034,
+      "learning_rate": 1.649303304023446e-07,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 155146975,
+      "step": 7266
+    },
+    {
+      "epoch": 0.873805086274274,
+      "grad_norm": 1.7716806477454652,
+      "learning_rate": 1.6462070730790222e-07,
+      "loss": 0.7783,
+      "num_input_tokens_seen": 155165855,
+      "step": 7267
+    },
+    {
+      "epoch": 0.8739253291649132,
+      "grad_norm": 25.477779563158002,
+      "learning_rate": 1.6431136263903912e-07,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 155184575,
+      "step": 7268
+    },
+    {
+      "epoch": 0.8740455720555522,
+      "grad_norm": 1.9790379960731481,
+      "learning_rate": 1.6400229644268282e-07,
+      "loss": 0.7358,
+      "num_input_tokens_seen": 155202650,
+      "step": 7269
+    },
+    {
+      "epoch": 0.8741658149461913,
+      "grad_norm": 2.0632569306880577,
+      "learning_rate": 1.6369350876571852e-07,
+      "loss": 0.8054,
+      "num_input_tokens_seen": 155220525,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8742860578368304,
+      "grad_norm": 4.119014060933731,
+      "learning_rate": 1.6338499965498874e-07,
+      "loss": 0.8055,
+      "num_input_tokens_seen": 155238975,
+      "step": 7271
+    },
+    {
+      "epoch": 0.8744063007274695,
+      "grad_norm": 1.67449513296944,
+      "learning_rate": 1.630767691572943e-07,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 155258715,
+      "step": 7272
+    },
+    {
+      "epoch": 0.8745265436181086,
+      "grad_norm": 0.7726221207645896,
+      "learning_rate": 1.627688173193935e-07,
+      "loss": 0.564,
+      "num_input_tokens_seen": 155320325,
+      "step": 7273
+    },
+    {
+      "epoch": 0.8746467865087477,
+      "grad_norm": 2.142295318549826,
+      "learning_rate": 1.6246114418800193e-07,
+      "loss": 0.7525,
+      "num_input_tokens_seen": 155340325,
+      "step": 7274
+    },
+    {
+      "epoch": 0.8747670293993868,
+      "grad_norm": 2.241612257790113,
+      "learning_rate": 1.6215374980979423e-07,
+      "loss": 0.7604,
+      "num_input_tokens_seen": 155360455,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8748872722900258,
+      "grad_norm": 5.116313393513584,
+      "learning_rate": 1.6184663423140133e-07,
+      "loss": 0.6847,
+      "num_input_tokens_seen": 155380475,
+      "step": 7276
+    },
+    {
+      "epoch": 0.875007515180665,
+      "grad_norm": 2.5639926860542843,
+      "learning_rate": 1.615397974994126e-07,
+      "loss": 0.6433,
+      "num_input_tokens_seen": 155398000,
+      "step": 7277
+    },
+    {
+      "epoch": 0.875127758071304,
+      "grad_norm": 1.6899886406677598,
+      "learning_rate": 1.6123323966037438e-07,
+      "loss": 0.7947,
+      "num_input_tokens_seen": 155416240,
+      "step": 7278
+    },
+    {
+      "epoch": 0.8752480009619431,
+      "grad_norm": 2.6514341977765112,
+      "learning_rate": 1.6092696076079216e-07,
+      "loss": 0.7765,
+      "num_input_tokens_seen": 155434335,
+      "step": 7279
+    },
+    {
+      "epoch": 0.8753682438525822,
+      "grad_norm": 2.1219837291097585,
+      "learning_rate": 1.6062096084712785e-07,
+      "loss": 0.7303,
+      "num_input_tokens_seen": 155455405,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8754884867432213,
+      "grad_norm": 2.0267282598001586,
+      "learning_rate": 1.6031523996580098e-07,
+      "loss": 0.7034,
+      "num_input_tokens_seen": 155472685,
+      "step": 7281
+    },
+    {
+      "epoch": 0.8756087296338604,
+      "grad_norm": 2.452921777603213,
+      "learning_rate": 1.600097981631894e-07,
+      "loss": 0.6552,
+      "num_input_tokens_seen": 155490870,
+      "step": 7282
+    },
+    {
+      "epoch": 0.8757289725244994,
+      "grad_norm": 2.6973055406511013,
+      "learning_rate": 1.5970463548562886e-07,
+      "loss": 0.7507,
+      "num_input_tokens_seen": 155509745,
+      "step": 7283
+    },
+    {
+      "epoch": 0.8758492154151386,
+      "grad_norm": 2.3713795344776427,
+      "learning_rate": 1.5939975197941192e-07,
+      "loss": 0.71,
+      "num_input_tokens_seen": 155531120,
+      "step": 7284
+    },
+    {
+      "epoch": 0.8759694583057777,
+      "grad_norm": 0.8371535968498088,
+      "learning_rate": 1.5909514769078892e-07,
+      "loss": 0.5623,
+      "num_input_tokens_seen": 155595945,
+      "step": 7285
+    },
+    {
+      "epoch": 0.8760897011964167,
+      "grad_norm": 1.7353195151683674,
+      "learning_rate": 1.5879082266596867e-07,
+      "loss": 0.7687,
+      "num_input_tokens_seen": 155617005,
+      "step": 7286
+    },
+    {
+      "epoch": 0.8762099440870559,
+      "grad_norm": 3.8741979204224317,
+      "learning_rate": 1.5848677695111645e-07,
+      "loss": 0.7137,
+      "num_input_tokens_seen": 155638325,
+      "step": 7287
+    },
+    {
+      "epoch": 0.8763301869776949,
+      "grad_norm": 3.1890496768097476,
+      "learning_rate": 1.5818301059235607e-07,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 155653220,
+      "step": 7288
+    },
+    {
+      "epoch": 0.876450429868334,
+      "grad_norm": 1.7995767709116486,
+      "learning_rate": 1.578795236357684e-07,
+      "loss": 0.811,
+      "num_input_tokens_seen": 155674405,
+      "step": 7289
+    },
+    {
+      "epoch": 0.8765706727589732,
+      "grad_norm": 2.8384049598305965,
+      "learning_rate": 1.5757631612739218e-07,
+      "loss": 0.8448,
+      "num_input_tokens_seen": 155687670,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8766909156496122,
+      "grad_norm": 0.8806307760284282,
+      "learning_rate": 1.572733881132242e-07,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 155748035,
+      "step": 7291
+    },
+    {
+      "epoch": 0.8768111585402513,
+      "grad_norm": 0.7984432667858014,
+      "learning_rate": 1.5697073963921814e-07,
+      "loss": 0.6136,
+      "num_input_tokens_seen": 155806995,
+      "step": 7292
+    },
+    {
+      "epoch": 0.8769314014308904,
+      "grad_norm": 2.3234911821240414,
+      "learning_rate": 1.566683707512857e-07,
+      "loss": 0.8472,
+      "num_input_tokens_seen": 155824390,
+      "step": 7293
+    },
+    {
+      "epoch": 0.8770516443215295,
+      "grad_norm": 2.372146328883043,
+      "learning_rate": 1.5636628149529508e-07,
+      "loss": 0.7875,
+      "num_input_tokens_seen": 155841900,
+      "step": 7294
+    },
+    {
+      "epoch": 0.8771718872121685,
+      "grad_norm": 2.4177423173776704,
+      "learning_rate": 1.560644719170743e-07,
+      "loss": 0.7818,
+      "num_input_tokens_seen": 155862490,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8772921301028077,
+      "grad_norm": 2.247949290823861,
+      "learning_rate": 1.5576294206240692e-07,
+      "loss": 0.7076,
+      "num_input_tokens_seen": 155881735,
+      "step": 7296
+    },
+    {
+      "epoch": 0.8774123729934468,
+      "grad_norm": 2.028692490742412,
+      "learning_rate": 1.5546169197703507e-07,
+      "loss": 0.6831,
+      "num_input_tokens_seen": 155907730,
+      "step": 7297
+    },
+    {
+      "epoch": 0.8775326158840858,
+      "grad_norm": 4.657685726373577,
+      "learning_rate": 1.551607217066575e-07,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 155925420,
+      "step": 7298
+    },
+    {
+      "epoch": 0.877652858774725,
+      "grad_norm": 1.7297149813438113,
+      "learning_rate": 1.5486003129693193e-07,
+      "loss": 0.8534,
+      "num_input_tokens_seen": 155942505,
+      "step": 7299
+    },
+    {
+      "epoch": 0.877773101665364,
+      "grad_norm": 2.006525888370653,
+      "learning_rate": 1.545596207934725e-07,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 155960710,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8778933445560031,
+      "grad_norm": 1.9001422590215593,
+      "learning_rate": 1.5425949024185147e-07,
+      "loss": 0.771,
+      "num_input_tokens_seen": 155980455,
+      "step": 7301
+    },
+    {
+      "epoch": 0.8780135874466423,
+      "grad_norm": 2.7000747616830045,
+      "learning_rate": 1.5395963968759818e-07,
+      "loss": 0.6772,
+      "num_input_tokens_seen": 156000450,
+      "step": 7302
+    },
+    {
+      "epoch": 0.8781338303372813,
+      "grad_norm": 1.702062347682195,
+      "learning_rate": 1.536600691761998e-07,
+      "loss": 0.6364,
+      "num_input_tokens_seen": 156026000,
+      "step": 7303
+    },
+    {
+      "epoch": 0.8782540732279204,
+      "grad_norm": 1.8085981974383063,
+      "learning_rate": 1.5336077875310084e-07,
+      "loss": 0.7056,
+      "num_input_tokens_seen": 156044945,
+      "step": 7304
+    },
+    {
+      "epoch": 0.8783743161185595,
+      "grad_norm": 3.598424529556889,
+      "learning_rate": 1.5306176846370345e-07,
+      "loss": 0.7386,
+      "num_input_tokens_seen": 156062810,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8784945590091986,
+      "grad_norm": 6.992587157182235,
+      "learning_rate": 1.5276303835336712e-07,
+      "loss": 0.7397,
+      "num_input_tokens_seen": 156083070,
+      "step": 7306
+    },
+    {
+      "epoch": 0.8786148018998376,
+      "grad_norm": 0.7998410538781606,
+      "learning_rate": 1.524645884674094e-07,
+      "loss": 0.5562,
+      "num_input_tokens_seen": 156139720,
+      "step": 7307
+    },
+    {
+      "epoch": 0.8787350447904768,
+      "grad_norm": 2.398837101047178,
+      "learning_rate": 1.521664188511047e-07,
+      "loss": 0.7854,
+      "num_input_tokens_seen": 156159465,
+      "step": 7308
+    },
+    {
+      "epoch": 0.8788552876811159,
+      "grad_norm": 2.5265979537210215,
+      "learning_rate": 1.518685295496851e-07,
+      "loss": 0.8103,
+      "num_input_tokens_seen": 156177045,
+      "step": 7309
+    },
+    {
+      "epoch": 0.8789755305717549,
+      "grad_norm": 1.7898697532212067,
+      "learning_rate": 1.5157092060833975e-07,
+      "loss": 0.8439,
+      "num_input_tokens_seen": 156196415,
+      "step": 7310
+    },
+    {
+      "epoch": 0.879095773462394,
+      "grad_norm": 1.7132263754821253,
+      "learning_rate": 1.5127359207221635e-07,
+      "loss": 0.6529,
+      "num_input_tokens_seen": 156215615,
+      "step": 7311
+    },
+    {
+      "epoch": 0.8792160163530331,
+      "grad_norm": 2.093428046917588,
+      "learning_rate": 1.5097654398641923e-07,
+      "loss": 0.7187,
+      "num_input_tokens_seen": 156233240,
+      "step": 7312
+    },
+    {
+      "epoch": 0.8793362592436722,
+      "grad_norm": 1.403515511192551,
+      "learning_rate": 1.5067977639601014e-07,
+      "loss": 0.7241,
+      "num_input_tokens_seen": 156255720,
+      "step": 7313
+    },
+    {
+      "epoch": 0.8794565021343113,
+      "grad_norm": 3.071228494945204,
+      "learning_rate": 1.5038328934600864e-07,
+      "loss": 0.7069,
+      "num_input_tokens_seen": 156272075,
+      "step": 7314
+    },
+    {
+      "epoch": 0.8795767450249504,
+      "grad_norm": 2.002934039991062,
+      "learning_rate": 1.5008708288139161e-07,
+      "loss": 0.6971,
+      "num_input_tokens_seen": 156294155,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8796969879155895,
+      "grad_norm": 2.0579384232195688,
+      "learning_rate": 1.497911570470931e-07,
+      "loss": 0.724,
+      "num_input_tokens_seen": 156313880,
+      "step": 7316
+    },
+    {
+      "epoch": 0.8798172308062285,
+      "grad_norm": 1.7319519990520393,
+      "learning_rate": 1.4949551188800502e-07,
+      "loss": 0.8483,
+      "num_input_tokens_seen": 156334585,
+      "step": 7317
+    },
+    {
+      "epoch": 0.8799374736968677,
+      "grad_norm": 1.6536701667612526,
+      "learning_rate": 1.4920014744897634e-07,
+      "loss": 0.7219,
+      "num_input_tokens_seen": 156353720,
+      "step": 7318
+    },
+    {
+      "epoch": 0.8800577165875068,
+      "grad_norm": 2.7264760224295217,
+      "learning_rate": 1.4890506377481392e-07,
+      "loss": 0.8608,
+      "num_input_tokens_seen": 156372530,
+      "step": 7319
+    },
+    {
+      "epoch": 0.8801779594781458,
+      "grad_norm": 1.8113670099325259,
+      "learning_rate": 1.486102609102815e-07,
+      "loss": 0.6342,
+      "num_input_tokens_seen": 156392800,
+      "step": 7320
+    },
+    {
+      "epoch": 0.880298202368785,
+      "grad_norm": 3.2734731952769187,
+      "learning_rate": 1.483157389001004e-07,
+      "loss": 0.8486,
+      "num_input_tokens_seen": 156410080,
+      "step": 7321
+    },
+    {
+      "epoch": 0.880418445259424,
+      "grad_norm": 2.767476507298718,
+      "learning_rate": 1.4802149778894933e-07,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 156428590,
+      "step": 7322
+    },
+    {
+      "epoch": 0.8805386881500631,
+      "grad_norm": 1.655509141066757,
+      "learning_rate": 1.4772753762146484e-07,
+      "loss": 0.8698,
+      "num_input_tokens_seen": 156447565,
+      "step": 7323
+    },
+    {
+      "epoch": 0.8806589310407023,
+      "grad_norm": 1.9936651901005138,
+      "learning_rate": 1.474338584422401e-07,
+      "loss": 0.696,
+      "num_input_tokens_seen": 156472495,
+      "step": 7324
+    },
+    {
+      "epoch": 0.8807791739313413,
+      "grad_norm": 2.30308829396143,
+      "learning_rate": 1.4714046029582595e-07,
+      "loss": 0.7511,
+      "num_input_tokens_seen": 156491280,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8808994168219804,
+      "grad_norm": 7.504067307446644,
+      "learning_rate": 1.4684734322673075e-07,
+      "loss": 0.7518,
+      "num_input_tokens_seen": 156512040,
+      "step": 7326
+    },
+    {
+      "epoch": 0.8810196597126195,
+      "grad_norm": 3.17519253007266,
+      "learning_rate": 1.465545072794203e-07,
+      "loss": 0.6877,
+      "num_input_tokens_seen": 156529635,
+      "step": 7327
+    },
+    {
+      "epoch": 0.8811399026032586,
+      "grad_norm": 1.725685443873617,
+      "learning_rate": 1.4626195249831753e-07,
+      "loss": 0.75,
+      "num_input_tokens_seen": 156550255,
+      "step": 7328
+    },
+    {
+      "epoch": 0.8812601454938976,
+      "grad_norm": 2.1093932448411263,
+      "learning_rate": 1.4596967892780244e-07,
+      "loss": 0.7118,
+      "num_input_tokens_seen": 156566305,
+      "step": 7329
+    },
+    {
+      "epoch": 0.8813803883845368,
+      "grad_norm": 2.2395660462486964,
+      "learning_rate": 1.4567768661221314e-07,
+      "loss": 0.744,
+      "num_input_tokens_seen": 156586595,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8815006312751759,
+      "grad_norm": 2.553811892529049,
+      "learning_rate": 1.4538597559584442e-07,
+      "loss": 0.7359,
+      "num_input_tokens_seen": 156604105,
+      "step": 7331
+    },
+    {
+      "epoch": 0.8816208741658149,
+      "grad_norm": 2.1129890530321607,
+      "learning_rate": 1.4509454592294868e-07,
+      "loss": 0.7674,
+      "num_input_tokens_seen": 156624310,
+      "step": 7332
+    },
+    {
+      "epoch": 0.8817411170564541,
+      "grad_norm": 1.9864245695085363,
+      "learning_rate": 1.448033976377354e-07,
+      "loss": 0.7879,
+      "num_input_tokens_seen": 156639015,
+      "step": 7333
+    },
+    {
+      "epoch": 0.8818613599470931,
+      "grad_norm": 2.850459602558032,
+      "learning_rate": 1.445125307843713e-07,
+      "loss": 0.7315,
+      "num_input_tokens_seen": 156656960,
+      "step": 7334
+    },
+    {
+      "epoch": 0.8819816028377322,
+      "grad_norm": 5.7091869677963345,
+      "learning_rate": 1.442219454069813e-07,
+      "loss": 0.7445,
+      "num_input_tokens_seen": 156677705,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8821018457283714,
+      "grad_norm": 2.4291813698257987,
+      "learning_rate": 1.4393164154964676e-07,
+      "loss": 0.6596,
+      "num_input_tokens_seen": 156696955,
+      "step": 7336
+    },
+    {
+      "epoch": 0.8822220886190104,
+      "grad_norm": 1.8521979586937316,
+      "learning_rate": 1.4364161925640649e-07,
+      "loss": 0.928,
+      "num_input_tokens_seen": 156718075,
+      "step": 7337
+    },
+    {
+      "epoch": 0.8823423315096495,
+      "grad_norm": 2.4251544194735564,
+      "learning_rate": 1.4335187857125618e-07,
+      "loss": 0.8438,
+      "num_input_tokens_seen": 156736495,
+      "step": 7338
+    },
+    {
+      "epoch": 0.8824625744002886,
+      "grad_norm": 2.143483609906486,
+      "learning_rate": 1.4306241953815023e-07,
+      "loss": 0.744,
+      "num_input_tokens_seen": 156757275,
+      "step": 7339
+    },
+    {
+      "epoch": 0.8825828172909277,
+      "grad_norm": 2.5192539981486455,
+      "learning_rate": 1.4277324220099862e-07,
+      "loss": 0.7025,
+      "num_input_tokens_seen": 156778905,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8827030601815667,
+      "grad_norm": 2.8641593289907576,
+      "learning_rate": 1.4248434660366938e-07,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 156798100,
+      "step": 7341
+    },
+    {
+      "epoch": 0.8828233030722058,
+      "grad_norm": 5.4814248673686174,
+      "learning_rate": 1.4219573278998765e-07,
+      "loss": 0.7033,
+      "num_input_tokens_seen": 156816280,
+      "step": 7342
+    },
+    {
+      "epoch": 0.882943545962845,
+      "grad_norm": 2.445820209226076,
+      "learning_rate": 1.4190740080373664e-07,
+      "loss": 0.648,
+      "num_input_tokens_seen": 156836280,
+      "step": 7343
+    },
+    {
+      "epoch": 0.883063788853484,
+      "grad_norm": 2.047721290995996,
+      "learning_rate": 1.4161935068865538e-07,
+      "loss": 0.8384,
+      "num_input_tokens_seen": 156851145,
+      "step": 7344
+    },
+    {
+      "epoch": 0.8831840317441231,
+      "grad_norm": 2.3130893788321925,
+      "learning_rate": 1.4133158248844113e-07,
+      "loss": 0.747,
+      "num_input_tokens_seen": 156869770,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8833042746347622,
+      "grad_norm": 2.0716009581216324,
+      "learning_rate": 1.4104409624674785e-07,
+      "loss": 0.724,
+      "num_input_tokens_seen": 156889275,
+      "step": 7346
+    },
+    {
+      "epoch": 0.8834245175254013,
+      "grad_norm": 1.890763442574728,
+      "learning_rate": 1.407568920071873e-07,
+      "loss": 0.7789,
+      "num_input_tokens_seen": 156907860,
+      "step": 7347
+    },
+    {
+      "epoch": 0.8835447604160404,
+      "grad_norm": 2.35163211453435,
+      "learning_rate": 1.4046996981332782e-07,
+      "loss": 0.6696,
+      "num_input_tokens_seen": 156927465,
+      "step": 7348
+    },
+    {
+      "epoch": 0.8836650033066795,
+      "grad_norm": 2.142423432451787,
+      "learning_rate": 1.4018332970869561e-07,
+      "loss": 0.7712,
+      "num_input_tokens_seen": 156945125,
+      "step": 7349
+    },
+    {
+      "epoch": 0.8837852461973186,
+      "grad_norm": 2.1744484176091605,
+      "learning_rate": 1.3989697173677305e-07,
+      "loss": 0.8389,
+      "num_input_tokens_seen": 156966170,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8839054890879576,
+      "grad_norm": 2.030314446021492,
+      "learning_rate": 1.396108959410014e-07,
+      "loss": 0.7615,
+      "num_input_tokens_seen": 156985105,
+      "step": 7351
+    },
+    {
+      "epoch": 0.8840257319785968,
+      "grad_norm": 1.9026112079934179,
+      "learning_rate": 1.3932510236477745e-07,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 157005495,
+      "step": 7352
+    },
+    {
+      "epoch": 0.8841459748692359,
+      "grad_norm": 2.083546629977031,
+      "learning_rate": 1.3903959105145636e-07,
+      "loss": 0.5489,
+      "num_input_tokens_seen": 157025705,
+      "step": 7353
+    },
+    {
+      "epoch": 0.8842662177598749,
+      "grad_norm": 2.2061970657174483,
+      "learning_rate": 1.387543620443492e-07,
+      "loss": 0.8213,
+      "num_input_tokens_seen": 157042270,
+      "step": 7354
+    },
+    {
+      "epoch": 0.8843864606505141,
+      "grad_norm": 1.861115844720276,
+      "learning_rate": 1.3846941538672564e-07,
+      "loss": 0.832,
+      "num_input_tokens_seen": 157060695,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8845067035411531,
+      "grad_norm": 2.42261155932535,
+      "learning_rate": 1.3818475112181193e-07,
+      "loss": 0.8073,
+      "num_input_tokens_seen": 157079210,
+      "step": 7356
+    },
+    {
+      "epoch": 0.8846269464317922,
+      "grad_norm": 2.167692392567915,
+      "learning_rate": 1.3790036929279091e-07,
+      "loss": 0.784,
+      "num_input_tokens_seen": 157096085,
+      "step": 7357
+    },
+    {
+      "epoch": 0.8847471893224313,
+      "grad_norm": 2.763826260668255,
+      "learning_rate": 1.3761626994280363e-07,
+      "loss": 0.5815,
+      "num_input_tokens_seen": 157113275,
+      "step": 7358
+    },
+    {
+      "epoch": 0.8848674322130704,
+      "grad_norm": 1.9188592581836388,
+      "learning_rate": 1.3733245311494735e-07,
+      "loss": 0.7286,
+      "num_input_tokens_seen": 157135650,
+      "step": 7359
+    },
+    {
+      "epoch": 0.8849876751037095,
+      "grad_norm": 8.317519376224942,
+      "learning_rate": 1.3704891885227676e-07,
+      "loss": 0.707,
+      "num_input_tokens_seen": 157155415,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8851079179943486,
+      "grad_norm": 2.245657821116547,
+      "learning_rate": 1.3676566719780414e-07,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 157172600,
+      "step": 7361
+    },
+    {
+      "epoch": 0.8852281608849877,
+      "grad_norm": 1.974864910709764,
+      "learning_rate": 1.36482698194498e-07,
+      "loss": 0.7406,
+      "num_input_tokens_seen": 157188865,
+      "step": 7362
+    },
+    {
+      "epoch": 0.8853484037756267,
+      "grad_norm": 2.455228309467207,
+      "learning_rate": 1.3620001188528506e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 157209305,
+      "step": 7363
+    },
+    {
+      "epoch": 0.8854686466662659,
+      "grad_norm": 2.5781171158206084,
+      "learning_rate": 1.3591760831304865e-07,
+      "loss": 0.7227,
+      "num_input_tokens_seen": 157226715,
+      "step": 7364
+    },
+    {
+      "epoch": 0.885588889556905,
+      "grad_norm": 9.118166295748768,
+      "learning_rate": 1.356354875206287e-07,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 157244270,
+      "step": 7365
+    },
+    {
+      "epoch": 0.885709132447544,
+      "grad_norm": 2.1402908511079652,
+      "learning_rate": 1.3535364955082296e-07,
+      "loss": 0.6854,
+      "num_input_tokens_seen": 157263840,
+      "step": 7366
+    },
+    {
+      "epoch": 0.8858293753381832,
+      "grad_norm": 1.8522966181911544,
+      "learning_rate": 1.3507209444638613e-07,
+      "loss": 0.6328,
+      "num_input_tokens_seen": 157285560,
+      "step": 7367
+    },
+    {
+      "epoch": 0.8859496182288222,
+      "grad_norm": 2.3928223980956136,
+      "learning_rate": 1.347908222500298e-07,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 157305355,
+      "step": 7368
+    },
+    {
+      "epoch": 0.8860698611194613,
+      "grad_norm": 2.7846449558826594,
+      "learning_rate": 1.3450983300442276e-07,
+      "loss": 0.695,
+      "num_input_tokens_seen": 157324305,
+      "step": 7369
+    },
+    {
+      "epoch": 0.8861901040101005,
+      "grad_norm": 1.938876681017681,
+      "learning_rate": 1.3422912675219044e-07,
+      "loss": 0.7266,
+      "num_input_tokens_seen": 157343780,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8863103469007395,
+      "grad_norm": 2.1202783117260533,
+      "learning_rate": 1.339487035359166e-07,
+      "loss": 0.7763,
+      "num_input_tokens_seen": 157363870,
+      "step": 7371
+    },
+    {
+      "epoch": 0.8864305897913786,
+      "grad_norm": 2.0667737806587216,
+      "learning_rate": 1.3366856339814049e-07,
+      "loss": 0.8444,
+      "num_input_tokens_seen": 157384675,
+      "step": 7372
+    },
+    {
+      "epoch": 0.8865508326820177,
+      "grad_norm": 2.2264483313600723,
+      "learning_rate": 1.333887063813597e-07,
+      "loss": 0.7374,
+      "num_input_tokens_seen": 157402500,
+      "step": 7373
+    },
+    {
+      "epoch": 0.8866710755726568,
+      "grad_norm": 6.017501354892182,
+      "learning_rate": 1.331091325280278e-07,
+      "loss": 0.6595,
+      "num_input_tokens_seen": 157421190,
+      "step": 7374
+    },
+    {
+      "epoch": 0.8867913184632958,
+      "grad_norm": 2.4631218893722178,
+      "learning_rate": 1.3282984188055625e-07,
+      "loss": 0.7792,
+      "num_input_tokens_seen": 157440700,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8869115613539349,
+      "grad_norm": 2.1149320955758952,
+      "learning_rate": 1.325508344813131e-07,
+      "loss": 0.7856,
+      "num_input_tokens_seen": 157459465,
+      "step": 7376
+    },
+    {
+      "epoch": 0.8870318042445741,
+      "grad_norm": 2.4888044248203482,
+      "learning_rate": 1.3227211037262365e-07,
+      "loss": 0.781,
+      "num_input_tokens_seen": 157476425,
+      "step": 7377
+    },
+    {
+      "epoch": 0.8871520471352131,
+      "grad_norm": 2.317447241394007,
+      "learning_rate": 1.319936695967696e-07,
+      "loss": 0.8553,
+      "num_input_tokens_seen": 157493970,
+      "step": 7378
+    },
+    {
+      "epoch": 0.8872722900258522,
+      "grad_norm": 2.5840236606613813,
+      "learning_rate": 1.3171551219599097e-07,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 157512215,
+      "step": 7379
+    },
+    {
+      "epoch": 0.8873925329164913,
+      "grad_norm": 2.3898624801931456,
+      "learning_rate": 1.3143763821248377e-07,
+      "loss": 0.7702,
+      "num_input_tokens_seen": 157529020,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8875127758071304,
+      "grad_norm": 1.8238026448901838,
+      "learning_rate": 1.3116004768840118e-07,
+      "loss": 0.7129,
+      "num_input_tokens_seen": 157547115,
+      "step": 7381
+    },
+    {
+      "epoch": 0.8876330186977694,
+      "grad_norm": 1.8422589716224764,
+      "learning_rate": 1.3088274066585303e-07,
+      "loss": 0.7262,
+      "num_input_tokens_seen": 157564445,
+      "step": 7382
+    },
+    {
+      "epoch": 0.8877532615884086,
+      "grad_norm": 2.2649777083044045,
+      "learning_rate": 1.3060571718690749e-07,
+      "loss": 0.8965,
+      "num_input_tokens_seen": 157581660,
+      "step": 7383
+    },
+    {
+      "epoch": 0.8878735044790477,
+      "grad_norm": 0.7839944321922209,
+      "learning_rate": 1.3032897729358805e-07,
+      "loss": 0.5974,
+      "num_input_tokens_seen": 157642335,
+      "step": 7384
+    },
+    {
+      "epoch": 0.8879937473696867,
+      "grad_norm": 2.6732217857331855,
+      "learning_rate": 1.3005252102787645e-07,
+      "loss": 0.7899,
+      "num_input_tokens_seen": 157660995,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8881139902603259,
+      "grad_norm": 1.7803093473260332,
+      "learning_rate": 1.2977634843171025e-07,
+      "loss": 0.7313,
+      "num_input_tokens_seen": 157679010,
+      "step": 7386
+    },
+    {
+      "epoch": 0.888234233150965,
+      "grad_norm": 2.4494793090637015,
+      "learning_rate": 1.295004595469853e-07,
+      "loss": 0.7028,
+      "num_input_tokens_seen": 157696565,
+      "step": 7387
+    },
+    {
+      "epoch": 0.888354476041604,
+      "grad_norm": 2.362765108782187,
+      "learning_rate": 1.2922485441555343e-07,
+      "loss": 0.7468,
+      "num_input_tokens_seen": 157715365,
+      "step": 7388
+    },
+    {
+      "epoch": 0.8884747189322432,
+      "grad_norm": 2.3954695551139538,
+      "learning_rate": 1.2894953307922363e-07,
+      "loss": 0.8136,
+      "num_input_tokens_seen": 157734045,
+      "step": 7389
+    },
+    {
+      "epoch": 0.8885949618228822,
+      "grad_norm": 2.0646279259217453,
+      "learning_rate": 1.2867449557976208e-07,
+      "loss": 0.8347,
+      "num_input_tokens_seen": 157751865,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8887152047135213,
+      "grad_norm": 2.3890209817780286,
+      "learning_rate": 1.283997419588916e-07,
+      "loss": 0.755,
+      "num_input_tokens_seen": 157771055,
+      "step": 7391
+    },
+    {
+      "epoch": 0.8888354476041604,
+      "grad_norm": 1.9512523927845635,
+      "learning_rate": 1.2812527225829216e-07,
+      "loss": 0.6118,
+      "num_input_tokens_seen": 157789000,
+      "step": 7392
+    },
+    {
+      "epoch": 0.8889556904947995,
+      "grad_norm": 7.857630504034833,
+      "learning_rate": 1.2785108651960076e-07,
+      "loss": 0.7555,
+      "num_input_tokens_seen": 157810355,
+      "step": 7393
+    },
+    {
+      "epoch": 0.8890759333854386,
+      "grad_norm": 2.7157426842107015,
+      "learning_rate": 1.275771847844105e-07,
+      "loss": 0.7885,
+      "num_input_tokens_seen": 157830820,
+      "step": 7394
+    },
+    {
+      "epoch": 0.8891961762760777,
+      "grad_norm": 1.900291048993054,
+      "learning_rate": 1.2730356709427302e-07,
+      "loss": 0.7691,
+      "num_input_tokens_seen": 157849220,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8893164191667168,
+      "grad_norm": 2.16600915867931,
+      "learning_rate": 1.2703023349069542e-07,
+      "loss": 0.5902,
+      "num_input_tokens_seen": 157873790,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8894366620573558,
+      "grad_norm": 2.4687199533663216,
+      "learning_rate": 1.2675718401514223e-07,
+      "loss": 0.6115,
+      "num_input_tokens_seen": 157897690,
+      "step": 7397
+    },
+    {
+      "epoch": 0.889556904947995,
+      "grad_norm": 2.477721961456672,
+      "learning_rate": 1.264844187090346e-07,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 157914535,
+      "step": 7398
+    },
+    {
+      "epoch": 0.889677147838634,
+      "grad_norm": 1.812463232675726,
+      "learning_rate": 1.2621193761375116e-07,
+      "loss": 0.7456,
+      "num_input_tokens_seen": 157935315,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8897973907292731,
+      "grad_norm": 1.807733222729297,
+      "learning_rate": 1.2593974077062707e-07,
+      "loss": 0.8404,
+      "num_input_tokens_seen": 157956655,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8899176336199123,
+      "grad_norm": 1.6680510105211415,
+      "learning_rate": 1.2566782822095423e-07,
+      "loss": 0.6183,
+      "num_input_tokens_seen": 157976630,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8900378765105513,
+      "grad_norm": 3.7925560935072222,
+      "learning_rate": 1.2539620000598162e-07,
+      "loss": 0.7152,
+      "num_input_tokens_seen": 157995685,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8901581194011904,
+      "grad_norm": 1.8352876282218142,
+      "learning_rate": 1.2512485616691492e-07,
+      "loss": 0.7905,
+      "num_input_tokens_seen": 158012460,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8902783622918296,
+      "grad_norm": 2.6441627331042903,
+      "learning_rate": 1.2485379674491681e-07,
+      "loss": 0.7992,
+      "num_input_tokens_seen": 158038375,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8903986051824686,
+      "grad_norm": 2.5340851572212753,
+      "learning_rate": 1.2458302178110702e-07,
+      "loss": 0.7878,
+      "num_input_tokens_seen": 158056460,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8905188480731077,
+      "grad_norm": 2.099011793590973,
+      "learning_rate": 1.2431253131656118e-07,
+      "loss": 0.8178,
+      "num_input_tokens_seen": 158075655,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8906390909637467,
+      "grad_norm": 2.7639861033090947,
+      "learning_rate": 1.240423253923133e-07,
+      "loss": 0.7528,
+      "num_input_tokens_seen": 158094980,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8907593338543859,
+      "grad_norm": 3.278476556308301,
+      "learning_rate": 1.237724040493533e-07,
+      "loss": 0.6816,
+      "num_input_tokens_seen": 158113325,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8908795767450249,
+      "grad_norm": 2.9028990544002333,
+      "learning_rate": 1.2350276732862773e-07,
+      "loss": 0.7308,
+      "num_input_tokens_seen": 158134070,
+      "step": 7409
+    },
+    {
+      "epoch": 0.890999819635664,
+      "grad_norm": 0.8592370892965188,
+      "learning_rate": 1.2323341527103993e-07,
+      "loss": 0.6169,
+      "num_input_tokens_seen": 158188990,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8911200625263032,
+      "grad_norm": 5.5175932348302235,
+      "learning_rate": 1.2296434791745135e-07,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 158207160,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8912403054169422,
+      "grad_norm": 4.249353177600977,
+      "learning_rate": 1.2269556530867875e-07,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 158225435,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8913605483075813,
+      "grad_norm": 2.3149239649135134,
+      "learning_rate": 1.2242706748549614e-07,
+      "loss": 0.815,
+      "num_input_tokens_seen": 158243150,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8914807911982204,
+      "grad_norm": 2.452027937539289,
+      "learning_rate": 1.2215885448863428e-07,
+      "loss": 0.8135,
+      "num_input_tokens_seen": 158263745,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8916010340888595,
+      "grad_norm": 2.0436152163925168,
+      "learning_rate": 1.2189092635878152e-07,
+      "loss": 0.7981,
+      "num_input_tokens_seen": 158284915,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8917212769794985,
+      "grad_norm": 1.8952763475669507,
+      "learning_rate": 1.21623283136582e-07,
+      "loss": 0.7725,
+      "num_input_tokens_seen": 158303580,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8918415198701377,
+      "grad_norm": 2.2136266171035612,
+      "learning_rate": 1.2135592486263678e-07,
+      "loss": 0.8058,
+      "num_input_tokens_seen": 158322550,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8919617627607768,
+      "grad_norm": 1.7865943064220071,
+      "learning_rate": 1.2108885157750415e-07,
+      "loss": 0.6047,
+      "num_input_tokens_seen": 158344630,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8920820056514158,
+      "grad_norm": 1.80477949447005,
+      "learning_rate": 1.2082206332169897e-07,
+      "loss": 0.7949,
+      "num_input_tokens_seen": 158364445,
+      "step": 7419
+    },
+    {
+      "epoch": 0.892202248542055,
+      "grad_norm": 2.4332201211360287,
+      "learning_rate": 1.2055556013569246e-07,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 158379675,
+      "step": 7420
+    },
+    {
+      "epoch": 0.892322491432694,
+      "grad_norm": 1.72965925276919,
+      "learning_rate": 1.2028934205991315e-07,
+      "loss": 0.8071,
+      "num_input_tokens_seen": 158398715,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8924427343233331,
+      "grad_norm": 1.7336509221232885,
+      "learning_rate": 1.2002340913474607e-07,
+      "loss": 0.7608,
+      "num_input_tokens_seen": 158422070,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8925629772139723,
+      "grad_norm": 2.078205066315998,
+      "learning_rate": 1.1975776140053317e-07,
+      "loss": 0.736,
+      "num_input_tokens_seen": 158441760,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8926832201046113,
+      "grad_norm": 2.6634472117488865,
+      "learning_rate": 1.194923988975729e-07,
+      "loss": 0.7294,
+      "num_input_tokens_seen": 158461080,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8928034629952504,
+      "grad_norm": 2.977311174455832,
+      "learning_rate": 1.192273216661206e-07,
+      "loss": 0.7316,
+      "num_input_tokens_seen": 158478890,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8929237058858895,
+      "grad_norm": 0.779681279661497,
+      "learning_rate": 1.1896252974638787e-07,
+      "loss": 0.6079,
+      "num_input_tokens_seen": 158540300,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8930439487765286,
+      "grad_norm": 2.592134327599314,
+      "learning_rate": 1.1869802317854394e-07,
+      "loss": 0.79,
+      "num_input_tokens_seen": 158563805,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8931641916671677,
+      "grad_norm": 2.0683654205797732,
+      "learning_rate": 1.1843380200271425e-07,
+      "loss": 0.7218,
+      "num_input_tokens_seen": 158582725,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8932844345578068,
+      "grad_norm": 2.2712199177089634,
+      "learning_rate": 1.1816986625898073e-07,
+      "loss": 0.7995,
+      "num_input_tokens_seen": 158602030,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8934046774484459,
+      "grad_norm": 2.043293741541181,
+      "learning_rate": 1.1790621598738204e-07,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 158620065,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8935249203390849,
+      "grad_norm": 2.3460672567136887,
+      "learning_rate": 1.176428512279144e-07,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 158640505,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8936451632297241,
+      "grad_norm": 2.142117129832284,
+      "learning_rate": 1.173797720205294e-07,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 158658260,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8937654061203631,
+      "grad_norm": 4.410563840176427,
+      "learning_rate": 1.1711697840513646e-07,
+      "loss": 0.7188,
+      "num_input_tokens_seen": 158677415,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8938856490110022,
+      "grad_norm": 2.981879446490988,
+      "learning_rate": 1.1685447042160012e-07,
+      "loss": 0.6999,
+      "num_input_tokens_seen": 158695170,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8940058919016414,
+      "grad_norm": 2.0064303955234886,
+      "learning_rate": 1.1659224810974367e-07,
+      "loss": 0.7063,
+      "num_input_tokens_seen": 158714850,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8941261347922804,
+      "grad_norm": 1.6583843680478008,
+      "learning_rate": 1.1633031150934591e-07,
+      "loss": 0.6771,
+      "num_input_tokens_seen": 158737600,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8942463776829195,
+      "grad_norm": 2.4075523325168495,
+      "learning_rate": 1.1606866066014199e-07,
+      "loss": 0.7962,
+      "num_input_tokens_seen": 158756370,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8943666205735585,
+      "grad_norm": 2.4275949545876614,
+      "learning_rate": 1.1580729560182412e-07,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 158771945,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8944868634641977,
+      "grad_norm": 2.4113990118787743,
+      "learning_rate": 1.1554621637404171e-07,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 158789755,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8946071063548368,
+      "grad_norm": 2.1845927393315874,
+      "learning_rate": 1.1528542301639999e-07,
+      "loss": 0.6025,
+      "num_input_tokens_seen": 158806265,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8947273492454758,
+      "grad_norm": 3.303139965580428,
+      "learning_rate": 1.1502491556846105e-07,
+      "loss": 0.8148,
+      "num_input_tokens_seen": 158824480,
+      "step": 7441
+    },
+    {
+      "epoch": 0.894847592136115,
+      "grad_norm": 2.746326716284547,
+      "learning_rate": 1.1476469406974331e-07,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 158839800,
+      "step": 7442
+    },
+    {
+      "epoch": 0.894967835026754,
+      "grad_norm": 2.1581172915131446,
+      "learning_rate": 1.1450475855972297e-07,
+      "loss": 0.7645,
+      "num_input_tokens_seen": 158860310,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8950880779173931,
+      "grad_norm": 2.127983045321389,
+      "learning_rate": 1.1424510907783158e-07,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 158877310,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8952083208080323,
+      "grad_norm": 1.6929488287054206,
+      "learning_rate": 1.1398574566345787e-07,
+      "loss": 0.8171,
+      "num_input_tokens_seen": 158897665,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8953285636986713,
+      "grad_norm": 2.2375815494616984,
+      "learning_rate": 1.1372666835594702e-07,
+      "loss": 0.8229,
+      "num_input_tokens_seen": 158915710,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8954488065893104,
+      "grad_norm": 2.9188268774038146,
+      "learning_rate": 1.1346787719460071e-07,
+      "loss": 0.7136,
+      "num_input_tokens_seen": 158934315,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8955690494799495,
+      "grad_norm": 1.961671483476635,
+      "learning_rate": 1.1320937221867732e-07,
+      "loss": 0.7241,
+      "num_input_tokens_seen": 158951615,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8956892923705886,
+      "grad_norm": 1.8541193162208953,
+      "learning_rate": 1.1295115346739215e-07,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 158971335,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8958095352612276,
+      "grad_norm": 2.5213796887906352,
+      "learning_rate": 1.1269322097991629e-07,
+      "loss": 0.7333,
+      "num_input_tokens_seen": 158994340,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8959297781518668,
+      "grad_norm": 2.2469988960220384,
+      "learning_rate": 1.1243557479537869e-07,
+      "loss": 0.679,
+      "num_input_tokens_seen": 159013950,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8960500210425059,
+      "grad_norm": 1.9917502640688571,
+      "learning_rate": 1.121782149528634e-07,
+      "loss": 0.6835,
+      "num_input_tokens_seen": 159030770,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8961702639331449,
+      "grad_norm": 2.165014867472244,
+      "learning_rate": 1.1192114149141208e-07,
+      "loss": 0.7872,
+      "num_input_tokens_seen": 159050125,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8962905068237841,
+      "grad_norm": 3.1238865484205225,
+      "learning_rate": 1.1166435445002197e-07,
+      "loss": 0.6501,
+      "num_input_tokens_seen": 159067515,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8964107497144231,
+      "grad_norm": 3.1551304464595242,
+      "learning_rate": 1.1140785386764818e-07,
+      "loss": 0.6768,
+      "num_input_tokens_seen": 159085935,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8965309926050622,
+      "grad_norm": 2.2656326700830607,
+      "learning_rate": 1.1115163978320153e-07,
+      "loss": 0.6961,
+      "num_input_tokens_seen": 159104385,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8966512354957014,
+      "grad_norm": 2.098923854081124,
+      "learning_rate": 1.1089571223554917e-07,
+      "loss": 0.8196,
+      "num_input_tokens_seen": 159124990,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8967714783863404,
+      "grad_norm": 1.821318755187835,
+      "learning_rate": 1.1064007126351494e-07,
+      "loss": 0.8347,
+      "num_input_tokens_seen": 159145425,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8968917212769795,
+      "grad_norm": 2.150388695466973,
+      "learning_rate": 1.1038471690588003e-07,
+      "loss": 0.753,
+      "num_input_tokens_seen": 159164290,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8970119641676186,
+      "grad_norm": 1.9845461245954064,
+      "learning_rate": 1.1012964920138124e-07,
+      "loss": 0.7941,
+      "num_input_tokens_seen": 159183595,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8971322070582577,
+      "grad_norm": 2.8961545752178472,
+      "learning_rate": 1.0987486818871205e-07,
+      "loss": 0.7521,
+      "num_input_tokens_seen": 159206905,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8972524499488967,
+      "grad_norm": 6.701494773359898,
+      "learning_rate": 1.0962037390652245e-07,
+      "loss": 0.7246,
+      "num_input_tokens_seen": 159225645,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8973726928395359,
+      "grad_norm": 2.24292520802731,
+      "learning_rate": 1.0936616639341911e-07,
+      "loss": 0.719,
+      "num_input_tokens_seen": 159245655,
+      "step": 7463
+    },
+    {
+      "epoch": 0.897492935730175,
+      "grad_norm": 0.9956366844433466,
+      "learning_rate": 1.0911224568796496e-07,
+      "loss": 0.5653,
+      "num_input_tokens_seen": 159303570,
+      "step": 7464
+    },
+    {
+      "epoch": 0.897613178620814,
+      "grad_norm": 2.110120525028927,
+      "learning_rate": 1.0885861182867984e-07,
+      "loss": 0.7025,
+      "num_input_tokens_seen": 159321395,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8977334215114532,
+      "grad_norm": 2.1731514485303545,
+      "learning_rate": 1.0860526485403942e-07,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 159342390,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8978536644020922,
+      "grad_norm": 1.7461191273470114,
+      "learning_rate": 1.0835220480247675e-07,
+      "loss": 0.7707,
+      "num_input_tokens_seen": 159360605,
+      "step": 7467
+    },
+    {
+      "epoch": 0.8979739072927313,
+      "grad_norm": 2.5862737864272782,
+      "learning_rate": 1.0809943171238067e-07,
+      "loss": 0.8323,
+      "num_input_tokens_seen": 159378250,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8980941501833704,
+      "grad_norm": 2.5868554760031923,
+      "learning_rate": 1.078469456220965e-07,
+      "loss": 0.6342,
+      "num_input_tokens_seen": 159398125,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8982143930740095,
+      "grad_norm": 2.414086964815238,
+      "learning_rate": 1.0759474656992584e-07,
+      "loss": 0.6861,
+      "num_input_tokens_seen": 159420615,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8983346359646486,
+      "grad_norm": 2.588448483697621,
+      "learning_rate": 1.0734283459412785e-07,
+      "loss": 0.7756,
+      "num_input_tokens_seen": 159437185,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8984548788552876,
+      "grad_norm": 1.870896329250706,
+      "learning_rate": 1.0709120973291707e-07,
+      "loss": 0.7982,
+      "num_input_tokens_seen": 159456685,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8985751217459268,
+      "grad_norm": 3.038158459974717,
+      "learning_rate": 1.0683987202446475e-07,
+      "loss": 0.7763,
+      "num_input_tokens_seen": 159474590,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8986953646365659,
+      "grad_norm": 2.0296187293502714,
+      "learning_rate": 1.0658882150689841e-07,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 159493170,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8988156075272049,
+      "grad_norm": 3.680987775662778,
+      "learning_rate": 1.0633805821830243e-07,
+      "loss": 0.7767,
+      "num_input_tokens_seen": 159509575,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8989358504178441,
+      "grad_norm": 2.6005847246806426,
+      "learning_rate": 1.0608758219671798e-07,
+      "loss": 0.8269,
+      "num_input_tokens_seen": 159528335,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8990560933084831,
+      "grad_norm": 1.7060250119074871,
+      "learning_rate": 1.0583739348014087e-07,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 159549140,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8991763361991222,
+      "grad_norm": 4.536282916458728,
+      "learning_rate": 1.0558749210652518e-07,
+      "loss": 0.841,
+      "num_input_tokens_seen": 159568790,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8992965790897613,
+      "grad_norm": 1.6573231953673202,
+      "learning_rate": 1.053378781137808e-07,
+      "loss": 0.8475,
+      "num_input_tokens_seen": 159589430,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8994168219804004,
+      "grad_norm": 1.862817181323433,
+      "learning_rate": 1.0508855153977392e-07,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 159605615,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8995370648710395,
+      "grad_norm": 2.860157491264734,
+      "learning_rate": 1.0483951242232714e-07,
+      "loss": 0.6626,
+      "num_input_tokens_seen": 159625810,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8996573077616786,
+      "grad_norm": 1.0992852768122545,
+      "learning_rate": 1.0459076079921913e-07,
+      "loss": 0.6085,
+      "num_input_tokens_seen": 159678190,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8997775506523177,
+      "grad_norm": 2.191421284110555,
+      "learning_rate": 1.0434229670818618e-07,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 159694585,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8998977935429567,
+      "grad_norm": 2.6993337303124676,
+      "learning_rate": 1.0409412018691944e-07,
+      "loss": 0.7942,
+      "num_input_tokens_seen": 159714770,
+      "step": 7484
+    },
+    {
+      "epoch": 0.9000180364335959,
+      "grad_norm": 2.538244067681212,
+      "learning_rate": 1.0384623127306724e-07,
+      "loss": 0.7521,
+      "num_input_tokens_seen": 159731835,
+      "step": 7485
+    },
+    {
+      "epoch": 0.900138279324235,
+      "grad_norm": 1.8307027562518343,
+      "learning_rate": 1.0359863000423397e-07,
+      "loss": 0.7813,
+      "num_input_tokens_seen": 159749690,
+      "step": 7486
+    },
+    {
+      "epoch": 0.900258522214874,
+      "grad_norm": 1.7498627808467078,
+      "learning_rate": 1.0335131641798089e-07,
+      "loss": 0.7109,
+      "num_input_tokens_seen": 159771370,
+      "step": 7487
+    },
+    {
+      "epoch": 0.9003787651055132,
+      "grad_norm": 0.8594861350953386,
+      "learning_rate": 1.0310429055182512e-07,
+      "loss": 0.6277,
+      "num_input_tokens_seen": 159825410,
+      "step": 7488
+    },
+    {
+      "epoch": 0.9004990079961522,
+      "grad_norm": 2.2680041214321682,
+      "learning_rate": 1.0285755244324024e-07,
+      "loss": 0.7254,
+      "num_input_tokens_seen": 159845875,
+      "step": 7489
+    },
+    {
+      "epoch": 0.9006192508867913,
+      "grad_norm": 1.5113856022716663,
+      "learning_rate": 1.0261110212965629e-07,
+      "loss": 0.6827,
+      "num_input_tokens_seen": 159867390,
+      "step": 7490
+    },
+    {
+      "epoch": 0.9007394937774305,
+      "grad_norm": 3.054985057124256,
+      "learning_rate": 1.023649396484596e-07,
+      "loss": 0.7891,
+      "num_input_tokens_seen": 159886165,
+      "step": 7491
+    },
+    {
+      "epoch": 0.9008597366680695,
+      "grad_norm": 2.624098147592181,
+      "learning_rate": 1.0211906503699275e-07,
+      "loss": 0.6699,
+      "num_input_tokens_seen": 159908860,
+      "step": 7492
+    },
+    {
+      "epoch": 0.9009799795587086,
+      "grad_norm": 3.0606136626489064,
+      "learning_rate": 1.0187347833255477e-07,
+      "loss": 0.8175,
+      "num_input_tokens_seen": 159924485,
+      "step": 7493
+    },
+    {
+      "epoch": 0.9011002224493477,
+      "grad_norm": 1.8862268761230878,
+      "learning_rate": 1.0162817957240056e-07,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 159944100,
+      "step": 7494
+    },
+    {
+      "epoch": 0.9012204653399868,
+      "grad_norm": 0.9996754349651223,
+      "learning_rate": 1.0138316879374253e-07,
+      "loss": 0.6832,
+      "num_input_tokens_seen": 160013110,
+      "step": 7495
+    },
+    {
+      "epoch": 0.9013407082306258,
+      "grad_norm": 2.3090022248454027,
+      "learning_rate": 1.0113844603374833e-07,
+      "loss": 0.7294,
+      "num_input_tokens_seen": 160029355,
+      "step": 7496
+    },
+    {
+      "epoch": 0.901460951121265,
+      "grad_norm": 2.7184849437388854,
+      "learning_rate": 1.0089401132954178e-07,
+      "loss": 0.713,
+      "num_input_tokens_seen": 160047055,
+      "step": 7497
+    },
+    {
+      "epoch": 0.9015811940119041,
+      "grad_norm": 3.1639211591771494,
+      "learning_rate": 1.006498647182037e-07,
+      "loss": 0.7247,
+      "num_input_tokens_seen": 160065430,
+      "step": 7498
+    },
+    {
+      "epoch": 0.9017014369025431,
+      "grad_norm": 2.515584669544843,
+      "learning_rate": 1.004060062367713e-07,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 160086245,
+      "step": 7499
+    },
+    {
+      "epoch": 0.9018216797931822,
+      "grad_norm": 2.207496324924907,
+      "learning_rate": 1.0016243592223728e-07,
+      "loss": 0.6892,
+      "num_input_tokens_seen": 160106365,
+      "step": 7500
+    },
+    {
+      "epoch": 0.9019419226838213,
+      "grad_norm": 2.0020856950507615,
+      "learning_rate": 9.991915381155114e-08,
+      "loss": 0.6456,
+      "num_input_tokens_seen": 160129065,
+      "step": 7501
+    },
+    {
+      "epoch": 0.9020621655744604,
+      "grad_norm": 6.1518308439457785,
+      "learning_rate": 9.967615994161849e-08,
+      "loss": 0.7562,
+      "num_input_tokens_seen": 160148445,
+      "step": 7502
+    },
+    {
+      "epoch": 0.9021824084650995,
+      "grad_norm": 2.881232592612134,
+      "learning_rate": 9.943345434930161e-08,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 160168415,
+      "step": 7503
+    },
+    {
+      "epoch": 0.9023026513557386,
+      "grad_norm": 2.3371083590113426,
+      "learning_rate": 9.919103707141862e-08,
+      "loss": 0.6815,
+      "num_input_tokens_seen": 160187015,
+      "step": 7504
+    },
+    {
+      "epoch": 0.9024228942463777,
+      "grad_norm": 15.34603127945343,
+      "learning_rate": 9.89489081447441e-08,
+      "loss": 0.7644,
+      "num_input_tokens_seen": 160203935,
+      "step": 7505
+    },
+    {
+      "epoch": 0.9025431371370167,
+      "grad_norm": 2.2168655036174623,
+      "learning_rate": 9.870706760600844e-08,
+      "loss": 0.8242,
+      "num_input_tokens_seen": 160223605,
+      "step": 7506
+    },
+    {
+      "epoch": 0.9026633800276559,
+      "grad_norm": 2.6545732032829754,
+      "learning_rate": 9.846551549189918e-08,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 160242930,
+      "step": 7507
+    },
+    {
+      "epoch": 0.902783622918295,
+      "grad_norm": 2.516511140837101,
+      "learning_rate": 9.822425183905925e-08,
+      "loss": 0.6799,
+      "num_input_tokens_seen": 160263175,
+      "step": 7508
+    },
+    {
+      "epoch": 0.902903865808934,
+      "grad_norm": 0.9948757554707756,
+      "learning_rate": 9.798327668408823e-08,
+      "loss": 0.7757,
+      "num_input_tokens_seen": 160324530,
+      "step": 7509
+    },
+    {
+      "epoch": 0.9030241086995732,
+      "grad_norm": 3.2401104748490774,
+      "learning_rate": 9.774259006354158e-08,
+      "loss": 0.6864,
+      "num_input_tokens_seen": 160344320,
+      "step": 7510
+    },
+    {
+      "epoch": 0.9031443515902122,
+      "grad_norm": 2.502940523264334,
+      "learning_rate": 9.750219201393184e-08,
+      "loss": 0.753,
+      "num_input_tokens_seen": 160364005,
+      "step": 7511
+    },
+    {
+      "epoch": 0.9032645944808513,
+      "grad_norm": 1.8601252712663567,
+      "learning_rate": 9.726208257172697e-08,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 160385420,
+      "step": 7512
+    },
+    {
+      "epoch": 0.9033848373714904,
+      "grad_norm": 2.3912634199003127,
+      "learning_rate": 9.702226177335115e-08,
+      "loss": 0.7459,
+      "num_input_tokens_seen": 160403635,
+      "step": 7513
+    },
+    {
+      "epoch": 0.9035050802621295,
+      "grad_norm": 1.7436473053112613,
+      "learning_rate": 9.678272965518508e-08,
+      "loss": 0.719,
+      "num_input_tokens_seen": 160424640,
+      "step": 7514
+    },
+    {
+      "epoch": 0.9036253231527686,
+      "grad_norm": 2.7474470451241904,
+      "learning_rate": 9.65434862535659e-08,
+      "loss": 0.6687,
+      "num_input_tokens_seen": 160443730,
+      "step": 7515
+    },
+    {
+      "epoch": 0.9037455660434077,
+      "grad_norm": 4.535659387411721,
+      "learning_rate": 9.630453160478635e-08,
+      "loss": 0.6416,
+      "num_input_tokens_seen": 160458805,
+      "step": 7516
+    },
+    {
+      "epoch": 0.9038658089340468,
+      "grad_norm": 1.969192699228269,
+      "learning_rate": 9.60658657450959e-08,
+      "loss": 0.8143,
+      "num_input_tokens_seen": 160478825,
+      "step": 7517
+    },
+    {
+      "epoch": 0.9039860518246858,
+      "grad_norm": 1.802472414416805,
+      "learning_rate": 9.582748871069957e-08,
+      "loss": 0.7881,
+      "num_input_tokens_seen": 160497985,
+      "step": 7518
+    },
+    {
+      "epoch": 0.904106294715325,
+      "grad_norm": 2.0273494435806922,
+      "learning_rate": 9.558940053775954e-08,
+      "loss": 0.8164,
+      "num_input_tokens_seen": 160516345,
+      "step": 7519
+    },
+    {
+      "epoch": 0.904226537605964,
+      "grad_norm": 2.229178075447839,
+      "learning_rate": 9.535160126239317e-08,
+      "loss": 0.6803,
+      "num_input_tokens_seen": 160532690,
+      "step": 7520
+    },
+    {
+      "epoch": 0.9043467804966031,
+      "grad_norm": 1.6338859475216079,
+      "learning_rate": 9.511409092067446e-08,
+      "loss": 0.7059,
+      "num_input_tokens_seen": 160552765,
+      "step": 7521
+    },
+    {
+      "epoch": 0.9044670233872423,
+      "grad_norm": 1.9095735221204646,
+      "learning_rate": 9.487686954863327e-08,
+      "loss": 0.6673,
+      "num_input_tokens_seen": 160572205,
+      "step": 7522
+    },
+    {
+      "epoch": 0.9045872662778813,
+      "grad_norm": 2.2797108697620443,
+      "learning_rate": 9.46399371822566e-08,
+      "loss": 0.7589,
+      "num_input_tokens_seen": 160591700,
+      "step": 7523
+    },
+    {
+      "epoch": 0.9047075091685204,
+      "grad_norm": 4.848339218755644,
+      "learning_rate": 9.440329385748657e-08,
+      "loss": 0.7122,
+      "num_input_tokens_seen": 160608490,
+      "step": 7524
+    },
+    {
+      "epoch": 0.9048277520591596,
+      "grad_norm": 1.8366682425941812,
+      "learning_rate": 9.41669396102216e-08,
+      "loss": 0.7065,
+      "num_input_tokens_seen": 160626460,
+      "step": 7525
+    },
+    {
+      "epoch": 0.9049479949497986,
+      "grad_norm": 2.1311443679812627,
+      "learning_rate": 9.393087447631631e-08,
+      "loss": 0.7696,
+      "num_input_tokens_seen": 160644460,
+      "step": 7526
+    },
+    {
+      "epoch": 0.9050682378404377,
+      "grad_norm": 1.9598539362045724,
+      "learning_rate": 9.36950984915823e-08,
+      "loss": 0.7187,
+      "num_input_tokens_seen": 160662535,
+      "step": 7527
+    },
+    {
+      "epoch": 0.9051884807310768,
+      "grad_norm": 2.1223043316954446,
+      "learning_rate": 9.345961169178607e-08,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 160681940,
+      "step": 7528
+    },
+    {
+      "epoch": 0.9053087236217159,
+      "grad_norm": 1.568713773382654,
+      "learning_rate": 9.322441411265081e-08,
+      "loss": 0.7197,
+      "num_input_tokens_seen": 160702645,
+      "step": 7529
+    },
+    {
+      "epoch": 0.9054289665123549,
+      "grad_norm": 2.146973649963694,
+      "learning_rate": 9.298950578985554e-08,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 160719440,
+      "step": 7530
+    },
+    {
+      "epoch": 0.905549209402994,
+      "grad_norm": 6.157937870752905,
+      "learning_rate": 9.275488675903642e-08,
+      "loss": 0.7066,
+      "num_input_tokens_seen": 160738105,
+      "step": 7531
+    },
+    {
+      "epoch": 0.9056694522936332,
+      "grad_norm": 2.1803037634972084,
+      "learning_rate": 9.252055705578454e-08,
+      "loss": 0.7271,
+      "num_input_tokens_seen": 160757325,
+      "step": 7532
+    },
+    {
+      "epoch": 0.9057896951842722,
+      "grad_norm": 1.7875094097835558,
+      "learning_rate": 9.228651671564747e-08,
+      "loss": 0.7097,
+      "num_input_tokens_seen": 160779075,
+      "step": 7533
+    },
+    {
+      "epoch": 0.9059099380749113,
+      "grad_norm": 1.6822523971466088,
+      "learning_rate": 9.205276577412901e-08,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 160801575,
+      "step": 7534
+    },
+    {
+      "epoch": 0.9060301809655504,
+      "grad_norm": 4.3546995619916595,
+      "learning_rate": 9.181930426668905e-08,
+      "loss": 0.7557,
+      "num_input_tokens_seen": 160818090,
+      "step": 7535
+    },
+    {
+      "epoch": 0.9061504238561895,
+      "grad_norm": 1.7066634306694526,
+      "learning_rate": 9.158613222874346e-08,
+      "loss": 0.6745,
+      "num_input_tokens_seen": 160839435,
+      "step": 7536
+    },
+    {
+      "epoch": 0.9062706667468285,
+      "grad_norm": 1.7270450184082642,
+      "learning_rate": 9.135324969566416e-08,
+      "loss": 0.8138,
+      "num_input_tokens_seen": 160858655,
+      "step": 7537
+    },
+    {
+      "epoch": 0.9063909096374677,
+      "grad_norm": 2.2086411011526232,
+      "learning_rate": 9.112065670277913e-08,
+      "loss": 0.7436,
+      "num_input_tokens_seen": 160874740,
+      "step": 7538
+    },
+    {
+      "epoch": 0.9065111525281068,
+      "grad_norm": 2.0603104100715526,
+      "learning_rate": 9.088835328537303e-08,
+      "loss": 0.7189,
+      "num_input_tokens_seen": 160896050,
+      "step": 7539
+    },
+    {
+      "epoch": 0.9066313954187458,
+      "grad_norm": 3.1993694370955117,
+      "learning_rate": 9.065633947868568e-08,
+      "loss": 0.7104,
+      "num_input_tokens_seen": 160915375,
+      "step": 7540
+    },
+    {
+      "epoch": 0.906751638309385,
+      "grad_norm": 11.465692091360646,
+      "learning_rate": 9.042461531791379e-08,
+      "loss": 0.7926,
+      "num_input_tokens_seen": 160933515,
+      "step": 7541
+    },
+    {
+      "epoch": 0.906871881200024,
+      "grad_norm": 1.867101866808219,
+      "learning_rate": 9.019318083820903e-08,
+      "loss": 0.7713,
+      "num_input_tokens_seen": 160951815,
+      "step": 7542
+    },
+    {
+      "epoch": 0.9069921240906631,
+      "grad_norm": 2.134552912810663,
+      "learning_rate": 8.996203607468045e-08,
+      "loss": 0.8428,
+      "num_input_tokens_seen": 160970535,
+      "step": 7543
+    },
+    {
+      "epoch": 0.9071123669813023,
+      "grad_norm": 1.776671957933852,
+      "learning_rate": 8.973118106239241e-08,
+      "loss": 0.7514,
+      "num_input_tokens_seen": 160992860,
+      "step": 7544
+    },
+    {
+      "epoch": 0.9072326098719413,
+      "grad_norm": 2.1123227885970226,
+      "learning_rate": 8.95006158363656e-08,
+      "loss": 0.94,
+      "num_input_tokens_seen": 161012765,
+      "step": 7545
+    },
+    {
+      "epoch": 0.9073528527625804,
+      "grad_norm": 5.205004338763424,
+      "learning_rate": 8.927034043157577e-08,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 161031575,
+      "step": 7546
+    },
+    {
+      "epoch": 0.9074730956532195,
+      "grad_norm": 2.5892651612581736,
+      "learning_rate": 8.904035488295658e-08,
+      "loss": 0.7298,
+      "num_input_tokens_seen": 161050795,
+      "step": 7547
+    },
+    {
+      "epoch": 0.9075933385438586,
+      "grad_norm": 0.7162675044292636,
+      "learning_rate": 8.881065922539588e-08,
+      "loss": 0.5642,
+      "num_input_tokens_seen": 161110955,
+      "step": 7548
+    },
+    {
+      "epoch": 0.9077135814344977,
+      "grad_norm": 1.9819170542957976,
+      "learning_rate": 8.85812534937389e-08,
+      "loss": 0.7366,
+      "num_input_tokens_seen": 161128775,
+      "step": 7549
+    },
+    {
+      "epoch": 0.9078338243251368,
+      "grad_norm": 2.641616198754066,
+      "learning_rate": 8.835213772278583e-08,
+      "loss": 0.6626,
+      "num_input_tokens_seen": 161145350,
+      "step": 7550
+    },
+    {
+      "epoch": 0.9079540672157759,
+      "grad_norm": 1.9311448888048481,
+      "learning_rate": 8.812331194729373e-08,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 161164715,
+      "step": 7551
+    },
+    {
+      "epoch": 0.9080743101064149,
+      "grad_norm": 2.3363297893957315,
+      "learning_rate": 8.789477620197484e-08,
+      "loss": 0.7186,
+      "num_input_tokens_seen": 161183960,
+      "step": 7552
+    },
+    {
+      "epoch": 0.9081945529970541,
+      "grad_norm": 9.535324200458785,
+      "learning_rate": 8.766653052149831e-08,
+      "loss": 0.7884,
+      "num_input_tokens_seen": 161198865,
+      "step": 7553
+    },
+    {
+      "epoch": 0.9083147958876931,
+      "grad_norm": 2.1869030122515847,
+      "learning_rate": 8.743857494048823e-08,
+      "loss": 0.7364,
+      "num_input_tokens_seen": 161215400,
+      "step": 7554
+    },
+    {
+      "epoch": 0.9084350387783322,
+      "grad_norm": 2.4098919368440184,
+      "learning_rate": 8.721090949352605e-08,
+      "loss": 0.6317,
+      "num_input_tokens_seen": 161231360,
+      "step": 7555
+    },
+    {
+      "epoch": 0.9085552816689714,
+      "grad_norm": 2.3997348097620343,
+      "learning_rate": 8.698353421514793e-08,
+      "loss": 0.7227,
+      "num_input_tokens_seen": 161249455,
+      "step": 7556
+    },
+    {
+      "epoch": 0.9086755245596104,
+      "grad_norm": 6.159752386165655,
+      "learning_rate": 8.67564491398467e-08,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 161266180,
+      "step": 7557
+    },
+    {
+      "epoch": 0.9087957674502495,
+      "grad_norm": 2.521642549208355,
+      "learning_rate": 8.652965430207082e-08,
+      "loss": 0.7367,
+      "num_input_tokens_seen": 161283805,
+      "step": 7558
+    },
+    {
+      "epoch": 0.9089160103408886,
+      "grad_norm": 2.690236497598766,
+      "learning_rate": 8.630314973622521e-08,
+      "loss": 0.6493,
+      "num_input_tokens_seen": 161301070,
+      "step": 7559
+    },
+    {
+      "epoch": 0.9090362532315277,
+      "grad_norm": 2.8108221454400737,
+      "learning_rate": 8.607693547666995e-08,
+      "loss": 0.7066,
+      "num_input_tokens_seen": 161323330,
+      "step": 7560
+    },
+    {
+      "epoch": 0.9091564961221668,
+      "grad_norm": 0.9286950482000047,
+      "learning_rate": 8.585101155772201e-08,
+      "loss": 0.625,
+      "num_input_tokens_seen": 161385170,
+      "step": 7561
+    },
+    {
+      "epoch": 0.9092767390128058,
+      "grad_norm": 1.9051620376967369,
+      "learning_rate": 8.562537801365354e-08,
+      "loss": 0.6815,
+      "num_input_tokens_seen": 161404625,
+      "step": 7562
+    },
+    {
+      "epoch": 0.909396981903445,
+      "grad_norm": 1.8887367327230244,
+      "learning_rate": 8.540003487869362e-08,
+      "loss": 0.6924,
+      "num_input_tokens_seen": 161426015,
+      "step": 7563
+    },
+    {
+      "epoch": 0.909517224794084,
+      "grad_norm": 2.272450348568467,
+      "learning_rate": 8.51749821870258e-08,
+      "loss": 0.7888,
+      "num_input_tokens_seen": 161443665,
+      "step": 7564
+    },
+    {
+      "epoch": 0.9096374676847231,
+      "grad_norm": 2.3608990691186404,
+      "learning_rate": 8.495021997279073e-08,
+      "loss": 0.6935,
+      "num_input_tokens_seen": 161461410,
+      "step": 7565
+    },
+    {
+      "epoch": 0.9097577105753623,
+      "grad_norm": 2.2850992885359744,
+      "learning_rate": 8.472574827008428e-08,
+      "loss": 0.6565,
+      "num_input_tokens_seen": 161482015,
+      "step": 7566
+    },
+    {
+      "epoch": 0.9098779534660013,
+      "grad_norm": 1.6910648294287771,
+      "learning_rate": 8.450156711295942e-08,
+      "loss": 0.8321,
+      "num_input_tokens_seen": 161501905,
+      "step": 7567
+    },
+    {
+      "epoch": 0.9099981963566404,
+      "grad_norm": 2.382902294918769,
+      "learning_rate": 8.427767653542383e-08,
+      "loss": 0.8612,
+      "num_input_tokens_seen": 161516795,
+      "step": 7568
+    },
+    {
+      "epoch": 0.9101184392472795,
+      "grad_norm": 2.672725355428453,
+      "learning_rate": 8.405407657144148e-08,
+      "loss": 0.7011,
+      "num_input_tokens_seen": 161535675,
+      "step": 7569
+    },
+    {
+      "epoch": 0.9102386821379186,
+      "grad_norm": 2.1751604996742224,
+      "learning_rate": 8.38307672549321e-08,
+      "loss": 0.7159,
+      "num_input_tokens_seen": 161552715,
+      "step": 7570
+    },
+    {
+      "epoch": 0.9103589250285576,
+      "grad_norm": 2.1264823940838316,
+      "learning_rate": 8.360774861977216e-08,
+      "loss": 0.6738,
+      "num_input_tokens_seen": 161571555,
+      "step": 7571
+    },
+    {
+      "epoch": 0.9104791679191968,
+      "grad_norm": 2.160953622723792,
+      "learning_rate": 8.338502069979281e-08,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 161591585,
+      "step": 7572
+    },
+    {
+      "epoch": 0.9105994108098359,
+      "grad_norm": 4.216202780987863,
+      "learning_rate": 8.316258352878214e-08,
+      "loss": 0.7924,
+      "num_input_tokens_seen": 161607725,
+      "step": 7573
+    },
+    {
+      "epoch": 0.9107196537004749,
+      "grad_norm": 3.191875191782117,
+      "learning_rate": 8.294043714048315e-08,
+      "loss": 0.7023,
+      "num_input_tokens_seen": 161626525,
+      "step": 7574
+    },
+    {
+      "epoch": 0.9108398965911141,
+      "grad_norm": 0.8626935519681387,
+      "learning_rate": 8.271858156859602e-08,
+      "loss": 0.6451,
+      "num_input_tokens_seen": 161691615,
+      "step": 7575
+    },
+    {
+      "epoch": 0.9109601394817531,
+      "grad_norm": 2.8407299383022813,
+      "learning_rate": 8.249701684677557e-08,
+      "loss": 0.7335,
+      "num_input_tokens_seen": 161712660,
+      "step": 7576
+    },
+    {
+      "epoch": 0.9110803823723922,
+      "grad_norm": 2.0580438973069257,
+      "learning_rate": 8.227574300863294e-08,
+      "loss": 0.8051,
+      "num_input_tokens_seen": 161732550,
+      "step": 7577
+    },
+    {
+      "epoch": 0.9112006252630314,
+      "grad_norm": 2.019764434310867,
+      "learning_rate": 8.205476008773548e-08,
+      "loss": 0.7011,
+      "num_input_tokens_seen": 161756270,
+      "step": 7578
+    },
+    {
+      "epoch": 0.9113208681536704,
+      "grad_norm": 2.420110162429577,
+      "learning_rate": 8.183406811760596e-08,
+      "loss": 0.8228,
+      "num_input_tokens_seen": 161775720,
+      "step": 7579
+    },
+    {
+      "epoch": 0.9114411110443095,
+      "grad_norm": 1.5226931927080167,
+      "learning_rate": 8.161366713172313e-08,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 161797830,
+      "step": 7580
+    },
+    {
+      "epoch": 0.9115613539349486,
+      "grad_norm": 3.487394882747052,
+      "learning_rate": 8.13935571635218e-08,
+      "loss": 0.847,
+      "num_input_tokens_seen": 161812390,
+      "step": 7581
+    },
+    {
+      "epoch": 0.9116815968255877,
+      "grad_norm": 2.350631755582932,
+      "learning_rate": 8.117373824639196e-08,
+      "loss": 0.6951,
+      "num_input_tokens_seen": 161832375,
+      "step": 7582
+    },
+    {
+      "epoch": 0.9118018397162267,
+      "grad_norm": 0.7445561490452474,
+      "learning_rate": 8.095421041368067e-08,
+      "loss": 0.6092,
+      "num_input_tokens_seen": 161891510,
+      "step": 7583
+    },
+    {
+      "epoch": 0.9119220826068659,
+      "grad_norm": 2.45747923774838,
+      "learning_rate": 8.073497369868999e-08,
+      "loss": 0.7074,
+      "num_input_tokens_seen": 161909690,
+      "step": 7584
+    },
+    {
+      "epoch": 0.912042325497505,
+      "grad_norm": 2.0020318037415072,
+      "learning_rate": 8.051602813467772e-08,
+      "loss": 0.7447,
+      "num_input_tokens_seen": 161931265,
+      "step": 7585
+    },
+    {
+      "epoch": 0.912162568388144,
+      "grad_norm": 1.870569300231878,
+      "learning_rate": 8.029737375485756e-08,
+      "loss": 0.7106,
+      "num_input_tokens_seen": 161950215,
+      "step": 7586
+    },
+    {
+      "epoch": 0.9122828112787832,
+      "grad_norm": 2.7125430342196126,
+      "learning_rate": 8.007901059239986e-08,
+      "loss": 0.7243,
+      "num_input_tokens_seen": 161969215,
+      "step": 7587
+    },
+    {
+      "epoch": 0.9124030541694222,
+      "grad_norm": 1.7769672611276608,
+      "learning_rate": 7.986093868042964e-08,
+      "loss": 0.7967,
+      "num_input_tokens_seen": 161989180,
+      "step": 7588
+    },
+    {
+      "epoch": 0.9125232970600613,
+      "grad_norm": 2.133263869857011,
+      "learning_rate": 7.964315805202826e-08,
+      "loss": 0.6742,
+      "num_input_tokens_seen": 162009480,
+      "step": 7589
+    },
+    {
+      "epoch": 0.9126435399507005,
+      "grad_norm": 3.2160695358914064,
+      "learning_rate": 7.94256687402326e-08,
+      "loss": 0.7297,
+      "num_input_tokens_seen": 162028385,
+      "step": 7590
+    },
+    {
+      "epoch": 0.9127637828413395,
+      "grad_norm": 2.4259216170256592,
+      "learning_rate": 7.920847077803649e-08,
+      "loss": 0.6948,
+      "num_input_tokens_seen": 162045895,
+      "step": 7591
+    },
+    {
+      "epoch": 0.9128840257319786,
+      "grad_norm": 2.1342458075863804,
+      "learning_rate": 7.899156419838804e-08,
+      "loss": 0.8181,
+      "num_input_tokens_seen": 162064585,
+      "step": 7592
+    },
+    {
+      "epoch": 0.9130042686226177,
+      "grad_norm": 2.820724667989107,
+      "learning_rate": 7.87749490341918e-08,
+      "loss": 0.6505,
+      "num_input_tokens_seen": 162084580,
+      "step": 7593
+    },
+    {
+      "epoch": 0.9131245115132568,
+      "grad_norm": 2.0430614147947286,
+      "learning_rate": 7.855862531830836e-08,
+      "loss": 0.8324,
+      "num_input_tokens_seen": 162100410,
+      "step": 7594
+    },
+    {
+      "epoch": 0.9132447544038959,
+      "grad_norm": 2.894161402487206,
+      "learning_rate": 7.834259308355373e-08,
+      "loss": 0.7194,
+      "num_input_tokens_seen": 162118895,
+      "step": 7595
+    },
+    {
+      "epoch": 0.9133649972945349,
+      "grad_norm": 2.0246382779070777,
+      "learning_rate": 7.812685236269989e-08,
+      "loss": 0.7387,
+      "num_input_tokens_seen": 162137275,
+      "step": 7596
+    },
+    {
+      "epoch": 0.9134852401851741,
+      "grad_norm": 0.8816646999174246,
+      "learning_rate": 7.791140318847445e-08,
+      "loss": 0.6149,
+      "num_input_tokens_seen": 162195130,
+      "step": 7597
+    },
+    {
+      "epoch": 0.9136054830758131,
+      "grad_norm": 1.787764871639484,
+      "learning_rate": 7.769624559356081e-08,
+      "loss": 0.7889,
+      "num_input_tokens_seen": 162218245,
+      "step": 7598
+    },
+    {
+      "epoch": 0.9137257259664522,
+      "grad_norm": 3.2829866993559094,
+      "learning_rate": 7.748137961059842e-08,
+      "loss": 0.7483,
+      "num_input_tokens_seen": 162231945,
+      "step": 7599
+    },
+    {
+      "epoch": 0.9138459688570914,
+      "grad_norm": 3.0265571629836345,
+      "learning_rate": 7.726680527218211e-08,
+      "loss": 0.6552,
+      "num_input_tokens_seen": 162248705,
+      "step": 7600
+    },
+    {
+      "epoch": 0.9139662117477304,
+      "grad_norm": 2.1066769458871057,
+      "learning_rate": 7.70525226108627e-08,
+      "loss": 0.7506,
+      "num_input_tokens_seen": 162272095,
+      "step": 7601
+    },
+    {
+      "epoch": 0.9140864546383695,
+      "grad_norm": 2.127204793214638,
+      "learning_rate": 7.683853165914644e-08,
+      "loss": 0.7939,
+      "num_input_tokens_seen": 162289585,
+      "step": 7602
+    },
+    {
+      "epoch": 0.9142066975290086,
+      "grad_norm": 1.797980066794584,
+      "learning_rate": 7.662483244949602e-08,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 162306565,
+      "step": 7603
+    },
+    {
+      "epoch": 0.9143269404196477,
+      "grad_norm": 2.4006741565354615,
+      "learning_rate": 7.641142501432951e-08,
+      "loss": 0.7972,
+      "num_input_tokens_seen": 162322480,
+      "step": 7604
+    },
+    {
+      "epoch": 0.9144471833102867,
+      "grad_norm": 2.48748038242498,
+      "learning_rate": 7.619830938602013e-08,
+      "loss": 0.7323,
+      "num_input_tokens_seen": 162343425,
+      "step": 7605
+    },
+    {
+      "epoch": 0.9145674262009259,
+      "grad_norm": 2.829063516006072,
+      "learning_rate": 7.598548559689777e-08,
+      "loss": 0.8067,
+      "num_input_tokens_seen": 162361545,
+      "step": 7606
+    },
+    {
+      "epoch": 0.914687669091565,
+      "grad_norm": 2.823479185294803,
+      "learning_rate": 7.577295367924751e-08,
+      "loss": 0.8043,
+      "num_input_tokens_seen": 162377665,
+      "step": 7607
+    },
+    {
+      "epoch": 0.914807911982204,
+      "grad_norm": 1.889590800438572,
+      "learning_rate": 7.556071366531025e-08,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 162398355,
+      "step": 7608
+    },
+    {
+      "epoch": 0.9149281548728432,
+      "grad_norm": 4.336539853757118,
+      "learning_rate": 7.534876558728242e-08,
+      "loss": 0.7776,
+      "num_input_tokens_seen": 162417245,
+      "step": 7609
+    },
+    {
+      "epoch": 0.9150483977634822,
+      "grad_norm": 4.194804596968461,
+      "learning_rate": 7.513710947731656e-08,
+      "loss": 0.7378,
+      "num_input_tokens_seen": 162438175,
+      "step": 7610
+    },
+    {
+      "epoch": 0.9151686406541213,
+      "grad_norm": 2.033333265548982,
+      "learning_rate": 7.492574536752095e-08,
+      "loss": 0.8443,
+      "num_input_tokens_seen": 162457885,
+      "step": 7611
+    },
+    {
+      "epoch": 0.9152888835447605,
+      "grad_norm": 1.7872056373206395,
+      "learning_rate": 7.471467328995907e-08,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 162476415,
+      "step": 7612
+    },
+    {
+      "epoch": 0.9154091264353995,
+      "grad_norm": 4.248516437476192,
+      "learning_rate": 7.450389327665041e-08,
+      "loss": 0.6096,
+      "num_input_tokens_seen": 162493970,
+      "step": 7613
+    },
+    {
+      "epoch": 0.9155293693260386,
+      "grad_norm": 2.6887946778669503,
+      "learning_rate": 7.429340535957007e-08,
+      "loss": 0.6774,
+      "num_input_tokens_seen": 162508885,
+      "step": 7614
+    },
+    {
+      "epoch": 0.9156496122166777,
+      "grad_norm": 2.5126813821631684,
+      "learning_rate": 7.40832095706494e-08,
+      "loss": 0.7062,
+      "num_input_tokens_seen": 162525300,
+      "step": 7615
+    },
+    {
+      "epoch": 0.9157698551073168,
+      "grad_norm": 1.9773764498642252,
+      "learning_rate": 7.387330594177443e-08,
+      "loss": 0.7947,
+      "num_input_tokens_seen": 162547095,
+      "step": 7616
+    },
+    {
+      "epoch": 0.9158900979979558,
+      "grad_norm": 2.463664845115131,
+      "learning_rate": 7.366369450478749e-08,
+      "loss": 0.7805,
+      "num_input_tokens_seen": 162567925,
+      "step": 7617
+    },
+    {
+      "epoch": 0.916010340888595,
+      "grad_norm": 1.7945315841860985,
+      "learning_rate": 7.345437529148623e-08,
+      "loss": 0.6497,
+      "num_input_tokens_seen": 162586655,
+      "step": 7618
+    },
+    {
+      "epoch": 0.9161305837792341,
+      "grad_norm": 2.7694333079269744,
+      "learning_rate": 7.324534833362461e-08,
+      "loss": 0.7283,
+      "num_input_tokens_seen": 162603950,
+      "step": 7619
+    },
+    {
+      "epoch": 0.9162508266698731,
+      "grad_norm": 1.8925551101515905,
+      "learning_rate": 7.303661366291192e-08,
+      "loss": 0.6785,
+      "num_input_tokens_seen": 162624340,
+      "step": 7620
+    },
+    {
+      "epoch": 0.9163710695605123,
+      "grad_norm": 2.2886122427724906,
+      "learning_rate": 7.28281713110126e-08,
+      "loss": 0.8145,
+      "num_input_tokens_seen": 162642135,
+      "step": 7621
+    },
+    {
+      "epoch": 0.9164913124511513,
+      "grad_norm": 2.4022685341152537,
+      "learning_rate": 7.262002130954759e-08,
+      "loss": 0.7693,
+      "num_input_tokens_seen": 162660310,
+      "step": 7622
+    },
+    {
+      "epoch": 0.9166115553417904,
+      "grad_norm": 2.004384622669353,
+      "learning_rate": 7.241216369009296e-08,
+      "loss": 0.784,
+      "num_input_tokens_seen": 162680215,
+      "step": 7623
+    },
+    {
+      "epoch": 0.9167317982324296,
+      "grad_norm": 2.402848142844697,
+      "learning_rate": 7.220459848418037e-08,
+      "loss": 0.6618,
+      "num_input_tokens_seen": 162700010,
+      "step": 7624
+    },
+    {
+      "epoch": 0.9168520411230686,
+      "grad_norm": 2.023539090655207,
+      "learning_rate": 7.199732572329731e-08,
+      "loss": 0.7881,
+      "num_input_tokens_seen": 162717630,
+      "step": 7625
+    },
+    {
+      "epoch": 0.9169722840137077,
+      "grad_norm": 3.0870319218966245,
+      "learning_rate": 7.179034543888684e-08,
+      "loss": 0.762,
+      "num_input_tokens_seen": 162736855,
+      "step": 7626
+    },
+    {
+      "epoch": 0.9170925269043467,
+      "grad_norm": 4.534386764551792,
+      "learning_rate": 7.158365766234808e-08,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 162755425,
+      "step": 7627
+    },
+    {
+      "epoch": 0.9172127697949859,
+      "grad_norm": 2.0964458591415425,
+      "learning_rate": 7.137726242503527e-08,
+      "loss": 0.7147,
+      "num_input_tokens_seen": 162774065,
+      "step": 7628
+    },
+    {
+      "epoch": 0.917333012685625,
+      "grad_norm": 2.8927536128091584,
+      "learning_rate": 7.11711597582585e-08,
+      "loss": 0.779,
+      "num_input_tokens_seen": 162791145,
+      "step": 7629
+    },
+    {
+      "epoch": 0.917453255576264,
+      "grad_norm": 2.616532091973877,
+      "learning_rate": 7.096534969328271e-08,
+      "loss": 0.7988,
+      "num_input_tokens_seen": 162808310,
+      "step": 7630
+    },
+    {
+      "epoch": 0.9175734984669032,
+      "grad_norm": 2.444107560883417,
+      "learning_rate": 7.075983226132987e-08,
+      "loss": 0.8364,
+      "num_input_tokens_seen": 162826960,
+      "step": 7631
+    },
+    {
+      "epoch": 0.9176937413575422,
+      "grad_norm": 2.830071813843808,
+      "learning_rate": 7.055460749357656e-08,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 162842960,
+      "step": 7632
+    },
+    {
+      "epoch": 0.9178139842481813,
+      "grad_norm": 2.5152038799735332,
+      "learning_rate": 7.034967542115521e-08,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 162860945,
+      "step": 7633
+    },
+    {
+      "epoch": 0.9179342271388204,
+      "grad_norm": 2.463994410315292,
+      "learning_rate": 7.014503607515366e-08,
+      "loss": 0.7455,
+      "num_input_tokens_seen": 162879970,
+      "step": 7634
+    },
+    {
+      "epoch": 0.9180544700294595,
+      "grad_norm": 2.3132634060940616,
+      "learning_rate": 6.994068948661592e-08,
+      "loss": 0.6691,
+      "num_input_tokens_seen": 162897845,
+      "step": 7635
+    },
+    {
+      "epoch": 0.9181747129200986,
+      "grad_norm": 2.900983597196399,
+      "learning_rate": 6.97366356865412e-08,
+      "loss": 0.7539,
+      "num_input_tokens_seen": 162915270,
+      "step": 7636
+    },
+    {
+      "epoch": 0.9182949558107377,
+      "grad_norm": 2.322618345785149,
+      "learning_rate": 6.953287470588386e-08,
+      "loss": 0.6467,
+      "num_input_tokens_seen": 162932945,
+      "step": 7637
+    },
+    {
+      "epoch": 0.9184151987013768,
+      "grad_norm": 3.0418981070969644,
+      "learning_rate": 6.932940657555452e-08,
+      "loss": 0.8516,
+      "num_input_tokens_seen": 162948795,
+      "step": 7638
+    },
+    {
+      "epoch": 0.9185354415920158,
+      "grad_norm": 2.082465536662542,
+      "learning_rate": 6.912623132641938e-08,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 162973605,
+      "step": 7639
+    },
+    {
+      "epoch": 0.918655684482655,
+      "grad_norm": 1.902650248504533,
+      "learning_rate": 6.892334898929952e-08,
+      "loss": 0.7512,
+      "num_input_tokens_seen": 162993570,
+      "step": 7640
+    },
+    {
+      "epoch": 0.918775927373294,
+      "grad_norm": 3.0512414338728653,
+      "learning_rate": 6.872075959497236e-08,
+      "loss": 0.8343,
+      "num_input_tokens_seen": 163012065,
+      "step": 7641
+    },
+    {
+      "epoch": 0.9188961702639331,
+      "grad_norm": 2.6949847214219194,
+      "learning_rate": 6.85184631741702e-08,
+      "loss": 0.8193,
+      "num_input_tokens_seen": 163032350,
+      "step": 7642
+    },
+    {
+      "epoch": 0.9190164131545723,
+      "grad_norm": 2.4017060426118664,
+      "learning_rate": 6.831645975758161e-08,
+      "loss": 0.7731,
+      "num_input_tokens_seen": 163050010,
+      "step": 7643
+    },
+    {
+      "epoch": 0.9191366560452113,
+      "grad_norm": 2.6698715287134633,
+      "learning_rate": 6.811474937585026e-08,
+      "loss": 0.6744,
+      "num_input_tokens_seen": 163069520,
+      "step": 7644
+    },
+    {
+      "epoch": 0.9192568989358504,
+      "grad_norm": 1.8434418050130366,
+      "learning_rate": 6.79133320595755e-08,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 163089160,
+      "step": 7645
+    },
+    {
+      "epoch": 0.9193771418264896,
+      "grad_norm": 2.1281752298736496,
+      "learning_rate": 6.771220783931175e-08,
+      "loss": 0.7466,
+      "num_input_tokens_seen": 163109040,
+      "step": 7646
+    },
+    {
+      "epoch": 0.9194973847171286,
+      "grad_norm": 0.8726606417077498,
+      "learning_rate": 6.751137674556994e-08,
+      "loss": 0.6767,
+      "num_input_tokens_seen": 163169145,
+      "step": 7647
+    },
+    {
+      "epoch": 0.9196176276077677,
+      "grad_norm": 2.1728873456451234,
+      "learning_rate": 6.731083880881572e-08,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 163185085,
+      "step": 7648
+    },
+    {
+      "epoch": 0.9197378704984068,
+      "grad_norm": 2.3115927579783695,
+      "learning_rate": 6.711059405947072e-08,
+      "loss": 0.8054,
+      "num_input_tokens_seen": 163202995,
+      "step": 7649
+    },
+    {
+      "epoch": 0.9198581133890459,
+      "grad_norm": 2.560097991471572,
+      "learning_rate": 6.691064252791156e-08,
+      "loss": 0.7694,
+      "num_input_tokens_seen": 163222190,
+      "step": 7650
+    },
+    {
+      "epoch": 0.9199783562796849,
+      "grad_norm": 1.9551818290424663,
+      "learning_rate": 6.67109842444713e-08,
+      "loss": 0.7699,
+      "num_input_tokens_seen": 163240840,
+      "step": 7651
+    },
+    {
+      "epoch": 0.9200985991703241,
+      "grad_norm": 3.25240196965659,
+      "learning_rate": 6.651161923943726e-08,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 163258465,
+      "step": 7652
+    },
+    {
+      "epoch": 0.9202188420609632,
+      "grad_norm": 3.024121756037315,
+      "learning_rate": 6.631254754305349e-08,
+      "loss": 0.7619,
+      "num_input_tokens_seen": 163277645,
+      "step": 7653
+    },
+    {
+      "epoch": 0.9203390849516022,
+      "grad_norm": 2.3864893769323388,
+      "learning_rate": 6.611376918551848e-08,
+      "loss": 0.7778,
+      "num_input_tokens_seen": 163296150,
+      "step": 7654
+    },
+    {
+      "epoch": 0.9204593278422414,
+      "grad_norm": 2.4106302106160684,
+      "learning_rate": 6.591528419698744e-08,
+      "loss": 0.7919,
+      "num_input_tokens_seen": 163315655,
+      "step": 7655
+    },
+    {
+      "epoch": 0.9205795707328804,
+      "grad_norm": 2.983989664660224,
+      "learning_rate": 6.571709260756986e-08,
+      "loss": 0.8333,
+      "num_input_tokens_seen": 163332020,
+      "step": 7656
+    },
+    {
+      "epoch": 0.9206998136235195,
+      "grad_norm": 3.445360355747961,
+      "learning_rate": 6.551919444733145e-08,
+      "loss": 0.7543,
+      "num_input_tokens_seen": 163349555,
+      "step": 7657
+    },
+    {
+      "epoch": 0.9208200565141585,
+      "grad_norm": 1.9893530282535243,
+      "learning_rate": 6.532158974629287e-08,
+      "loss": 0.6436,
+      "num_input_tokens_seen": 163373030,
+      "step": 7658
+    },
+    {
+      "epoch": 0.9209402994047977,
+      "grad_norm": 2.275459375063756,
+      "learning_rate": 6.512427853443103e-08,
+      "loss": 0.758,
+      "num_input_tokens_seen": 163394830,
+      "step": 7659
+    },
+    {
+      "epoch": 0.9210605422954368,
+      "grad_norm": 1.6606415919907909,
+      "learning_rate": 6.492726084167799e-08,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 163416665,
+      "step": 7660
+    },
+    {
+      "epoch": 0.9211807851860758,
+      "grad_norm": 0.807823609834835,
+      "learning_rate": 6.473053669792072e-08,
+      "loss": 0.5804,
+      "num_input_tokens_seen": 163471075,
+      "step": 7661
+    },
+    {
+      "epoch": 0.921301028076715,
+      "grad_norm": 2.606458243927914,
+      "learning_rate": 6.453410613300225e-08,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 163488725,
+      "step": 7662
+    },
+    {
+      "epoch": 0.921421270967354,
+      "grad_norm": 1.8033922760765775,
+      "learning_rate": 6.433796917672118e-08,
+      "loss": 0.578,
+      "num_input_tokens_seen": 163507650,
+      "step": 7663
+    },
+    {
+      "epoch": 0.9215415138579931,
+      "grad_norm": 0.7869964700232591,
+      "learning_rate": 6.414212585883105e-08,
+      "loss": 0.6168,
+      "num_input_tokens_seen": 163570000,
+      "step": 7664
+    },
+    {
+      "epoch": 0.9216617567486323,
+      "grad_norm": 2.07232525265648,
+      "learning_rate": 6.394657620904143e-08,
+      "loss": 0.6949,
+      "num_input_tokens_seen": 163592830,
+      "step": 7665
+    },
+    {
+      "epoch": 0.9217819996392713,
+      "grad_norm": 1.9840034656730698,
+      "learning_rate": 6.375132025701657e-08,
+      "loss": 0.7092,
+      "num_input_tokens_seen": 163614850,
+      "step": 7666
+    },
+    {
+      "epoch": 0.9219022425299104,
+      "grad_norm": 7.31759322867845,
+      "learning_rate": 6.355635803237724e-08,
+      "loss": 0.6872,
+      "num_input_tokens_seen": 163630270,
+      "step": 7667
+    },
+    {
+      "epoch": 0.9220224854205495,
+      "grad_norm": 2.2343063216639405,
+      "learning_rate": 6.336168956469867e-08,
+      "loss": 0.7996,
+      "num_input_tokens_seen": 163648465,
+      "step": 7668
+    },
+    {
+      "epoch": 0.9221427283111886,
+      "grad_norm": 1.767091412407001,
+      "learning_rate": 6.31673148835119e-08,
+      "loss": 0.715,
+      "num_input_tokens_seen": 163669375,
+      "step": 7669
+    },
+    {
+      "epoch": 0.9222629712018277,
+      "grad_norm": 1.784182565960609,
+      "learning_rate": 6.297323401830334e-08,
+      "loss": 0.6345,
+      "num_input_tokens_seen": 163687880,
+      "step": 7670
+    },
+    {
+      "epoch": 0.9223832140924668,
+      "grad_norm": 2.3693360289981107,
+      "learning_rate": 6.277944699851523e-08,
+      "loss": 0.6881,
+      "num_input_tokens_seen": 163707120,
+      "step": 7671
+    },
+    {
+      "epoch": 0.9225034569831059,
+      "grad_norm": 1.9841378315237865,
+      "learning_rate": 6.25859538535447e-08,
+      "loss": 0.7275,
+      "num_input_tokens_seen": 163727635,
+      "step": 7672
+    },
+    {
+      "epoch": 0.9226236998737449,
+      "grad_norm": 3.5849558296867365,
+      "learning_rate": 6.239275461274474e-08,
+      "loss": 0.7663,
+      "num_input_tokens_seen": 163743730,
+      "step": 7673
+    },
+    {
+      "epoch": 0.9227439427643841,
+      "grad_norm": 2.0079527815440104,
+      "learning_rate": 6.219984930542299e-08,
+      "loss": 0.8525,
+      "num_input_tokens_seen": 163764190,
+      "step": 7674
+    },
+    {
+      "epoch": 0.9228641856550232,
+      "grad_norm": 2.567093971340067,
+      "learning_rate": 6.200723796084362e-08,
+      "loss": 0.7538,
+      "num_input_tokens_seen": 163782005,
+      "step": 7675
+    },
+    {
+      "epoch": 0.9229844285456622,
+      "grad_norm": 0.7644028197374373,
+      "learning_rate": 6.181492060822546e-08,
+      "loss": 0.6434,
+      "num_input_tokens_seen": 163841900,
+      "step": 7676
+    },
+    {
+      "epoch": 0.9231046714363014,
+      "grad_norm": 2.210626430229041,
+      "learning_rate": 6.162289727674274e-08,
+      "loss": 0.8144,
+      "num_input_tokens_seen": 163859300,
+      "step": 7677
+    },
+    {
+      "epoch": 0.9232249143269404,
+      "grad_norm": 2.9426182512994665,
+      "learning_rate": 6.143116799552506e-08,
+      "loss": 0.8769,
+      "num_input_tokens_seen": 163875265,
+      "step": 7678
+    },
+    {
+      "epoch": 0.9233451572175795,
+      "grad_norm": 2.5919334311516073,
+      "learning_rate": 6.123973279365802e-08,
+      "loss": 0.5499,
+      "num_input_tokens_seen": 163893960,
+      "step": 7679
+    },
+    {
+      "epoch": 0.9234654001082186,
+      "grad_norm": 1.8493410164895827,
+      "learning_rate": 6.104859170018218e-08,
+      "loss": 0.7756,
+      "num_input_tokens_seen": 163911535,
+      "step": 7680
+    },
+    {
+      "epoch": 0.9235856429988577,
+      "grad_norm": 1.689468347824574,
+      "learning_rate": 6.085774474409322e-08,
+      "loss": 0.802,
+      "num_input_tokens_seen": 163931815,
+      "step": 7681
+    },
+    {
+      "epoch": 0.9237058858894968,
+      "grad_norm": 2.169122796669728,
+      "learning_rate": 6.066719195434267e-08,
+      "loss": 0.6983,
+      "num_input_tokens_seen": 163949335,
+      "step": 7682
+    },
+    {
+      "epoch": 0.9238261287801359,
+      "grad_norm": 4.430542826929758,
+      "learning_rate": 6.047693335983717e-08,
+      "loss": 0.661,
+      "num_input_tokens_seen": 163971400,
+      "step": 7683
+    },
+    {
+      "epoch": 0.923946371670775,
+      "grad_norm": 4.4135184896564414,
+      "learning_rate": 6.028696898943853e-08,
+      "loss": 0.817,
+      "num_input_tokens_seen": 163990180,
+      "step": 7684
+    },
+    {
+      "epoch": 0.924066614561414,
+      "grad_norm": 4.047769178950823,
+      "learning_rate": 6.00972988719648e-08,
+      "loss": 0.7062,
+      "num_input_tokens_seen": 164008135,
+      "step": 7685
+    },
+    {
+      "epoch": 0.9241868574520532,
+      "grad_norm": 6.575575374632869,
+      "learning_rate": 5.990792303618807e-08,
+      "loss": 0.7103,
+      "num_input_tokens_seen": 164027435,
+      "step": 7686
+    },
+    {
+      "epoch": 0.9243071003426923,
+      "grad_norm": 1.8826012948425552,
+      "learning_rate": 5.971884151083695e-08,
+      "loss": 0.6934,
+      "num_input_tokens_seen": 164049565,
+      "step": 7687
+    },
+    {
+      "epoch": 0.9244273432333313,
+      "grad_norm": 1.829255598550161,
+      "learning_rate": 5.9530054324595124e-08,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 164069400,
+      "step": 7688
+    },
+    {
+      "epoch": 0.9245475861239704,
+      "grad_norm": 0.7772465622589103,
+      "learning_rate": 5.934156150610103e-08,
+      "loss": 0.5959,
+      "num_input_tokens_seen": 164126485,
+      "step": 7689
+    },
+    {
+      "epoch": 0.9246678290146095,
+      "grad_norm": 2.4899874784998315,
+      "learning_rate": 5.915336308394891e-08,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 164145040,
+      "step": 7690
+    },
+    {
+      "epoch": 0.9247880719052486,
+      "grad_norm": 1.6930480601185507,
+      "learning_rate": 5.89654590866886e-08,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 164164260,
+      "step": 7691
+    },
+    {
+      "epoch": 0.9249083147958876,
+      "grad_norm": 2.3324529964104532,
+      "learning_rate": 5.877784954282483e-08,
+      "loss": 0.8775,
+      "num_input_tokens_seen": 164183320,
+      "step": 7692
+    },
+    {
+      "epoch": 0.9250285576865268,
+      "grad_norm": 3.111661559376203,
+      "learning_rate": 5.8590534480817963e-08,
+      "loss": 0.7191,
+      "num_input_tokens_seen": 164204765,
+      "step": 7693
+    },
+    {
+      "epoch": 0.9251488005771659,
+      "grad_norm": 2.9735658744086466,
+      "learning_rate": 5.840351392908349e-08,
+      "loss": 0.723,
+      "num_input_tokens_seen": 164220205,
+      "step": 7694
+    },
+    {
+      "epoch": 0.9252690434678049,
+      "grad_norm": 3.2971726652754314,
+      "learning_rate": 5.821678791599205e-08,
+      "loss": 0.7095,
+      "num_input_tokens_seen": 164239370,
+      "step": 7695
+    },
+    {
+      "epoch": 0.9253892863584441,
+      "grad_norm": 1.8403975449503616,
+      "learning_rate": 5.803035646986987e-08,
+      "loss": 0.801,
+      "num_input_tokens_seen": 164258425,
+      "step": 7696
+    },
+    {
+      "epoch": 0.9255095292490831,
+      "grad_norm": 2.825252842672205,
+      "learning_rate": 5.7844219618998766e-08,
+      "loss": 0.6687,
+      "num_input_tokens_seen": 164272470,
+      "step": 7697
+    },
+    {
+      "epoch": 0.9256297721397222,
+      "grad_norm": 2.2738473960864876,
+      "learning_rate": 5.765837739161505e-08,
+      "loss": 0.7096,
+      "num_input_tokens_seen": 164291310,
+      "step": 7698
+    },
+    {
+      "epoch": 0.9257500150303614,
+      "grad_norm": 3.8134690384264545,
+      "learning_rate": 5.7472829815911504e-08,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 164309855,
+      "step": 7699
+    },
+    {
+      "epoch": 0.9258702579210004,
+      "grad_norm": 1.7736622015278536,
+      "learning_rate": 5.7287576920035164e-08,
+      "loss": 0.8023,
+      "num_input_tokens_seen": 164328590,
+      "step": 7700
+    },
+    {
+      "epoch": 0.9259905008116395,
+      "grad_norm": 1.9273500005798785,
+      "learning_rate": 5.710261873208866e-08,
+      "loss": 0.7544,
+      "num_input_tokens_seen": 164347640,
+      "step": 7701
+    },
+    {
+      "epoch": 0.9261107437022786,
+      "grad_norm": 1.6368246176197416,
+      "learning_rate": 5.691795528012999e-08,
+      "loss": 0.7347,
+      "num_input_tokens_seen": 164368840,
+      "step": 7702
+    },
+    {
+      "epoch": 0.9262309865929177,
+      "grad_norm": 2.5428603472362052,
+      "learning_rate": 5.6733586592172755e-08,
+      "loss": 0.7208,
+      "num_input_tokens_seen": 164388055,
+      "step": 7703
+    },
+    {
+      "epoch": 0.9263512294835567,
+      "grad_norm": 2.8726336453957617,
+      "learning_rate": 5.6549512696185244e-08,
+      "loss": 0.8029,
+      "num_input_tokens_seen": 164406275,
+      "step": 7704
+    },
+    {
+      "epoch": 0.9264714723741959,
+      "grad_norm": 2.180029138224243,
+      "learning_rate": 5.636573362009156e-08,
+      "loss": 0.6788,
+      "num_input_tokens_seen": 164426055,
+      "step": 7705
+    },
+    {
+      "epoch": 0.926591715264835,
+      "grad_norm": 2.1096575202205017,
+      "learning_rate": 5.618224939177052e-08,
+      "loss": 0.7558,
+      "num_input_tokens_seen": 164443680,
+      "step": 7706
+    },
+    {
+      "epoch": 0.926711958155474,
+      "grad_norm": 2.5465581893520937,
+      "learning_rate": 5.5999060039056964e-08,
+      "loss": 0.6965,
+      "num_input_tokens_seen": 164465945,
+      "step": 7707
+    },
+    {
+      "epoch": 0.9268322010461132,
+      "grad_norm": 2.590806558290182,
+      "learning_rate": 5.581616558974023e-08,
+      "loss": 0.8204,
+      "num_input_tokens_seen": 164484230,
+      "step": 7708
+    },
+    {
+      "epoch": 0.9269524439367522,
+      "grad_norm": 3.7070702873445693,
+      "learning_rate": 5.5633566071565444e-08,
+      "loss": 0.7849,
+      "num_input_tokens_seen": 164503190,
+      "step": 7709
+    },
+    {
+      "epoch": 0.9270726868273913,
+      "grad_norm": 2.8166443618556927,
+      "learning_rate": 5.5451261512232896e-08,
+      "loss": 0.6937,
+      "num_input_tokens_seen": 164525590,
+      "step": 7710
+    },
+    {
+      "epoch": 0.9271929297180305,
+      "grad_norm": 2.451214690879537,
+      "learning_rate": 5.5269251939397576e-08,
+      "loss": 0.618,
+      "num_input_tokens_seen": 164544825,
+      "step": 7711
+    },
+    {
+      "epoch": 0.9273131726086695,
+      "grad_norm": 2.2009864936452934,
+      "learning_rate": 5.508753738067073e-08,
+      "loss": 0.759,
+      "num_input_tokens_seen": 164564085,
+      "step": 7712
+    },
+    {
+      "epoch": 0.9274334154993086,
+      "grad_norm": 2.0569710206288003,
+      "learning_rate": 5.4906117863617875e-08,
+      "loss": 0.7829,
+      "num_input_tokens_seen": 164583190,
+      "step": 7713
+    },
+    {
+      "epoch": 0.9275536583899477,
+      "grad_norm": 1.8967905050436942,
+      "learning_rate": 5.4724993415760533e-08,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 164601265,
+      "step": 7714
+    },
+    {
+      "epoch": 0.9276739012805868,
+      "grad_norm": 3.4088639460890398,
+      "learning_rate": 5.454416406457496e-08,
+      "loss": 0.7467,
+      "num_input_tokens_seen": 164620080,
+      "step": 7715
+    },
+    {
+      "epoch": 0.9277941441712259,
+      "grad_norm": 3.5453436908898968,
+      "learning_rate": 5.436362983749299e-08,
+      "loss": 0.7404,
+      "num_input_tokens_seen": 164634970,
+      "step": 7716
+    },
+    {
+      "epoch": 0.927914387061865,
+      "grad_norm": 2.306579451410043,
+      "learning_rate": 5.418339076190137e-08,
+      "loss": 0.6332,
+      "num_input_tokens_seen": 164654200,
+      "step": 7717
+    },
+    {
+      "epoch": 0.9280346299525041,
+      "grad_norm": 2.281436237660145,
+      "learning_rate": 5.400344686514202e-08,
+      "loss": 0.8822,
+      "num_input_tokens_seen": 164671505,
+      "step": 7718
+    },
+    {
+      "epoch": 0.9281548728431431,
+      "grad_norm": 2.0002020547021266,
+      "learning_rate": 5.382379817451288e-08,
+      "loss": 0.6597,
+      "num_input_tokens_seen": 164689340,
+      "step": 7719
+    },
+    {
+      "epoch": 0.9282751157337822,
+      "grad_norm": 1.7624345531522285,
+      "learning_rate": 5.364444471726592e-08,
+      "loss": 0.8065,
+      "num_input_tokens_seen": 164708265,
+      "step": 7720
+    },
+    {
+      "epoch": 0.9283953586244214,
+      "grad_norm": 2.751545909416791,
+      "learning_rate": 5.346538652060939e-08,
+      "loss": 0.7954,
+      "num_input_tokens_seen": 164729340,
+      "step": 7721
+    },
+    {
+      "epoch": 0.9285156015150604,
+      "grad_norm": 2.249500590737962,
+      "learning_rate": 5.3286623611705775e-08,
+      "loss": 0.6981,
+      "num_input_tokens_seen": 164747105,
+      "step": 7722
+    },
+    {
+      "epoch": 0.9286358444056995,
+      "grad_norm": 0.9058493096277834,
+      "learning_rate": 5.3108156017673824e-08,
+      "loss": 0.6502,
+      "num_input_tokens_seen": 164808585,
+      "step": 7723
+    },
+    {
+      "epoch": 0.9287560872963386,
+      "grad_norm": 2.6429313806698795,
+      "learning_rate": 5.292998376558655e-08,
+      "loss": 0.7077,
+      "num_input_tokens_seen": 164827085,
+      "step": 7724
+    },
+    {
+      "epoch": 0.9288763301869777,
+      "grad_norm": 1.9818057872456156,
+      "learning_rate": 5.275210688247278e-08,
+      "loss": 0.6238,
+      "num_input_tokens_seen": 164847130,
+      "step": 7725
+    },
+    {
+      "epoch": 0.9289965730776167,
+      "grad_norm": 2.813342292186378,
+      "learning_rate": 5.257452539531604e-08,
+      "loss": 0.8399,
+      "num_input_tokens_seen": 164863920,
+      "step": 7726
+    },
+    {
+      "epoch": 0.9291168159682559,
+      "grad_norm": 2.1237894101746386,
+      "learning_rate": 5.2397239331055445e-08,
+      "loss": 0.6806,
+      "num_input_tokens_seen": 164885640,
+      "step": 7727
+    },
+    {
+      "epoch": 0.929237058858895,
+      "grad_norm": 2.6320667404706826,
+      "learning_rate": 5.2220248716585036e-08,
+      "loss": 0.7978,
+      "num_input_tokens_seen": 164903040,
+      "step": 7728
+    },
+    {
+      "epoch": 0.929357301749534,
+      "grad_norm": 3.2097207003638486,
+      "learning_rate": 5.204355357875445e-08,
+      "loss": 0.7491,
+      "num_input_tokens_seen": 164921105,
+      "step": 7729
+    },
+    {
+      "epoch": 0.9294775446401732,
+      "grad_norm": 3.2066326279463175,
+      "learning_rate": 5.1867153944367584e-08,
+      "loss": 0.701,
+      "num_input_tokens_seen": 164937215,
+      "step": 7730
+    },
+    {
+      "epoch": 0.9295977875308122,
+      "grad_norm": 1.6386494531643085,
+      "learning_rate": 5.16910498401848e-08,
+      "loss": 0.7311,
+      "num_input_tokens_seen": 164956385,
+      "step": 7731
+    },
+    {
+      "epoch": 0.9297180304214513,
+      "grad_norm": 2.12365505333182,
+      "learning_rate": 5.151524129292073e-08,
+      "loss": 0.831,
+      "num_input_tokens_seen": 164974000,
+      "step": 7732
+    },
+    {
+      "epoch": 0.9298382733120905,
+      "grad_norm": 4.194903200666486,
+      "learning_rate": 5.1339728329245155e-08,
+      "loss": 0.6703,
+      "num_input_tokens_seen": 164994285,
+      "step": 7733
+    },
+    {
+      "epoch": 0.9299585162027295,
+      "grad_norm": 3.9698574765343566,
+      "learning_rate": 5.116451097578345e-08,
+      "loss": 0.7847,
+      "num_input_tokens_seen": 165013045,
+      "step": 7734
+    },
+    {
+      "epoch": 0.9300787590933686,
+      "grad_norm": 1.8423210063791695,
+      "learning_rate": 5.0989589259115895e-08,
+      "loss": 0.7401,
+      "num_input_tokens_seen": 165033650,
+      "step": 7735
+    },
+    {
+      "epoch": 0.9301990019840077,
+      "grad_norm": 1.9752022694647853,
+      "learning_rate": 5.081496320577816e-08,
+      "loss": 0.7021,
+      "num_input_tokens_seen": 165050490,
+      "step": 7736
+    },
+    {
+      "epoch": 0.9303192448746468,
+      "grad_norm": 0.9637446798467478,
+      "learning_rate": 5.0640632842260835e-08,
+      "loss": 0.6601,
+      "num_input_tokens_seen": 165110470,
+      "step": 7737
+    },
+    {
+      "epoch": 0.9304394877652858,
+      "grad_norm": 1.5982502814379467,
+      "learning_rate": 5.0466598195009426e-08,
+      "loss": 0.7147,
+      "num_input_tokens_seen": 165137060,
+      "step": 7738
+    },
+    {
+      "epoch": 0.930559730655925,
+      "grad_norm": 2.3424280890928078,
+      "learning_rate": 5.0292859290425036e-08,
+      "loss": 0.7005,
+      "num_input_tokens_seen": 165154650,
+      "step": 7739
+    },
+    {
+      "epoch": 0.9306799735465641,
+      "grad_norm": 2.0358421300537564,
+      "learning_rate": 5.01194161548637e-08,
+      "loss": 0.773,
+      "num_input_tokens_seen": 165173485,
+      "step": 7740
+    },
+    {
+      "epoch": 0.9308002164372031,
+      "grad_norm": 2.112256078091361,
+      "learning_rate": 4.994626881463659e-08,
+      "loss": 0.8452,
+      "num_input_tokens_seen": 165189460,
+      "step": 7741
+    },
+    {
+      "epoch": 0.9309204593278423,
+      "grad_norm": 1.832913599036993,
+      "learning_rate": 4.9773417296009814e-08,
+      "loss": 0.7021,
+      "num_input_tokens_seen": 165210700,
+      "step": 7742
+    },
+    {
+      "epoch": 0.9310407022184813,
+      "grad_norm": 2.0711501273961463,
+      "learning_rate": 4.960086162520527e-08,
+      "loss": 0.6507,
+      "num_input_tokens_seen": 165230510,
+      "step": 7743
+    },
+    {
+      "epoch": 0.9311609451091204,
+      "grad_norm": 2.9479820749355037,
+      "learning_rate": 4.942860182839936e-08,
+      "loss": 0.8139,
+      "num_input_tokens_seen": 165248575,
+      "step": 7744
+    },
+    {
+      "epoch": 0.9312811879997596,
+      "grad_norm": 2.006368254204446,
+      "learning_rate": 4.925663793172363e-08,
+      "loss": 0.7832,
+      "num_input_tokens_seen": 165266255,
+      "step": 7745
+    },
+    {
+      "epoch": 0.9314014308903986,
+      "grad_norm": 0.8604809808810056,
+      "learning_rate": 4.9084969961264544e-08,
+      "loss": 0.6107,
+      "num_input_tokens_seen": 165329435,
+      "step": 7746
+    },
+    {
+      "epoch": 0.9315216737810377,
+      "grad_norm": 1.6442784922257239,
+      "learning_rate": 4.89135979430646e-08,
+      "loss": 0.7568,
+      "num_input_tokens_seen": 165349200,
+      "step": 7747
+    },
+    {
+      "epoch": 0.9316419166716768,
+      "grad_norm": 1.8634660754024868,
+      "learning_rate": 4.874252190312078e-08,
+      "loss": 0.8363,
+      "num_input_tokens_seen": 165369305,
+      "step": 7748
+    },
+    {
+      "epoch": 0.9317621595623159,
+      "grad_norm": 4.262197833330716,
+      "learning_rate": 4.857174186738477e-08,
+      "loss": 0.643,
+      "num_input_tokens_seen": 165392375,
+      "step": 7749
+    },
+    {
+      "epoch": 0.931882402452955,
+      "grad_norm": 3.0136117953515273,
+      "learning_rate": 4.840125786176385e-08,
+      "loss": 0.7314,
+      "num_input_tokens_seen": 165408300,
+      "step": 7750
+    },
+    {
+      "epoch": 0.932002645343594,
+      "grad_norm": 1.9877469090677302,
+      "learning_rate": 4.823106991212045e-08,
+      "loss": 0.7627,
+      "num_input_tokens_seen": 165427260,
+      "step": 7751
+    },
+    {
+      "epoch": 0.9321228882342332,
+      "grad_norm": 2.082490565239733,
+      "learning_rate": 4.806117804427212e-08,
+      "loss": 0.8362,
+      "num_input_tokens_seen": 165444915,
+      "step": 7752
+    },
+    {
+      "epoch": 0.9322431311248722,
+      "grad_norm": 1.9981519634381384,
+      "learning_rate": 4.7891582283990926e-08,
+      "loss": 0.6407,
+      "num_input_tokens_seen": 165463360,
+      "step": 7753
+    },
+    {
+      "epoch": 0.9323633740155113,
+      "grad_norm": 1.70360463337487,
+      "learning_rate": 4.772228265700473e-08,
+      "loss": 0.7194,
+      "num_input_tokens_seen": 165483940,
+      "step": 7754
+    },
+    {
+      "epoch": 0.9324836169061504,
+      "grad_norm": 2.17087333477564,
+      "learning_rate": 4.75532791889961e-08,
+      "loss": 0.7446,
+      "num_input_tokens_seen": 165500360,
+      "step": 7755
+    },
+    {
+      "epoch": 0.9326038597967895,
+      "grad_norm": 2.1014668839916117,
+      "learning_rate": 4.738457190560252e-08,
+      "loss": 0.6518,
+      "num_input_tokens_seen": 165519190,
+      "step": 7756
+    },
+    {
+      "epoch": 0.9327241026874286,
+      "grad_norm": 2.5190607165203907,
+      "learning_rate": 4.721616083241686e-08,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 165537165,
+      "step": 7757
+    },
+    {
+      "epoch": 0.9328443455780677,
+      "grad_norm": 1.7832295340407653,
+      "learning_rate": 4.7048045994986684e-08,
+      "loss": 0.7673,
+      "num_input_tokens_seen": 165557745,
+      "step": 7758
+    },
+    {
+      "epoch": 0.9329645884687068,
+      "grad_norm": 2.2463172532169464,
+      "learning_rate": 4.688022741881559e-08,
+      "loss": 0.9088,
+      "num_input_tokens_seen": 165577990,
+      "step": 7759
+    },
+    {
+      "epoch": 0.9330848313593458,
+      "grad_norm": 1.676237432646866,
+      "learning_rate": 4.671270512936076e-08,
+      "loss": 0.753,
+      "num_input_tokens_seen": 165596870,
+      "step": 7760
+    },
+    {
+      "epoch": 0.933205074249985,
+      "grad_norm": 1.886040018968966,
+      "learning_rate": 4.6545479152035884e-08,
+      "loss": 0.8234,
+      "num_input_tokens_seen": 165615760,
+      "step": 7761
+    },
+    {
+      "epoch": 0.9333253171406241,
+      "grad_norm": 2.2079434040633323,
+      "learning_rate": 4.637854951220821e-08,
+      "loss": 0.7679,
+      "num_input_tokens_seen": 165632265,
+      "step": 7762
+    },
+    {
+      "epoch": 0.9334455600312631,
+      "grad_norm": 1.9321950544389397,
+      "learning_rate": 4.6211916235201485e-08,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 165650415,
+      "step": 7763
+    },
+    {
+      "epoch": 0.9335658029219023,
+      "grad_norm": 4.065464584551626,
+      "learning_rate": 4.604557934629372e-08,
+      "loss": 0.8343,
+      "num_input_tokens_seen": 165669210,
+      "step": 7764
+    },
+    {
+      "epoch": 0.9336860458125413,
+      "grad_norm": 1.9331438800195957,
+      "learning_rate": 4.587953887071805e-08,
+      "loss": 0.7987,
+      "num_input_tokens_seen": 165688750,
+      "step": 7765
+    },
+    {
+      "epoch": 0.9338062887031804,
+      "grad_norm": 2.164590352950927,
+      "learning_rate": 4.5713794833662336e-08,
+      "loss": 0.8618,
+      "num_input_tokens_seen": 165707685,
+      "step": 7766
+    },
+    {
+      "epoch": 0.9339265315938196,
+      "grad_norm": 2.1502681466433566,
+      "learning_rate": 4.5548347260270236e-08,
+      "loss": 0.6328,
+      "num_input_tokens_seen": 165726695,
+      "step": 7767
+    },
+    {
+      "epoch": 0.9340467744844586,
+      "grad_norm": 2.0988134568412047,
+      "learning_rate": 4.5383196175639905e-08,
+      "loss": 0.6946,
+      "num_input_tokens_seen": 165745435,
+      "step": 7768
+    },
+    {
+      "epoch": 0.9341670173750977,
+      "grad_norm": 2.231215763484151,
+      "learning_rate": 4.521834160482485e-08,
+      "loss": 0.742,
+      "num_input_tokens_seen": 165763895,
+      "step": 7769
+    },
+    {
+      "epoch": 0.9342872602657368,
+      "grad_norm": 2.0708266523925687,
+      "learning_rate": 4.5053783572832846e-08,
+      "loss": 0.8084,
+      "num_input_tokens_seen": 165783795,
+      "step": 7770
+    },
+    {
+      "epoch": 0.9344075031563759,
+      "grad_norm": 1.8226344806387074,
+      "learning_rate": 4.488952210462771e-08,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 165803720,
+      "step": 7771
+    },
+    {
+      "epoch": 0.9345277460470149,
+      "grad_norm": 2.109034799002808,
+      "learning_rate": 4.4725557225127495e-08,
+      "loss": 0.8507,
+      "num_input_tokens_seen": 165821780,
+      "step": 7772
+    },
+    {
+      "epoch": 0.9346479889376541,
+      "grad_norm": 1.6380201568195163,
+      "learning_rate": 4.456188895920565e-08,
+      "loss": 0.7895,
+      "num_input_tokens_seen": 165843255,
+      "step": 7773
+    },
+    {
+      "epoch": 0.9347682318282932,
+      "grad_norm": 1.9813355877921004,
+      "learning_rate": 4.439851733169031e-08,
+      "loss": 0.8417,
+      "num_input_tokens_seen": 165860765,
+      "step": 7774
+    },
+    {
+      "epoch": 0.9348884747189322,
+      "grad_norm": 2.75155907317542,
+      "learning_rate": 4.4235442367365204e-08,
+      "loss": 0.6947,
+      "num_input_tokens_seen": 165880795,
+      "step": 7775
+    },
+    {
+      "epoch": 0.9350087176095714,
+      "grad_norm": 2.182436864339903,
+      "learning_rate": 4.4072664090968327e-08,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 165898900,
+      "step": 7776
+    },
+    {
+      "epoch": 0.9351289605002104,
+      "grad_norm": 2.1022582666502005,
+      "learning_rate": 4.391018252719347e-08,
+      "loss": 0.838,
+      "num_input_tokens_seen": 165918415,
+      "step": 7777
+    },
+    {
+      "epoch": 0.9352492033908495,
+      "grad_norm": 2.1710050756791532,
+      "learning_rate": 4.3747997700688264e-08,
+      "loss": 0.685,
+      "num_input_tokens_seen": 165934810,
+      "step": 7778
+    },
+    {
+      "epoch": 0.9353694462814887,
+      "grad_norm": 3.0018462457921795,
+      "learning_rate": 4.358610963605658e-08,
+      "loss": 0.7398,
+      "num_input_tokens_seen": 165954980,
+      "step": 7779
+    },
+    {
+      "epoch": 0.9354896891721277,
+      "grad_norm": 2.4629164433005224,
+      "learning_rate": 4.342451835785677e-08,
+      "loss": 0.679,
+      "num_input_tokens_seen": 165975610,
+      "step": 7780
+    },
+    {
+      "epoch": 0.9356099320627668,
+      "grad_norm": 2.1366441601098596,
+      "learning_rate": 4.3263223890601665e-08,
+      "loss": 0.7443,
+      "num_input_tokens_seen": 165994040,
+      "step": 7781
+    },
+    {
+      "epoch": 0.9357301749534058,
+      "grad_norm": 1.7920647430720333,
+      "learning_rate": 4.31022262587597e-08,
+      "loss": 0.7955,
+      "num_input_tokens_seen": 166012435,
+      "step": 7782
+    },
+    {
+      "epoch": 0.935850417844045,
+      "grad_norm": 1.8015132620110743,
+      "learning_rate": 4.2941525486754225e-08,
+      "loss": 0.6537,
+      "num_input_tokens_seen": 166032475,
+      "step": 7783
+    },
+    {
+      "epoch": 0.935970660734684,
+      "grad_norm": 2.5462421615259325,
+      "learning_rate": 4.2781121598963076e-08,
+      "loss": 0.7851,
+      "num_input_tokens_seen": 166050035,
+      "step": 7784
+    },
+    {
+      "epoch": 0.9360909036253231,
+      "grad_norm": 13.549006132997718,
+      "learning_rate": 4.2621014619719896e-08,
+      "loss": 0.6718,
+      "num_input_tokens_seen": 166067520,
+      "step": 7785
+    },
+    {
+      "epoch": 0.9362111465159623,
+      "grad_norm": 0.7700988321416368,
+      "learning_rate": 4.246120457331215e-08,
+      "loss": 0.6298,
+      "num_input_tokens_seen": 166129415,
+      "step": 7786
+    },
+    {
+      "epoch": 0.9363313894066013,
+      "grad_norm": 2.145353666808575,
+      "learning_rate": 4.2301691483983325e-08,
+      "loss": 0.7176,
+      "num_input_tokens_seen": 166149255,
+      "step": 7787
+    },
+    {
+      "epoch": 0.9364516322972404,
+      "grad_norm": 5.790223092642301,
+      "learning_rate": 4.214247537593163e-08,
+      "loss": 0.75,
+      "num_input_tokens_seen": 166168225,
+      "step": 7788
+    },
+    {
+      "epoch": 0.9365718751878795,
+      "grad_norm": 2.111044250953407,
+      "learning_rate": 4.198355627330952e-08,
+      "loss": 0.7964,
+      "num_input_tokens_seen": 166186695,
+      "step": 7789
+    },
+    {
+      "epoch": 0.9366921180785186,
+      "grad_norm": 3.2707046007380423,
+      "learning_rate": 4.1824934200225034e-08,
+      "loss": 0.6911,
+      "num_input_tokens_seen": 166202085,
+      "step": 7790
+    },
+    {
+      "epoch": 0.9368123609691577,
+      "grad_norm": 1.9799627666173143,
+      "learning_rate": 4.166660918074139e-08,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 166221710,
+      "step": 7791
+    },
+    {
+      "epoch": 0.9369326038597968,
+      "grad_norm": 1.606768299346578,
+      "learning_rate": 4.15085812388758e-08,
+      "loss": 0.7276,
+      "num_input_tokens_seen": 166243650,
+      "step": 7792
+    },
+    {
+      "epoch": 0.9370528467504359,
+      "grad_norm": 1.7967776462907175,
+      "learning_rate": 4.135085039860153e-08,
+      "loss": 0.7818,
+      "num_input_tokens_seen": 166262770,
+      "step": 7793
+    },
+    {
+      "epoch": 0.9371730896410749,
+      "grad_norm": 2.8206937712409825,
+      "learning_rate": 4.119341668384568e-08,
+      "loss": 0.787,
+      "num_input_tokens_seen": 166281420,
+      "step": 7794
+    },
+    {
+      "epoch": 0.9372933325317141,
+      "grad_norm": 2.5563290745335956,
+      "learning_rate": 4.103628011849136e-08,
+      "loss": 0.8232,
+      "num_input_tokens_seen": 166296500,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9374135754223532,
+      "grad_norm": 2.056509023377302,
+      "learning_rate": 4.0879440726375506e-08,
+      "loss": 0.749,
+      "num_input_tokens_seen": 166314005,
+      "step": 7796
+    },
+    {
+      "epoch": 0.9375338183129922,
+      "grad_norm": 3.0503071618314284,
+      "learning_rate": 4.0722898531291074e-08,
+      "loss": 0.5528,
+      "num_input_tokens_seen": 166330965,
+      "step": 7797
+    },
+    {
+      "epoch": 0.9376540612036314,
+      "grad_norm": 2.241635628607921,
+      "learning_rate": 4.056665355698508e-08,
+      "loss": 0.7577,
+      "num_input_tokens_seen": 166351230,
+      "step": 7798
+    },
+    {
+      "epoch": 0.9377743040942704,
+      "grad_norm": 6.259066509423454,
+      "learning_rate": 4.0410705827159886e-08,
+      "loss": 0.8079,
+      "num_input_tokens_seen": 166368245,
+      "step": 7799
+    },
+    {
+      "epoch": 0.9378945469849095,
+      "grad_norm": 2.5048287502838043,
+      "learning_rate": 4.0255055365472356e-08,
+      "loss": 0.7063,
+      "num_input_tokens_seen": 166386060,
+      "step": 7800
+    },
+    {
+      "epoch": 0.9380147898755486,
+      "grad_norm": 2.3440089017728196,
+      "learning_rate": 4.0099702195534935e-08,
+      "loss": 0.7327,
+      "num_input_tokens_seen": 166402730,
+      "step": 7801
+    },
+    {
+      "epoch": 0.9381350327661877,
+      "grad_norm": 3.328941824586175,
+      "learning_rate": 3.99446463409141e-08,
+      "loss": 0.7586,
+      "num_input_tokens_seen": 166420305,
+      "step": 7802
+    },
+    {
+      "epoch": 0.9382552756568268,
+      "grad_norm": 3.0479777481105814,
+      "learning_rate": 3.978988782513215e-08,
+      "loss": 0.6842,
+      "num_input_tokens_seen": 166437520,
+      "step": 7803
+    },
+    {
+      "epoch": 0.9383755185474659,
+      "grad_norm": 2.086627613233314,
+      "learning_rate": 3.963542667166586e-08,
+      "loss": 0.7552,
+      "num_input_tokens_seen": 166457345,
+      "step": 7804
+    },
+    {
+      "epoch": 0.938495761438105,
+      "grad_norm": 2.2312363176614536,
+      "learning_rate": 3.9481262903946486e-08,
+      "loss": 0.6846,
+      "num_input_tokens_seen": 166476510,
+      "step": 7805
+    },
+    {
+      "epoch": 0.938616004328744,
+      "grad_norm": 0.8010731541405713,
+      "learning_rate": 3.932739654536066e-08,
+      "loss": 0.5702,
+      "num_input_tokens_seen": 166538930,
+      "step": 7806
+    },
+    {
+      "epoch": 0.9387362472193832,
+      "grad_norm": 2.2624080324645086,
+      "learning_rate": 3.917382761925014e-08,
+      "loss": 0.7303,
+      "num_input_tokens_seen": 166554485,
+      "step": 7807
+    },
+    {
+      "epoch": 0.9388564901100223,
+      "grad_norm": 1.798461760517054,
+      "learning_rate": 3.9020556148910754e-08,
+      "loss": 0.782,
+      "num_input_tokens_seen": 166573560,
+      "step": 7808
+    },
+    {
+      "epoch": 0.9389767330006613,
+      "grad_norm": 0.7378359564794622,
+      "learning_rate": 3.8867582157593895e-08,
+      "loss": 0.589,
+      "num_input_tokens_seen": 166627485,
+      "step": 7809
+    },
+    {
+      "epoch": 0.9390969758913005,
+      "grad_norm": 2.026459683623381,
+      "learning_rate": 3.871490566850544e-08,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 166651415,
+      "step": 7810
+    },
+    {
+      "epoch": 0.9392172187819395,
+      "grad_norm": 1.8556379036847632,
+      "learning_rate": 3.856252670480642e-08,
+      "loss": 0.6996,
+      "num_input_tokens_seen": 166669795,
+      "step": 7811
+    },
+    {
+      "epoch": 0.9393374616725786,
+      "grad_norm": 6.922869819428638,
+      "learning_rate": 3.841044528961279e-08,
+      "loss": 0.8079,
+      "num_input_tokens_seen": 166687310,
+      "step": 7812
+    },
+    {
+      "epoch": 0.9394577045632178,
+      "grad_norm": 2.5436438358092097,
+      "learning_rate": 3.825866144599499e-08,
+      "loss": 0.7846,
+      "num_input_tokens_seen": 166706085,
+      "step": 7813
+    },
+    {
+      "epoch": 0.9395779474538568,
+      "grad_norm": 2.460778425754977,
+      "learning_rate": 3.8107175196978145e-08,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 166722110,
+      "step": 7814
+    },
+    {
+      "epoch": 0.9396981903444959,
+      "grad_norm": 2.311435586361119,
+      "learning_rate": 3.7955986565542996e-08,
+      "loss": 0.7607,
+      "num_input_tokens_seen": 166739910,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9398184332351349,
+      "grad_norm": 2.3373259681092944,
+      "learning_rate": 3.780509557462497e-08,
+      "loss": 0.6843,
+      "num_input_tokens_seen": 166759830,
+      "step": 7816
+    },
+    {
+      "epoch": 0.9399386761257741,
+      "grad_norm": 8.600736525993172,
+      "learning_rate": 3.765450224711375e-08,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 166780055,
+      "step": 7817
+    },
+    {
+      "epoch": 0.9400589190164131,
+      "grad_norm": 1.7273285346485665,
+      "learning_rate": 3.750420660585396e-08,
+      "loss": 0.7868,
+      "num_input_tokens_seen": 166801715,
+      "step": 7818
+    },
+    {
+      "epoch": 0.9401791619070522,
+      "grad_norm": 3.6524712755973527,
+      "learning_rate": 3.735420867364603e-08,
+      "loss": 0.7935,
+      "num_input_tokens_seen": 166822415,
+      "step": 7819
+    },
+    {
+      "epoch": 0.9402994047976914,
+      "grad_norm": 2.1609752360013883,
+      "learning_rate": 3.720450847324397e-08,
+      "loss": 0.6153,
+      "num_input_tokens_seen": 166845760,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9404196476883304,
+      "grad_norm": 2.3865701095750778,
+      "learning_rate": 3.7055106027357395e-08,
+      "loss": 0.6867,
+      "num_input_tokens_seen": 166865345,
+      "step": 7821
+    },
+    {
+      "epoch": 0.9405398905789695,
+      "grad_norm": 2.893641042497842,
+      "learning_rate": 3.690600135865041e-08,
+      "loss": 0.7156,
+      "num_input_tokens_seen": 166881990,
+      "step": 7822
+    },
+    {
+      "epoch": 0.9406601334696086,
+      "grad_norm": 0.7907660928297727,
+      "learning_rate": 3.675719448974246e-08,
+      "loss": 0.6011,
+      "num_input_tokens_seen": 166946800,
+      "step": 7823
+    },
+    {
+      "epoch": 0.9407803763602477,
+      "grad_norm": 2.3144180039820137,
+      "learning_rate": 3.6608685443207054e-08,
+      "loss": 0.5961,
+      "num_input_tokens_seen": 166965670,
+      "step": 7824
+    },
+    {
+      "epoch": 0.9409006192508867,
+      "grad_norm": 2.3973680170974956,
+      "learning_rate": 3.646047424157306e-08,
+      "loss": 0.6692,
+      "num_input_tokens_seen": 166982365,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9410208621415259,
+      "grad_norm": 3.0587815088750583,
+      "learning_rate": 3.631256090732382e-08,
+      "loss": 0.6694,
+      "num_input_tokens_seen": 167002545,
+      "step": 7826
+    },
+    {
+      "epoch": 0.941141105032165,
+      "grad_norm": 2.3181352039744834,
+      "learning_rate": 3.6164945462897833e-08,
+      "loss": 0.8168,
+      "num_input_tokens_seen": 167021555,
+      "step": 7827
+    },
+    {
+      "epoch": 0.941261347922804,
+      "grad_norm": 1.9469453572826758,
+      "learning_rate": 3.6017627930688074e-08,
+      "loss": 0.7461,
+      "num_input_tokens_seen": 167041100,
+      "step": 7828
+    },
+    {
+      "epoch": 0.9413815908134432,
+      "grad_norm": 2.4890781555091306,
+      "learning_rate": 3.587060833304267e-08,
+      "loss": 0.7656,
+      "num_input_tokens_seen": 167059010,
+      "step": 7829
+    },
+    {
+      "epoch": 0.9415018337040822,
+      "grad_norm": 2.2299327797370063,
+      "learning_rate": 3.5723886692264225e-08,
+      "loss": 0.6369,
+      "num_input_tokens_seen": 167076270,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9416220765947213,
+      "grad_norm": 2.354287789934089,
+      "learning_rate": 3.557746303061071e-08,
+      "loss": 0.6112,
+      "num_input_tokens_seen": 167097745,
+      "step": 7831
+    },
+    {
+      "epoch": 0.9417423194853605,
+      "grad_norm": 2.2445783405849786,
+      "learning_rate": 3.543133737029391e-08,
+      "loss": 0.716,
+      "num_input_tokens_seen": 167117975,
+      "step": 7832
+    },
+    {
+      "epoch": 0.9418625623759995,
+      "grad_norm": 1.964919552161831,
+      "learning_rate": 3.5285509733481214e-08,
+      "loss": 0.676,
+      "num_input_tokens_seen": 167137420,
+      "step": 7833
+    },
+    {
+      "epoch": 0.9419828052666386,
+      "grad_norm": 2.0875254969405788,
+      "learning_rate": 3.513998014229447e-08,
+      "loss": 0.7632,
+      "num_input_tokens_seen": 167156090,
+      "step": 7834
+    },
+    {
+      "epoch": 0.9421030481572777,
+      "grad_norm": 4.967629022609805,
+      "learning_rate": 3.499474861881069e-08,
+      "loss": 0.8554,
+      "num_input_tokens_seen": 167173035,
+      "step": 7835
+    },
+    {
+      "epoch": 0.9422232910479168,
+      "grad_norm": 2.4773409967926403,
+      "learning_rate": 3.4849815185061136e-08,
+      "loss": 0.677,
+      "num_input_tokens_seen": 167192645,
+      "step": 7836
+    },
+    {
+      "epoch": 0.9423435339385559,
+      "grad_norm": 2.378971646462686,
+      "learning_rate": 3.470517986303223e-08,
+      "loss": 0.7582,
+      "num_input_tokens_seen": 167211350,
+      "step": 7837
+    },
+    {
+      "epoch": 0.942463776829195,
+      "grad_norm": 1.8114841893128897,
+      "learning_rate": 3.4560842674664856e-08,
+      "loss": 0.7873,
+      "num_input_tokens_seen": 167229585,
+      "step": 7838
+    },
+    {
+      "epoch": 0.9425840197198341,
+      "grad_norm": 6.504006656085906,
+      "learning_rate": 3.441680364185484e-08,
+      "loss": 0.7485,
+      "num_input_tokens_seen": 167249175,
+      "step": 7839
+    },
+    {
+      "epoch": 0.9427042626104731,
+      "grad_norm": 3.1991518237384624,
+      "learning_rate": 3.427306278645314e-08,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 167267350,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9428245055011123,
+      "grad_norm": 2.005488035228287,
+      "learning_rate": 3.4129620130264767e-08,
+      "loss": 0.7217,
+      "num_input_tokens_seen": 167285430,
+      "step": 7841
+    },
+    {
+      "epoch": 0.9429447483917514,
+      "grad_norm": 2.810382928468579,
+      "learning_rate": 3.398647569505009e-08,
+      "loss": 0.7835,
+      "num_input_tokens_seen": 167302575,
+      "step": 7842
+    },
+    {
+      "epoch": 0.9430649912823904,
+      "grad_norm": 3.051985669490127,
+      "learning_rate": 3.384362950252373e-08,
+      "loss": 0.7376,
+      "num_input_tokens_seen": 167319265,
+      "step": 7843
+    },
+    {
+      "epoch": 0.9431852341730296,
+      "grad_norm": 2.718053253896293,
+      "learning_rate": 3.3701081574355473e-08,
+      "loss": 0.5579,
+      "num_input_tokens_seen": 167340945,
+      "step": 7844
+    },
+    {
+      "epoch": 0.9433054770636686,
+      "grad_norm": 0.685433690681999,
+      "learning_rate": 3.355883193217002e-08,
+      "loss": 0.536,
+      "num_input_tokens_seen": 167409335,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9434257199543077,
+      "grad_norm": 1.9728851464632355,
+      "learning_rate": 3.341688059754588e-08,
+      "loss": 0.8762,
+      "num_input_tokens_seen": 167424710,
+      "step": 7846
+    },
+    {
+      "epoch": 0.9435459628449467,
+      "grad_norm": 2.602949673862356,
+      "learning_rate": 3.327522759201762e-08,
+      "loss": 0.7741,
+      "num_input_tokens_seen": 167444300,
+      "step": 7847
+    },
+    {
+      "epoch": 0.9436662057355859,
+      "grad_norm": 4.027816666404336,
+      "learning_rate": 3.313387293707359e-08,
+      "loss": 0.66,
+      "num_input_tokens_seen": 167462725,
+      "step": 7848
+    },
+    {
+      "epoch": 0.943786448626225,
+      "grad_norm": 2.8430588210180856,
+      "learning_rate": 3.29928166541571e-08,
+      "loss": 0.677,
+      "num_input_tokens_seen": 167481400,
+      "step": 7849
+    },
+    {
+      "epoch": 0.943906691516864,
+      "grad_norm": 2.330422352990951,
+      "learning_rate": 3.2852058764666346e-08,
+      "loss": 0.7952,
+      "num_input_tokens_seen": 167500220,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9440269344075032,
+      "grad_norm": 2.373335213428048,
+      "learning_rate": 3.2711599289954264e-08,
+      "loss": 0.685,
+      "num_input_tokens_seen": 167523975,
+      "step": 7851
+    },
+    {
+      "epoch": 0.9441471772981422,
+      "grad_norm": 2.001050152499835,
+      "learning_rate": 3.257143825132847e-08,
+      "loss": 0.7717,
+      "num_input_tokens_seen": 167541865,
+      "step": 7852
+    },
+    {
+      "epoch": 0.9442674201887813,
+      "grad_norm": 1.8185334276795595,
+      "learning_rate": 3.243157567005106e-08,
+      "loss": 0.7503,
+      "num_input_tokens_seen": 167559765,
+      "step": 7853
+    },
+    {
+      "epoch": 0.9443876630794205,
+      "grad_norm": 2.3825594706741646,
+      "learning_rate": 3.2292011567339296e-08,
+      "loss": 0.6356,
+      "num_input_tokens_seen": 167577290,
+      "step": 7854
+    },
+    {
+      "epoch": 0.9445079059700595,
+      "grad_norm": 3.7214351567185737,
+      "learning_rate": 3.21527459643649e-08,
+      "loss": 0.5607,
+      "num_input_tokens_seen": 167593895,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9446281488606986,
+      "grad_norm": 4.870754460671157,
+      "learning_rate": 3.2013778882254536e-08,
+      "loss": 0.7355,
+      "num_input_tokens_seen": 167612410,
+      "step": 7856
+    },
+    {
+      "epoch": 0.9447483917513377,
+      "grad_norm": 2.0263256189947194,
+      "learning_rate": 3.1875110342088895e-08,
+      "loss": 0.7541,
+      "num_input_tokens_seen": 167633580,
+      "step": 7857
+    },
+    {
+      "epoch": 0.9448686346419768,
+      "grad_norm": 1.8548097932559486,
+      "learning_rate": 3.1736740364904035e-08,
+      "loss": 0.6521,
+      "num_input_tokens_seen": 167653830,
+      "step": 7858
+    },
+    {
+      "epoch": 0.9449888775326158,
+      "grad_norm": 2.861439742553642,
+      "learning_rate": 3.159866897169094e-08,
+      "loss": 0.761,
+      "num_input_tokens_seen": 167671750,
+      "step": 7859
+    },
+    {
+      "epoch": 0.945109120423255,
+      "grad_norm": 2.3923390008519343,
+      "learning_rate": 3.146089618339487e-08,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 167688325,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9452293633138941,
+      "grad_norm": 2.1232775061426215,
+      "learning_rate": 3.132342202091554e-08,
+      "loss": 0.6732,
+      "num_input_tokens_seen": 167708270,
+      "step": 7861
+    },
+    {
+      "epoch": 0.9453496062045331,
+      "grad_norm": 2.9503404982755304,
+      "learning_rate": 3.1186246505107595e-08,
+      "loss": 0.6729,
+      "num_input_tokens_seen": 167727130,
+      "step": 7862
+    },
+    {
+      "epoch": 0.9454698490951723,
+      "grad_norm": 1.788639537107173,
+      "learning_rate": 3.104936965678084e-08,
+      "loss": 0.8316,
+      "num_input_tokens_seen": 167745180,
+      "step": 7863
+    },
+    {
+      "epoch": 0.9455900919858113,
+      "grad_norm": 2.3924124210275846,
+      "learning_rate": 3.091279149669934e-08,
+      "loss": 0.8039,
+      "num_input_tokens_seen": 167763690,
+      "step": 7864
+    },
+    {
+      "epoch": 0.9457103348764504,
+      "grad_norm": 2.0877745496750193,
+      "learning_rate": 3.0776512045581624e-08,
+      "loss": 0.7323,
+      "num_input_tokens_seen": 167782200,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9458305777670896,
+      "grad_norm": 2.892406012889742,
+      "learning_rate": 3.064053132410116e-08,
+      "loss": 0.7642,
+      "num_input_tokens_seen": 167799685,
+      "step": 7866
+    },
+    {
+      "epoch": 0.9459508206577286,
+      "grad_norm": 1.8794005960632765,
+      "learning_rate": 3.0504849352886554e-08,
+      "loss": 0.7503,
+      "num_input_tokens_seen": 167817550,
+      "step": 7867
+    },
+    {
+      "epoch": 0.9460710635483677,
+      "grad_norm": 2.5103010199407363,
+      "learning_rate": 3.036946615252023e-08,
+      "loss": 0.7067,
+      "num_input_tokens_seen": 167832800,
+      "step": 7868
+    },
+    {
+      "epoch": 0.9461913064390068,
+      "grad_norm": 2.1213907429428223,
+      "learning_rate": 3.0234381743539984e-08,
+      "loss": 0.6585,
+      "num_input_tokens_seen": 167850135,
+      "step": 7869
+    },
+    {
+      "epoch": 0.9463115493296459,
+      "grad_norm": 2.2415656602949827,
+      "learning_rate": 3.0099596146437863e-08,
+      "loss": 0.7976,
+      "num_input_tokens_seen": 167866960,
+      "step": 7870
+    },
+    {
+      "epoch": 0.946431792220285,
+      "grad_norm": 0.9844997154264823,
+      "learning_rate": 2.996510938166086e-08,
+      "loss": 0.6279,
+      "num_input_tokens_seen": 167929655,
+      "step": 7871
+    },
+    {
+      "epoch": 0.9465520351109241,
+      "grad_norm": 1.9779463855153512,
+      "learning_rate": 2.9830921469610196e-08,
+      "loss": 0.726,
+      "num_input_tokens_seen": 167946720,
+      "step": 7872
+    },
+    {
+      "epoch": 0.9466722780015632,
+      "grad_norm": 2.2817562370736724,
+      "learning_rate": 2.9697032430642256e-08,
+      "loss": 0.7932,
+      "num_input_tokens_seen": 167964655,
+      "step": 7873
+    },
+    {
+      "epoch": 0.9467925208922022,
+      "grad_norm": 2.4811427145918126,
+      "learning_rate": 2.9563442285067906e-08,
+      "loss": 0.7329,
+      "num_input_tokens_seen": 167981420,
+      "step": 7874
+    },
+    {
+      "epoch": 0.9469127637828414,
+      "grad_norm": 3.238490785175216,
+      "learning_rate": 2.943015105315294e-08,
+      "loss": 0.7958,
+      "num_input_tokens_seen": 168001335,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9470330066734804,
+      "grad_norm": 2.4895358851466507,
+      "learning_rate": 2.929715875511718e-08,
+      "loss": 0.6601,
+      "num_input_tokens_seen": 168020090,
+      "step": 7876
+    },
+    {
+      "epoch": 0.9471532495641195,
+      "grad_norm": 2.465899396047182,
+      "learning_rate": 2.9164465411135375e-08,
+      "loss": 0.6909,
+      "num_input_tokens_seen": 168039580,
+      "step": 7877
+    },
+    {
+      "epoch": 0.9472734924547586,
+      "grad_norm": 2.2963694950006883,
+      "learning_rate": 2.9032071041337204e-08,
+      "loss": 0.7963,
+      "num_input_tokens_seen": 168057535,
+      "step": 7878
+    },
+    {
+      "epoch": 0.9473937353453977,
+      "grad_norm": 1.9506390774048077,
+      "learning_rate": 2.889997566580704e-08,
+      "loss": 0.7248,
+      "num_input_tokens_seen": 168075410,
+      "step": 7879
+    },
+    {
+      "epoch": 0.9475139782360368,
+      "grad_norm": 1.874611849228241,
+      "learning_rate": 2.8768179304583086e-08,
+      "loss": 0.6946,
+      "num_input_tokens_seen": 168097185,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9476342211266758,
+      "grad_norm": 1.8285824138980884,
+      "learning_rate": 2.8636681977659117e-08,
+      "loss": 0.733,
+      "num_input_tokens_seen": 168116555,
+      "step": 7881
+    },
+    {
+      "epoch": 0.947754464017315,
+      "grad_norm": 2.3750887309025224,
+      "learning_rate": 2.850548370498296e-08,
+      "loss": 0.7717,
+      "num_input_tokens_seen": 168134115,
+      "step": 7882
+    },
+    {
+      "epoch": 0.9478747069079541,
+      "grad_norm": 1.859595708140476,
+      "learning_rate": 2.8374584506457577e-08,
+      "loss": 0.7057,
+      "num_input_tokens_seen": 168155110,
+      "step": 7883
+    },
+    {
+      "epoch": 0.9479949497985931,
+      "grad_norm": 2.9710440866155983,
+      "learning_rate": 2.824398440193998e-08,
+      "loss": 0.667,
+      "num_input_tokens_seen": 168173630,
+      "step": 7884
+    },
+    {
+      "epoch": 0.9481151926892323,
+      "grad_norm": 3.345057298641073,
+      "learning_rate": 2.811368341124232e-08,
+      "loss": 0.7059,
+      "num_input_tokens_seen": 168192420,
+      "step": 7885
+    },
+    {
+      "epoch": 0.9482354355798713,
+      "grad_norm": 4.299865995872961,
+      "learning_rate": 2.7983681554131222e-08,
+      "loss": 0.6728,
+      "num_input_tokens_seen": 168212400,
+      "step": 7886
+    },
+    {
+      "epoch": 0.9483556784705104,
+      "grad_norm": 3.42910105571297,
+      "learning_rate": 2.7853978850327365e-08,
+      "loss": 0.692,
+      "num_input_tokens_seen": 168231290,
+      "step": 7887
+    },
+    {
+      "epoch": 0.9484759213611496,
+      "grad_norm": 1.9063330145742587,
+      "learning_rate": 2.7724575319507225e-08,
+      "loss": 0.8663,
+      "num_input_tokens_seen": 168250720,
+      "step": 7888
+    },
+    {
+      "epoch": 0.9485961642517886,
+      "grad_norm": 1.8641401795724934,
+      "learning_rate": 2.759547098130066e-08,
+      "loss": 0.7735,
+      "num_input_tokens_seen": 168269170,
+      "step": 7889
+    },
+    {
+      "epoch": 0.9487164071424277,
+      "grad_norm": 2.137271093436953,
+      "learning_rate": 2.746666585529267e-08,
+      "loss": 0.7615,
+      "num_input_tokens_seen": 168289165,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9488366500330668,
+      "grad_norm": 2.2389786577764266,
+      "learning_rate": 2.73381599610234e-08,
+      "loss": 0.7319,
+      "num_input_tokens_seen": 168309285,
+      "step": 7891
+    },
+    {
+      "epoch": 0.9489568929237059,
+      "grad_norm": 2.0373743053437936,
+      "learning_rate": 2.7209953317987033e-08,
+      "loss": 0.7076,
+      "num_input_tokens_seen": 168330045,
+      "step": 7892
+    },
+    {
+      "epoch": 0.9490771358143449,
+      "grad_norm": 2.259652604021654,
+      "learning_rate": 2.7082045945631793e-08,
+      "loss": 0.7655,
+      "num_input_tokens_seen": 168351980,
+      "step": 7893
+    },
+    {
+      "epoch": 0.9491973787049841,
+      "grad_norm": 2.7159464477700173,
+      "learning_rate": 2.6954437863361712e-08,
+      "loss": 0.6915,
+      "num_input_tokens_seen": 168369615,
+      "step": 7894
+    },
+    {
+      "epoch": 0.9493176215956232,
+      "grad_norm": 7.62534997736172,
+      "learning_rate": 2.682712909053464e-08,
+      "loss": 0.7091,
+      "num_input_tokens_seen": 168389635,
+      "step": 7895
+    },
+    {
+      "epoch": 0.9494378644862622,
+      "grad_norm": 2.098902253118414,
+      "learning_rate": 2.670011964646335e-08,
+      "loss": 0.7761,
+      "num_input_tokens_seen": 168408035,
+      "step": 7896
+    },
+    {
+      "epoch": 0.9495581073769014,
+      "grad_norm": 2.420571720420862,
+      "learning_rate": 2.657340955041487e-08,
+      "loss": 0.6844,
+      "num_input_tokens_seen": 168426530,
+      "step": 7897
+    },
+    {
+      "epoch": 0.9496783502675404,
+      "grad_norm": 2.2621318880658197,
+      "learning_rate": 2.6446998821610945e-08,
+      "loss": 0.7103,
+      "num_input_tokens_seen": 168446445,
+      "step": 7898
+    },
+    {
+      "epoch": 0.9497985931581795,
+      "grad_norm": 3.5804442966929555,
+      "learning_rate": 2.632088747922845e-08,
+      "loss": 0.7121,
+      "num_input_tokens_seen": 168462765,
+      "step": 7899
+    },
+    {
+      "epoch": 0.9499188360488187,
+      "grad_norm": 3.055979145460257,
+      "learning_rate": 2.619507554239786e-08,
+      "loss": 0.7175,
+      "num_input_tokens_seen": 168481045,
+      "step": 7900
+    },
+    {
+      "epoch": 0.9500390789394577,
+      "grad_norm": 2.0833230995363703,
+      "learning_rate": 2.606956303020502e-08,
+      "loss": 0.6992,
+      "num_input_tokens_seen": 168501570,
+      "step": 7901
+    },
+    {
+      "epoch": 0.9501593218300968,
+      "grad_norm": 1.8408385206878377,
+      "learning_rate": 2.5944349961690036e-08,
+      "loss": 0.8311,
+      "num_input_tokens_seen": 168518310,
+      "step": 7902
+    },
+    {
+      "epoch": 0.9502795647207359,
+      "grad_norm": 1.7946286642215257,
+      "learning_rate": 2.581943635584749e-08,
+      "loss": 0.7257,
+      "num_input_tokens_seen": 168540860,
+      "step": 7903
+    },
+    {
+      "epoch": 0.950399807611375,
+      "grad_norm": 2.104113168371056,
+      "learning_rate": 2.569482223162689e-08,
+      "loss": 0.6541,
+      "num_input_tokens_seen": 168564555,
+      "step": 7904
+    },
+    {
+      "epoch": 0.950520050502014,
+      "grad_norm": 1.831970054004136,
+      "learning_rate": 2.5570507607932e-08,
+      "loss": 0.7189,
+      "num_input_tokens_seen": 168584190,
+      "step": 7905
+    },
+    {
+      "epoch": 0.9506402933926532,
+      "grad_norm": 3.36053132209125,
+      "learning_rate": 2.54464925036213e-08,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 168601200,
+      "step": 7906
+    },
+    {
+      "epoch": 0.9507605362832923,
+      "grad_norm": 2.5850478358333913,
+      "learning_rate": 2.532277693750773e-08,
+      "loss": 0.597,
+      "num_input_tokens_seen": 168621845,
+      "step": 7907
+    },
+    {
+      "epoch": 0.9508807791739313,
+      "grad_norm": 4.271286290063512,
+      "learning_rate": 2.5199360928358726e-08,
+      "loss": 0.7499,
+      "num_input_tokens_seen": 168638800,
+      "step": 7908
+    },
+    {
+      "epoch": 0.9510010220645704,
+      "grad_norm": 2.051570321444406,
+      "learning_rate": 2.507624449489665e-08,
+      "loss": 0.862,
+      "num_input_tokens_seen": 168657150,
+      "step": 7909
+    },
+    {
+      "epoch": 0.9511212649552095,
+      "grad_norm": 4.908450471661576,
+      "learning_rate": 2.4953427655797888e-08,
+      "loss": 0.6457,
+      "num_input_tokens_seen": 168675530,
+      "step": 7910
+    },
+    {
+      "epoch": 0.9512415078458486,
+      "grad_norm": 1.845008486164713,
+      "learning_rate": 2.4830910429693984e-08,
+      "loss": 0.7082,
+      "num_input_tokens_seen": 168695210,
+      "step": 7911
+    },
+    {
+      "epoch": 0.9513617507364877,
+      "grad_norm": 2.2125042004841315,
+      "learning_rate": 2.470869283517052e-08,
+      "loss": 0.7845,
+      "num_input_tokens_seen": 168712965,
+      "step": 7912
+    },
+    {
+      "epoch": 0.9514819936271268,
+      "grad_norm": 1.8260647504810248,
+      "learning_rate": 2.458677489076777e-08,
+      "loss": 0.7666,
+      "num_input_tokens_seen": 168733695,
+      "step": 7913
+    },
+    {
+      "epoch": 0.9516022365177659,
+      "grad_norm": 1.8879815643061244,
+      "learning_rate": 2.446515661498072e-08,
+      "loss": 0.8238,
+      "num_input_tokens_seen": 168752745,
+      "step": 7914
+    },
+    {
+      "epoch": 0.9517224794084049,
+      "grad_norm": 5.020952357132159,
+      "learning_rate": 2.434383802625861e-08,
+      "loss": 0.735,
+      "num_input_tokens_seen": 168771420,
+      "step": 7915
+    },
+    {
+      "epoch": 0.9518427222990441,
+      "grad_norm": 2.288130826957981,
+      "learning_rate": 2.422281914300539e-08,
+      "loss": 0.7349,
+      "num_input_tokens_seen": 168790735,
+      "step": 7916
+    },
+    {
+      "epoch": 0.9519629651896832,
+      "grad_norm": 2.1950592436228598,
+      "learning_rate": 2.4102099983579706e-08,
+      "loss": 0.8019,
+      "num_input_tokens_seen": 168809605,
+      "step": 7917
+    },
+    {
+      "epoch": 0.9520832080803222,
+      "grad_norm": 1.8318396159718264,
+      "learning_rate": 2.3981680566294236e-08,
+      "loss": 0.7549,
+      "num_input_tokens_seen": 168828925,
+      "step": 7918
+    },
+    {
+      "epoch": 0.9522034509709614,
+      "grad_norm": 2.0990833969288776,
+      "learning_rate": 2.3861560909416822e-08,
+      "loss": 0.7312,
+      "num_input_tokens_seen": 168848195,
+      "step": 7919
+    },
+    {
+      "epoch": 0.9523236938616004,
+      "grad_norm": 2.3775611187163412,
+      "learning_rate": 2.3741741031169325e-08,
+      "loss": 0.8179,
+      "num_input_tokens_seen": 168867485,
+      "step": 7920
+    },
+    {
+      "epoch": 0.9524439367522395,
+      "grad_norm": 2.161232264195119,
+      "learning_rate": 2.3622220949728544e-08,
+      "loss": 0.7035,
+      "num_input_tokens_seen": 168886090,
+      "step": 7921
+    },
+    {
+      "epoch": 0.9525641796428787,
+      "grad_norm": 3.072965574445768,
+      "learning_rate": 2.3503000683225304e-08,
+      "loss": 0.6154,
+      "num_input_tokens_seen": 168903525,
+      "step": 7922
+    },
+    {
+      "epoch": 0.9526844225335177,
+      "grad_norm": 2.4445056856672696,
+      "learning_rate": 2.3384080249745585e-08,
+      "loss": 0.8349,
+      "num_input_tokens_seen": 168921135,
+      "step": 7923
+    },
+    {
+      "epoch": 0.9528046654241568,
+      "grad_norm": 2.8082466760971396,
+      "learning_rate": 2.3265459667329178e-08,
+      "loss": 0.8351,
+      "num_input_tokens_seen": 168940345,
+      "step": 7924
+    },
+    {
+      "epoch": 0.9529249083147959,
+      "grad_norm": 3.2794835538733422,
+      "learning_rate": 2.31471389539708e-08,
+      "loss": 0.8589,
+      "num_input_tokens_seen": 168957190,
+      "step": 7925
+    },
+    {
+      "epoch": 0.953045151205435,
+      "grad_norm": 2.3487699881912922,
+      "learning_rate": 2.302911812761965e-08,
+      "loss": 0.7259,
+      "num_input_tokens_seen": 168976625,
+      "step": 7926
+    },
+    {
+      "epoch": 0.953165394096074,
+      "grad_norm": 2.244578565167704,
+      "learning_rate": 2.2911397206179628e-08,
+      "loss": 0.8589,
+      "num_input_tokens_seen": 168993095,
+      "step": 7927
+    },
+    {
+      "epoch": 0.9532856369867132,
+      "grad_norm": 5.287023861445611,
+      "learning_rate": 2.279397620750845e-08,
+      "loss": 0.6197,
+      "num_input_tokens_seen": 169011860,
+      "step": 7928
+    },
+    {
+      "epoch": 0.9534058798773523,
+      "grad_norm": 3.2592722599664192,
+      "learning_rate": 2.2676855149419195e-08,
+      "loss": 0.7823,
+      "num_input_tokens_seen": 169028750,
+      "step": 7929
+    },
+    {
+      "epoch": 0.9535261227679913,
+      "grad_norm": 2.4184075683096546,
+      "learning_rate": 2.2560034049678766e-08,
+      "loss": 0.7406,
+      "num_input_tokens_seen": 169042820,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9536463656586305,
+      "grad_norm": 2.767569650434561,
+      "learning_rate": 2.2443512926008988e-08,
+      "loss": 0.7494,
+      "num_input_tokens_seen": 169061870,
+      "step": 7931
+    },
+    {
+      "epoch": 0.9537666085492695,
+      "grad_norm": 2.571659299413914,
+      "learning_rate": 2.2327291796085946e-08,
+      "loss": 0.7,
+      "num_input_tokens_seen": 169079950,
+      "step": 7932
+    },
+    {
+      "epoch": 0.9538868514399086,
+      "grad_norm": 4.5221436510617705,
+      "learning_rate": 2.221137067754042e-08,
+      "loss": 0.7643,
+      "num_input_tokens_seen": 169096195,
+      "step": 7933
+    },
+    {
+      "epoch": 0.9540070943305478,
+      "grad_norm": 3.122665925840265,
+      "learning_rate": 2.2095749587957012e-08,
+      "loss": 0.7753,
+      "num_input_tokens_seen": 169113820,
+      "step": 7934
+    },
+    {
+      "epoch": 0.9541273372211868,
+      "grad_norm": 2.371007277750419,
+      "learning_rate": 2.1980428544876138e-08,
+      "loss": 0.6927,
+      "num_input_tokens_seen": 169132180,
+      "step": 7935
+    },
+    {
+      "epoch": 0.9542475801118259,
+      "grad_norm": 1.7739209527619382,
+      "learning_rate": 2.1865407565791584e-08,
+      "loss": 0.7308,
+      "num_input_tokens_seen": 169153470,
+      "step": 7936
+    },
+    {
+      "epoch": 0.954367823002465,
+      "grad_norm": 2.2624711270871396,
+      "learning_rate": 2.175068666815183e-08,
+      "loss": 0.7673,
+      "num_input_tokens_seen": 169174030,
+      "step": 7937
+    },
+    {
+      "epoch": 0.9544880658931041,
+      "grad_norm": 2.5346677480768722,
+      "learning_rate": 2.163626586935985e-08,
+      "loss": 0.7883,
+      "num_input_tokens_seen": 169190290,
+      "step": 7938
+    },
+    {
+      "epoch": 0.9546083087837431,
+      "grad_norm": 2.474563057907463,
+      "learning_rate": 2.1522145186773533e-08,
+      "loss": 0.6266,
+      "num_input_tokens_seen": 169208930,
+      "step": 7939
+    },
+    {
+      "epoch": 0.9547285516743822,
+      "grad_norm": 2.131475109895167,
+      "learning_rate": 2.140832463770481e-08,
+      "loss": 0.8506,
+      "num_input_tokens_seen": 169227845,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9548487945650214,
+      "grad_norm": 8.565388463467404,
+      "learning_rate": 2.129480423941987e-08,
+      "loss": 0.7501,
+      "num_input_tokens_seen": 169244235,
+      "step": 7941
+    },
+    {
+      "epoch": 0.9549690374556604,
+      "grad_norm": 2.497887772340388,
+      "learning_rate": 2.1181584009140052e-08,
+      "loss": 0.8,
+      "num_input_tokens_seen": 169263495,
+      "step": 7942
+    },
+    {
+      "epoch": 0.9550892803462995,
+      "grad_norm": 2.6260386267379134,
+      "learning_rate": 2.10686639640405e-08,
+      "loss": 0.8307,
+      "num_input_tokens_seen": 169277305,
+      "step": 7943
+    },
+    {
+      "epoch": 0.9552095232369386,
+      "grad_norm": 2.307565353533796,
+      "learning_rate": 2.0956044121251294e-08,
+      "loss": 0.8082,
+      "num_input_tokens_seen": 169295810,
+      "step": 7944
+    },
+    {
+      "epoch": 0.9553297661275777,
+      "grad_norm": 2.353113366858493,
+      "learning_rate": 2.084372449785654e-08,
+      "loss": 0.8096,
+      "num_input_tokens_seen": 169315365,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9554500090182168,
+      "grad_norm": 1.8942030976690745,
+      "learning_rate": 2.0731705110895282e-08,
+      "loss": 0.6725,
+      "num_input_tokens_seen": 169332575,
+      "step": 7946
+    },
+    {
+      "epoch": 0.9555702519088559,
+      "grad_norm": 2.0011680526915065,
+      "learning_rate": 2.0619985977360587e-08,
+      "loss": 0.8651,
+      "num_input_tokens_seen": 169350615,
+      "step": 7947
+    },
+    {
+      "epoch": 0.955690494799495,
+      "grad_norm": 2.0123845546260317,
+      "learning_rate": 2.0508567114200237e-08,
+      "loss": 0.7664,
+      "num_input_tokens_seen": 169370250,
+      "step": 7948
+    },
+    {
+      "epoch": 0.955810737690134,
+      "grad_norm": 2.8528708311849935,
+      "learning_rate": 2.0397448538316485e-08,
+      "loss": 0.7735,
+      "num_input_tokens_seen": 169391010,
+      "step": 7949
+    },
+    {
+      "epoch": 0.9559309805807732,
+      "grad_norm": 2.1818762262611218,
+      "learning_rate": 2.028663026656563e-08,
+      "loss": 0.6527,
+      "num_input_tokens_seen": 169409585,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9560512234714122,
+      "grad_norm": 2.187762253747632,
+      "learning_rate": 2.0176112315758885e-08,
+      "loss": 0.7079,
+      "num_input_tokens_seen": 169427095,
+      "step": 7951
+    },
+    {
+      "epoch": 0.9561714663620513,
+      "grad_norm": 3.565405894107871,
+      "learning_rate": 2.0065894702661957e-08,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 169443490,
+      "step": 7952
+    },
+    {
+      "epoch": 0.9562917092526905,
+      "grad_norm": 2.251478513492884,
+      "learning_rate": 1.9955977443994577e-08,
+      "loss": 0.7781,
+      "num_input_tokens_seen": 169463200,
+      "step": 7953
+    },
+    {
+      "epoch": 0.9564119521433295,
+      "grad_norm": 2.674929553048704,
+      "learning_rate": 1.9846360556430965e-08,
+      "loss": 0.6315,
+      "num_input_tokens_seen": 169481220,
+      "step": 7954
+    },
+    {
+      "epoch": 0.9565321950339686,
+      "grad_norm": 2.6161275933809938,
+      "learning_rate": 1.973704405660004e-08,
+      "loss": 0.614,
+      "num_input_tokens_seen": 169502055,
+      "step": 7955
+    },
+    {
+      "epoch": 0.9566524379246077,
+      "grad_norm": 1.8275144177077192,
+      "learning_rate": 1.9628027961085203e-08,
+      "loss": 0.7686,
+      "num_input_tokens_seen": 169525005,
+      "step": 7956
+    },
+    {
+      "epoch": 0.9567726808152468,
+      "grad_norm": 1.971944240910578,
+      "learning_rate": 1.9519312286423894e-08,
+      "loss": 0.8267,
+      "num_input_tokens_seen": 169547920,
+      "step": 7957
+    },
+    {
+      "epoch": 0.9568929237058859,
+      "grad_norm": 4.072879737707969,
+      "learning_rate": 1.9410897049108255e-08,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 169566920,
+      "step": 7958
+    },
+    {
+      "epoch": 0.957013166596525,
+      "grad_norm": 2.1094889135384163,
+      "learning_rate": 1.9302782265584905e-08,
+      "loss": 0.9061,
+      "num_input_tokens_seen": 169587305,
+      "step": 7959
+    },
+    {
+      "epoch": 0.9571334094871641,
+      "grad_norm": 2.391583435135503,
+      "learning_rate": 1.9194967952254504e-08,
+      "loss": 0.8609,
+      "num_input_tokens_seen": 169605600,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9572536523778031,
+      "grad_norm": 2.4597325502862613,
+      "learning_rate": 1.9087454125472635e-08,
+      "loss": 0.7983,
+      "num_input_tokens_seen": 169619795,
+      "step": 7961
+    },
+    {
+      "epoch": 0.9573738952684423,
+      "grad_norm": 2.1077675693187388,
+      "learning_rate": 1.8980240801548696e-08,
+      "loss": 0.7799,
+      "num_input_tokens_seen": 169638705,
+      "step": 7962
+    },
+    {
+      "epoch": 0.9574941381590814,
+      "grad_norm": 1.8438501368236029,
+      "learning_rate": 1.8873327996747458e-08,
+      "loss": 0.7348,
+      "num_input_tokens_seen": 169656925,
+      "step": 7963
+    },
+    {
+      "epoch": 0.9576143810497204,
+      "grad_norm": 2.150654740561062,
+      "learning_rate": 1.8766715727287053e-08,
+      "loss": 0.6595,
+      "num_input_tokens_seen": 169678350,
+      "step": 7964
+    },
+    {
+      "epoch": 0.9577346239403596,
+      "grad_norm": 1.8844785279048308,
+      "learning_rate": 1.8660404009340546e-08,
+      "loss": 0.7811,
+      "num_input_tokens_seen": 169698520,
+      "step": 7965
+    },
+    {
+      "epoch": 0.9578548668309986,
+      "grad_norm": 0.9324936308729534,
+      "learning_rate": 1.8554392859035485e-08,
+      "loss": 0.664,
+      "num_input_tokens_seen": 169755990,
+      "step": 7966
+    },
+    {
+      "epoch": 0.9579751097216377,
+      "grad_norm": 1.919559263784753,
+      "learning_rate": 1.8448682292453444e-08,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 169774785,
+      "step": 7967
+    },
+    {
+      "epoch": 0.9580953526122769,
+      "grad_norm": 2.4468980376831193,
+      "learning_rate": 1.8343272325631154e-08,
+      "loss": 0.6675,
+      "num_input_tokens_seen": 169793450,
+      "step": 7968
+    },
+    {
+      "epoch": 0.9582155955029159,
+      "grad_norm": 2.9481665710402356,
+      "learning_rate": 1.8238162974558492e-08,
+      "loss": 0.7684,
+      "num_input_tokens_seen": 169807100,
+      "step": 7969
+    },
+    {
+      "epoch": 0.958335838393555,
+      "grad_norm": 2.0919724748888733,
+      "learning_rate": 1.8133354255180922e-08,
+      "loss": 0.7409,
+      "num_input_tokens_seen": 169827135,
+      "step": 7970
+    },
+    {
+      "epoch": 0.958456081284194,
+      "grad_norm": 3.1962032399822204,
+      "learning_rate": 1.8028846183397727e-08,
+      "loss": 0.7532,
+      "num_input_tokens_seen": 169845660,
+      "step": 7971
+    },
+    {
+      "epoch": 0.9585763241748332,
+      "grad_norm": 5.430738491819635,
+      "learning_rate": 1.7924638775062894e-08,
+      "loss": 0.8062,
+      "num_input_tokens_seen": 169864500,
+      "step": 7972
+    },
+    {
+      "epoch": 0.9586965670654722,
+      "grad_norm": 1.9866314431317613,
+      "learning_rate": 1.7820732045984444e-08,
+      "loss": 0.8141,
+      "num_input_tokens_seen": 169884365,
+      "step": 7973
+    },
+    {
+      "epoch": 0.9588168099561113,
+      "grad_norm": 1.9135452002848408,
+      "learning_rate": 1.7717126011924655e-08,
+      "loss": 0.7381,
+      "num_input_tokens_seen": 169905670,
+      "step": 7974
+    },
+    {
+      "epoch": 0.9589370528467505,
+      "grad_norm": 3.0833277736629165,
+      "learning_rate": 1.7613820688600957e-08,
+      "loss": 0.7561,
+      "num_input_tokens_seen": 169921295,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9590572957373895,
+      "grad_norm": 2.8037767955591897,
+      "learning_rate": 1.7510816091684588e-08,
+      "loss": 0.7831,
+      "num_input_tokens_seen": 169940940,
+      "step": 7976
+    },
+    {
+      "epoch": 0.9591775386280286,
+      "grad_norm": 3.1708717613445248,
+      "learning_rate": 1.7408112236801053e-08,
+      "loss": 0.7924,
+      "num_input_tokens_seen": 169957515,
+      "step": 7977
+    },
+    {
+      "epoch": 0.9592977815186677,
+      "grad_norm": 13.37495886578553,
+      "learning_rate": 1.7305709139530334e-08,
+      "loss": 0.7396,
+      "num_input_tokens_seen": 169976015,
+      "step": 7978
+    },
+    {
+      "epoch": 0.9594180244093068,
+      "grad_norm": 2.423448386526746,
+      "learning_rate": 1.7203606815407334e-08,
+      "loss": 0.7463,
+      "num_input_tokens_seen": 169990330,
+      "step": 7979
+    },
+    {
+      "epoch": 0.9595382672999458,
+      "grad_norm": 2.0307758473377198,
+      "learning_rate": 1.7101805279920557e-08,
+      "loss": 0.7916,
+      "num_input_tokens_seen": 170008210,
+      "step": 7980
+    },
+    {
+      "epoch": 0.959658510190585,
+      "grad_norm": 3.0735222189430376,
+      "learning_rate": 1.7000304548513643e-08,
+      "loss": 0.8102,
+      "num_input_tokens_seen": 170028035,
+      "step": 7981
+    },
+    {
+      "epoch": 0.9597787530812241,
+      "grad_norm": 2.580253512112275,
+      "learning_rate": 1.6899104636583394e-08,
+      "loss": 0.812,
+      "num_input_tokens_seen": 170045805,
+      "step": 7982
+    },
+    {
+      "epoch": 0.9598989959718631,
+      "grad_norm": 0.763674509277108,
+      "learning_rate": 1.6798205559482638e-08,
+      "loss": 0.6557,
+      "num_input_tokens_seen": 170107905,
+      "step": 7983
+    },
+    {
+      "epoch": 0.9600192388625023,
+      "grad_norm": 2.5518254776811484,
+      "learning_rate": 1.669760733251713e-08,
+      "loss": 0.7605,
+      "num_input_tokens_seen": 170126500,
+      "step": 7984
+    },
+    {
+      "epoch": 0.9601394817531413,
+      "grad_norm": 2.094349304433427,
+      "learning_rate": 1.659730997094755e-08,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 170144710,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9602597246437804,
+      "grad_norm": 1.8968075376209723,
+      "learning_rate": 1.6497313489989283e-08,
+      "loss": 0.6127,
+      "num_input_tokens_seen": 170164255,
+      "step": 7986
+    },
+    {
+      "epoch": 0.9603799675344196,
+      "grad_norm": 3.5113935915170638,
+      "learning_rate": 1.639761790481131e-08,
+      "loss": 0.6948,
+      "num_input_tokens_seen": 170184855,
+      "step": 7987
+    },
+    {
+      "epoch": 0.9605002104250586,
+      "grad_norm": 2.3703774336825965,
+      "learning_rate": 1.6298223230537754e-08,
+      "loss": 0.7852,
+      "num_input_tokens_seen": 170202375,
+      "step": 7988
+    },
+    {
+      "epoch": 0.9606204533156977,
+      "grad_norm": 2.4605232139940822,
+      "learning_rate": 1.6199129482246333e-08,
+      "loss": 0.6846,
+      "num_input_tokens_seen": 170223300,
+      "step": 7989
+    },
+    {
+      "epoch": 0.9607406962063368,
+      "grad_norm": 3.7221899342562614,
+      "learning_rate": 1.6100336674969682e-08,
+      "loss": 0.6013,
+      "num_input_tokens_seen": 170241860,
+      "step": 7990
+    },
+    {
+      "epoch": 0.9608609390969759,
+      "grad_norm": 2.5758386934124604,
+      "learning_rate": 1.600184482369449e-08,
+      "loss": 0.766,
+      "num_input_tokens_seen": 170261495,
+      "step": 7991
+    },
+    {
+      "epoch": 0.960981181987615,
+      "grad_norm": 2.376932590939063,
+      "learning_rate": 1.5903653943362126e-08,
+      "loss": 0.894,
+      "num_input_tokens_seen": 170280210,
+      "step": 7992
+    },
+    {
+      "epoch": 0.9611014248782541,
+      "grad_norm": 2.7639449677992003,
+      "learning_rate": 1.580576404886802e-08,
+      "loss": 0.7566,
+      "num_input_tokens_seen": 170298460,
+      "step": 7993
+    },
+    {
+      "epoch": 0.9612216677688932,
+      "grad_norm": 1.9954917169327204,
+      "learning_rate": 1.570817515506162e-08,
+      "loss": 0.7937,
+      "num_input_tokens_seen": 170316870,
+      "step": 7994
+    },
+    {
+      "epoch": 0.9613419106595322,
+      "grad_norm": 2.1146164838727146,
+      "learning_rate": 1.561088727674753e-08,
+      "loss": 0.804,
+      "num_input_tokens_seen": 170330800,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9614621535501714,
+      "grad_norm": 2.8287339152070974,
+      "learning_rate": 1.551390042868417e-08,
+      "loss": 0.7035,
+      "num_input_tokens_seen": 170352290,
+      "step": 7996
+    },
+    {
+      "epoch": 0.9615823964408104,
+      "grad_norm": 2.0839700740286173,
+      "learning_rate": 1.5417214625584207e-08,
+      "loss": 0.7027,
+      "num_input_tokens_seen": 170369665,
+      "step": 7997
+    },
+    {
+      "epoch": 0.9617026393314495,
+      "grad_norm": 1.8568633179769516,
+      "learning_rate": 1.5320829882114584e-08,
+      "loss": 0.8479,
+      "num_input_tokens_seen": 170387460,
+      "step": 7998
+    },
+    {
+      "epoch": 0.9618228822220887,
+      "grad_norm": 2.121928857208219,
+      "learning_rate": 1.5224746212897378e-08,
+      "loss": 0.7788,
+      "num_input_tokens_seen": 170406475,
+      "step": 7999
+    },
+    {
+      "epoch": 0.9619431251127277,
+      "grad_norm": 1.660845754178301,
+      "learning_rate": 1.512896363250804e-08,
+      "loss": 0.7621,
+      "num_input_tokens_seen": 170426305,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9620633680033668,
+      "grad_norm": 2.331924161128743,
+      "learning_rate": 1.503348215547673e-08,
+      "loss": 0.7476,
+      "num_input_tokens_seen": 170447115,
+      "step": 8001
+    },
+    {
+      "epoch": 0.962183610894006,
+      "grad_norm": 1.8891233275701773,
+      "learning_rate": 1.4938301796288078e-08,
+      "loss": 0.7957,
+      "num_input_tokens_seen": 170463405,
+      "step": 8002
+    },
+    {
+      "epoch": 0.962303853784645,
+      "grad_norm": 2.712257359943641,
+      "learning_rate": 1.4843422569380537e-08,
+      "loss": 0.8182,
+      "num_input_tokens_seen": 170479880,
+      "step": 8003
+    },
+    {
+      "epoch": 0.9624240966752841,
+      "grad_norm": 2.1430669697397913,
+      "learning_rate": 1.4748844489147483e-08,
+      "loss": 0.8223,
+      "num_input_tokens_seen": 170496590,
+      "step": 8004
+    },
+    {
+      "epoch": 0.9625443395659231,
+      "grad_norm": 2.2377112735916644,
+      "learning_rate": 1.4654567569936326e-08,
+      "loss": 0.7039,
+      "num_input_tokens_seen": 170513885,
+      "step": 8005
+    },
+    {
+      "epoch": 0.9626645824565623,
+      "grad_norm": 2.8734296156319075,
+      "learning_rate": 1.456059182604874e-08,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 170532410,
+      "step": 8006
+    },
+    {
+      "epoch": 0.9627848253472013,
+      "grad_norm": 2.8486158705629547,
+      "learning_rate": 1.4466917271740653e-08,
+      "loss": 0.7558,
+      "num_input_tokens_seen": 170550330,
+      "step": 8007
+    },
+    {
+      "epoch": 0.9629050682378404,
+      "grad_norm": 2.3415190061961337,
+      "learning_rate": 1.4373543921222697e-08,
+      "loss": 0.6699,
+      "num_input_tokens_seen": 170569635,
+      "step": 8008
+    },
+    {
+      "epoch": 0.9630253111284796,
+      "grad_norm": 3.1784240889830198,
+      "learning_rate": 1.428047178865932e-08,
+      "loss": 0.7731,
+      "num_input_tokens_seen": 170586145,
+      "step": 8009
+    },
+    {
+      "epoch": 0.9631455540191186,
+      "grad_norm": 3.8905884786682123,
+      "learning_rate": 1.418770088816923e-08,
+      "loss": 0.7343,
+      "num_input_tokens_seen": 170605040,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9632657969097577,
+      "grad_norm": 0.833427001999858,
+      "learning_rate": 1.40952312338265e-08,
+      "loss": 0.6487,
+      "num_input_tokens_seen": 170669405,
+      "step": 8011
+    },
+    {
+      "epoch": 0.9633860398003968,
+      "grad_norm": 1.6966099832892405,
+      "learning_rate": 1.4003062839657909e-08,
+      "loss": 0.678,
+      "num_input_tokens_seen": 170691605,
+      "step": 8012
+    },
+    {
+      "epoch": 0.9635062826910359,
+      "grad_norm": 1.7189036580018384,
+      "learning_rate": 1.391119571964583e-08,
+      "loss": 0.7925,
+      "num_input_tokens_seen": 170712265,
+      "step": 8013
+    },
+    {
+      "epoch": 0.9636265255816749,
+      "grad_norm": 1.9626285073365846,
+      "learning_rate": 1.3819629887726003e-08,
+      "loss": 0.7282,
+      "num_input_tokens_seen": 170730075,
+      "step": 8014
+    },
+    {
+      "epoch": 0.9637467684723141,
+      "grad_norm": 2.4162597051472328,
+      "learning_rate": 1.3728365357789317e-08,
+      "loss": 0.763,
+      "num_input_tokens_seen": 170749160,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9638670113629532,
+      "grad_norm": 3.891502961263289,
+      "learning_rate": 1.3637402143680254e-08,
+      "loss": 0.7608,
+      "num_input_tokens_seen": 170763780,
+      "step": 8016
+    },
+    {
+      "epoch": 0.9639872542535922,
+      "grad_norm": 0.7841700082856119,
+      "learning_rate": 1.3546740259197998e-08,
+      "loss": 0.5744,
+      "num_input_tokens_seen": 170816310,
+      "step": 8017
+    },
+    {
+      "epoch": 0.9641074971442314,
+      "grad_norm": 2.339438057956562,
+      "learning_rate": 1.3456379718095989e-08,
+      "loss": 0.6929,
+      "num_input_tokens_seen": 170836445,
+      "step": 8018
+    },
+    {
+      "epoch": 0.9642277400348704,
+      "grad_norm": 0.8922137860030963,
+      "learning_rate": 1.3366320534081487e-08,
+      "loss": 0.6545,
+      "num_input_tokens_seen": 170898845,
+      "step": 8019
+    },
+    {
+      "epoch": 0.9643479829255095,
+      "grad_norm": 2.8914530739701556,
+      "learning_rate": 1.3276562720816675e-08,
+      "loss": 0.7499,
+      "num_input_tokens_seen": 170920075,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9644682258161487,
+      "grad_norm": 3.2364401260642817,
+      "learning_rate": 1.3187106291917549e-08,
+      "loss": 0.8274,
+      "num_input_tokens_seen": 170936785,
+      "step": 8021
+    },
+    {
+      "epoch": 0.9645884687067877,
+      "grad_norm": 1.9971944082193316,
+      "learning_rate": 1.3097951260954809e-08,
+      "loss": 0.6966,
+      "num_input_tokens_seen": 170954805,
+      "step": 8022
+    },
+    {
+      "epoch": 0.9647087115974268,
+      "grad_norm": 2.4455279274616686,
+      "learning_rate": 1.3009097641453192e-08,
+      "loss": 0.8022,
+      "num_input_tokens_seen": 170972375,
+      "step": 8023
+    },
+    {
+      "epoch": 0.9648289544880659,
+      "grad_norm": 1.8223879618476428,
+      "learning_rate": 1.2920545446891474e-08,
+      "loss": 0.7577,
+      "num_input_tokens_seen": 170988815,
+      "step": 8024
+    },
+    {
+      "epoch": 0.964949197378705,
+      "grad_norm": 2.0937659587575155,
+      "learning_rate": 1.2832294690703127e-08,
+      "loss": 0.6937,
+      "num_input_tokens_seen": 171007510,
+      "step": 8025
+    },
+    {
+      "epoch": 0.965069440269344,
+      "grad_norm": 2.3079748720877977,
+      "learning_rate": 1.2744345386275668e-08,
+      "loss": 0.7692,
+      "num_input_tokens_seen": 171026770,
+      "step": 8026
+    },
+    {
+      "epoch": 0.9651896831599832,
+      "grad_norm": 10.118050113344793,
+      "learning_rate": 1.265669754695109e-08,
+      "loss": 0.7771,
+      "num_input_tokens_seen": 171046060,
+      "step": 8027
+    },
+    {
+      "epoch": 0.9653099260506223,
+      "grad_norm": 4.050076615553602,
+      "learning_rate": 1.2569351186025201e-08,
+      "loss": 0.815,
+      "num_input_tokens_seen": 171064235,
+      "step": 8028
+    },
+    {
+      "epoch": 0.9654301689412613,
+      "grad_norm": 1.6154677376918176,
+      "learning_rate": 1.2482306316748737e-08,
+      "loss": 0.7453,
+      "num_input_tokens_seen": 171084400,
+      "step": 8029
+    },
+    {
+      "epoch": 0.9655504118319005,
+      "grad_norm": 2.1825478964128235,
+      "learning_rate": 1.2395562952326021e-08,
+      "loss": 0.7827,
+      "num_input_tokens_seen": 171101280,
+      "step": 8030
+    },
+    {
+      "epoch": 0.9656706547225395,
+      "grad_norm": 2.6160818897027065,
+      "learning_rate": 1.2309121105916309e-08,
+      "loss": 0.8131,
+      "num_input_tokens_seen": 171119290,
+      "step": 8031
+    },
+    {
+      "epoch": 0.9657908976131786,
+      "grad_norm": 2.454567307861694,
+      "learning_rate": 1.222298079063222e-08,
+      "loss": 0.6807,
+      "num_input_tokens_seen": 171140150,
+      "step": 8032
+    },
+    {
+      "epoch": 0.9659111405038178,
+      "grad_norm": 2.145079813702674,
+      "learning_rate": 1.2137142019541747e-08,
+      "loss": 0.724,
+      "num_input_tokens_seen": 171158425,
+      "step": 8033
+    },
+    {
+      "epoch": 0.9660313833944568,
+      "grad_norm": 2.496751086261041,
+      "learning_rate": 1.2051604805666027e-08,
+      "loss": 0.7339,
+      "num_input_tokens_seen": 171175270,
+      "step": 8034
+    },
+    {
+      "epoch": 0.9661516262850959,
+      "grad_norm": 2.232208227466493,
+      "learning_rate": 1.196636916198135e-08,
+      "loss": 0.7759,
+      "num_input_tokens_seen": 171192530,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9662718691757349,
+      "grad_norm": 2.2418081602637034,
+      "learning_rate": 1.1881435101418036e-08,
+      "loss": 0.768,
+      "num_input_tokens_seen": 171211665,
+      "step": 8036
+    },
+    {
+      "epoch": 0.9663921120663741,
+      "grad_norm": 0.7890566035850896,
+      "learning_rate": 1.1796802636860003e-08,
+      "loss": 0.7021,
+      "num_input_tokens_seen": 171279915,
+      "step": 8037
+    },
+    {
+      "epoch": 0.9665123549570132,
+      "grad_norm": 2.633004589320094,
+      "learning_rate": 1.1712471781146316e-08,
+      "loss": 0.7292,
+      "num_input_tokens_seen": 171298970,
+      "step": 8038
+    },
+    {
+      "epoch": 0.9666325978476522,
+      "grad_norm": 2.2410098803118297,
+      "learning_rate": 1.1628442547069628e-08,
+      "loss": 0.6734,
+      "num_input_tokens_seen": 171320890,
+      "step": 8039
+    },
+    {
+      "epoch": 0.9667528407382914,
+      "grad_norm": 3.9047966056289662,
+      "learning_rate": 1.1544714947377521e-08,
+      "loss": 0.7543,
+      "num_input_tokens_seen": 171338295,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9668730836289304,
+      "grad_norm": 4.617361842263025,
+      "learning_rate": 1.1461288994770945e-08,
+      "loss": 0.699,
+      "num_input_tokens_seen": 171357090,
+      "step": 8041
+    },
+    {
+      "epoch": 0.9669933265195695,
+      "grad_norm": 1.8365444211448254,
+      "learning_rate": 1.1378164701905778e-08,
+      "loss": 0.7725,
+      "num_input_tokens_seen": 171378575,
+      "step": 8042
+    },
+    {
+      "epoch": 0.9671135694102087,
+      "grad_norm": 2.049513339049453,
+      "learning_rate": 1.1295342081392156e-08,
+      "loss": 0.6593,
+      "num_input_tokens_seen": 171397655,
+      "step": 8043
+    },
+    {
+      "epoch": 0.9672338123008477,
+      "grad_norm": 1.8331255606877064,
+      "learning_rate": 1.1212821145793804e-08,
+      "loss": 0.6853,
+      "num_input_tokens_seen": 171416990,
+      "step": 8044
+    },
+    {
+      "epoch": 0.9673540551914868,
+      "grad_norm": 2.105240127533223,
+      "learning_rate": 1.1130601907629156e-08,
+      "loss": 0.7801,
+      "num_input_tokens_seen": 171434440,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9674742980821259,
+      "grad_norm": 0.8652868216238374,
+      "learning_rate": 1.1048684379370899e-08,
+      "loss": 0.6807,
+      "num_input_tokens_seen": 171494845,
+      "step": 8046
+    },
+    {
+      "epoch": 0.967594540972765,
+      "grad_norm": 2.3931982480796345,
+      "learning_rate": 1.0967068573445759e-08,
+      "loss": 0.7433,
+      "num_input_tokens_seen": 171512050,
+      "step": 8047
+    },
+    {
+      "epoch": 0.967714783863404,
+      "grad_norm": 2.633423996656163,
+      "learning_rate": 1.0885754502234945e-08,
+      "loss": 0.6513,
+      "num_input_tokens_seen": 171531430,
+      "step": 8048
+    },
+    {
+      "epoch": 0.9678350267540432,
+      "grad_norm": 2.1158135533162103,
+      "learning_rate": 1.08047421780737e-08,
+      "loss": 0.7794,
+      "num_input_tokens_seen": 171550340,
+      "step": 8049
+    },
+    {
+      "epoch": 0.9679552696446823,
+      "grad_norm": 2.312157510222528,
+      "learning_rate": 1.0724031613251305e-08,
+      "loss": 0.7414,
+      "num_input_tokens_seen": 171567960,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9680755125353213,
+      "grad_norm": 2.778771965009616,
+      "learning_rate": 1.0643622820011744e-08,
+      "loss": 0.6581,
+      "num_input_tokens_seen": 171588735,
+      "step": 8051
+    },
+    {
+      "epoch": 0.9681957554259605,
+      "grad_norm": 3.6330454600610715,
+      "learning_rate": 1.0563515810552814e-08,
+      "loss": 0.6762,
+      "num_input_tokens_seen": 171605425,
+      "step": 8052
+    },
+    {
+      "epoch": 0.9683159983165995,
+      "grad_norm": 1.5918218422794614,
+      "learning_rate": 1.0483710597026795e-08,
+      "loss": 0.7289,
+      "num_input_tokens_seen": 171625005,
+      "step": 8053
+    },
+    {
+      "epoch": 0.9684362412072386,
+      "grad_norm": 2.4468582438185225,
+      "learning_rate": 1.0404207191540004e-08,
+      "loss": 0.7333,
+      "num_input_tokens_seen": 171645180,
+      "step": 8054
+    },
+    {
+      "epoch": 0.9685564840978778,
+      "grad_norm": 2.169848576469213,
+      "learning_rate": 1.0325005606153236e-08,
+      "loss": 0.7491,
+      "num_input_tokens_seen": 171664360,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9686767269885168,
+      "grad_norm": 2.9045212662637767,
+      "learning_rate": 1.0246105852881104e-08,
+      "loss": 0.778,
+      "num_input_tokens_seen": 171679180,
+      "step": 8056
+    },
+    {
+      "epoch": 0.9687969698791559,
+      "grad_norm": 1.922016499138457,
+      "learning_rate": 1.0167507943692476e-08,
+      "loss": 0.7812,
+      "num_input_tokens_seen": 171697985,
+      "step": 8057
+    },
+    {
+      "epoch": 0.968917212769795,
+      "grad_norm": 2.5707928562644673,
+      "learning_rate": 1.008921189051093e-08,
+      "loss": 0.707,
+      "num_input_tokens_seen": 171715050,
+      "step": 8058
+    },
+    {
+      "epoch": 0.9690374556604341,
+      "grad_norm": 2.018293448098207,
+      "learning_rate": 1.0011217705213848e-08,
+      "loss": 0.7664,
+      "num_input_tokens_seen": 171732645,
+      "step": 8059
+    },
+    {
+      "epoch": 0.9691576985510731,
+      "grad_norm": 1.8592714532205175,
+      "learning_rate": 9.933525399632658e-09,
+      "loss": 0.7486,
+      "num_input_tokens_seen": 171750600,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9692779414417123,
+      "grad_norm": 2.051628554743265,
+      "learning_rate": 9.856134985553488e-09,
+      "loss": 0.6485,
+      "num_input_tokens_seen": 171770045,
+      "step": 8061
+    },
+    {
+      "epoch": 0.9693981843323514,
+      "grad_norm": 2.3525745578273565,
+      "learning_rate": 9.77904647471628e-09,
+      "loss": 0.7299,
+      "num_input_tokens_seen": 171792945,
+      "step": 8062
+    },
+    {
+      "epoch": 0.9695184272229904,
+      "grad_norm": 2.18426259739792,
+      "learning_rate": 9.702259878815454e-09,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 171812990,
+      "step": 8063
+    },
+    {
+      "epoch": 0.9696386701136296,
+      "grad_norm": 3.2437377377041323,
+      "learning_rate": 9.625775209499254e-09,
+      "loss": 0.74,
+      "num_input_tokens_seen": 171832715,
+      "step": 8064
+    },
+    {
+      "epoch": 0.9697589130042686,
+      "grad_norm": 2.1817471257718837,
+      "learning_rate": 9.549592478370394e-09,
+      "loss": 0.7373,
+      "num_input_tokens_seen": 171850615,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9698791558949077,
+      "grad_norm": 2.17472122246022,
+      "learning_rate": 9.473711696985632e-09,
+      "loss": 0.7955,
+      "num_input_tokens_seen": 171869665,
+      "step": 8066
+    },
+    {
+      "epoch": 0.9699993987855468,
+      "grad_norm": 2.5050576842799046,
+      "learning_rate": 9.398132876856201e-09,
+      "loss": 0.7515,
+      "num_input_tokens_seen": 171888350,
+      "step": 8067
+    },
+    {
+      "epoch": 0.9701196416761859,
+      "grad_norm": 0.7904499167935078,
+      "learning_rate": 9.322856029447379e-09,
+      "loss": 0.6453,
+      "num_input_tokens_seen": 171949255,
+      "step": 8068
+    },
+    {
+      "epoch": 0.970239884566825,
+      "grad_norm": 3.8226679822290865,
+      "learning_rate": 9.247881166178695e-09,
+      "loss": 0.7931,
+      "num_input_tokens_seen": 171967685,
+      "step": 8069
+    },
+    {
+      "epoch": 0.970360127457464,
+      "grad_norm": 2.8873106668989172,
+      "learning_rate": 9.173208298423274e-09,
+      "loss": 0.7713,
+      "num_input_tokens_seen": 171988610,
+      "step": 8070
+    },
+    {
+      "epoch": 0.9704803703481032,
+      "grad_norm": 1.9418277279599057,
+      "learning_rate": 9.098837437509389e-09,
+      "loss": 0.7564,
+      "num_input_tokens_seen": 172011220,
+      "step": 8071
+    },
+    {
+      "epoch": 0.9706006132387422,
+      "grad_norm": 1.7997187337770832,
+      "learning_rate": 9.024768594719124e-09,
+      "loss": 0.8343,
+      "num_input_tokens_seen": 172029320,
+      "step": 8072
+    },
+    {
+      "epoch": 0.9707208561293813,
+      "grad_norm": 2.4917085762968383,
+      "learning_rate": 8.95100178128816e-09,
+      "loss": 0.7198,
+      "num_input_tokens_seen": 172048180,
+      "step": 8073
+    },
+    {
+      "epoch": 0.9708410990200205,
+      "grad_norm": 2.2418718082442832,
+      "learning_rate": 8.877537008407321e-09,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 172067950,
+      "step": 8074
+    },
+    {
+      "epoch": 0.9709613419106595,
+      "grad_norm": 4.469347936666569,
+      "learning_rate": 8.804374287221028e-09,
+      "loss": 0.6742,
+      "num_input_tokens_seen": 172088905,
+      "step": 8075
+    },
+    {
+      "epoch": 0.9710815848012986,
+      "grad_norm": 1.6330953968385515,
+      "learning_rate": 8.731513628827958e-09,
+      "loss": 0.8399,
+      "num_input_tokens_seen": 172107990,
+      "step": 8076
+    },
+    {
+      "epoch": 0.9712018276919377,
+      "grad_norm": 2.041194962644168,
+      "learning_rate": 8.658955044281047e-09,
+      "loss": 0.8179,
+      "num_input_tokens_seen": 172126635,
+      "step": 8077
+    },
+    {
+      "epoch": 0.9713220705825768,
+      "grad_norm": 1.7451976900278796,
+      "learning_rate": 8.586698544587268e-09,
+      "loss": 0.7642,
+      "num_input_tokens_seen": 172147965,
+      "step": 8078
+    },
+    {
+      "epoch": 0.9714423134732159,
+      "grad_norm": 2.08309210406066,
+      "learning_rate": 8.514744140707853e-09,
+      "loss": 0.7359,
+      "num_input_tokens_seen": 172166825,
+      "step": 8079
+    },
+    {
+      "epoch": 0.971562556363855,
+      "grad_norm": 1.87159604320456,
+      "learning_rate": 8.443091843558515e-09,
+      "loss": 0.7573,
+      "num_input_tokens_seen": 172185630,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9716827992544941,
+      "grad_norm": 2.6323068351592824,
+      "learning_rate": 8.37174166400878e-09,
+      "loss": 0.6443,
+      "num_input_tokens_seen": 172200925,
+      "step": 8081
+    },
+    {
+      "epoch": 0.9718030421451331,
+      "grad_norm": 2.196851156171053,
+      "learning_rate": 8.300693612881992e-09,
+      "loss": 0.8477,
+      "num_input_tokens_seen": 172220710,
+      "step": 8082
+    },
+    {
+      "epoch": 0.9719232850357723,
+      "grad_norm": 2.5740158929288337,
+      "learning_rate": 8.22994770095664e-09,
+      "loss": 0.8008,
+      "num_input_tokens_seen": 172239005,
+      "step": 8083
+    },
+    {
+      "epoch": 0.9720435279264114,
+      "grad_norm": 2.593024829472518,
+      "learning_rate": 8.159503938964585e-09,
+      "loss": 0.7484,
+      "num_input_tokens_seen": 172256045,
+      "step": 8084
+    },
+    {
+      "epoch": 0.9721637708170504,
+      "grad_norm": 1.7659990580690959,
+      "learning_rate": 8.089362337592164e-09,
+      "loss": 0.7007,
+      "num_input_tokens_seen": 172279390,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9722840137076896,
+      "grad_norm": 1.9805584641378822,
+      "learning_rate": 8.019522907479536e-09,
+      "loss": 0.7165,
+      "num_input_tokens_seen": 172299470,
+      "step": 8086
+    },
+    {
+      "epoch": 0.9724042565983286,
+      "grad_norm": 2.135805961351635,
+      "learning_rate": 7.949985659221558e-09,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 172316455,
+      "step": 8087
+    },
+    {
+      "epoch": 0.9725244994889677,
+      "grad_norm": 2.2411719132482713,
+      "learning_rate": 7.880750603366904e-09,
+      "loss": 0.7826,
+      "num_input_tokens_seen": 172335045,
+      "step": 8088
+    },
+    {
+      "epoch": 0.9726447423796069,
+      "grad_norm": 2.420216177479266,
+      "learning_rate": 7.811817750418282e-09,
+      "loss": 0.7917,
+      "num_input_tokens_seen": 172353525,
+      "step": 8089
+    },
+    {
+      "epoch": 0.9727649852702459,
+      "grad_norm": 2.1522786532070413,
+      "learning_rate": 7.743187110833105e-09,
+      "loss": 0.792,
+      "num_input_tokens_seen": 172376005,
+      "step": 8090
+    },
+    {
+      "epoch": 0.972885228160885,
+      "grad_norm": 1.7732818278657758,
+      "learning_rate": 7.674858695022602e-09,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 172394080,
+      "step": 8091
+    },
+    {
+      "epoch": 0.9730054710515241,
+      "grad_norm": 2.6817188526706093,
+      "learning_rate": 7.606832513351591e-09,
+      "loss": 0.7572,
+      "num_input_tokens_seen": 172411750,
+      "step": 8092
+    },
+    {
+      "epoch": 0.9731257139421632,
+      "grad_norm": 0.8172461126494237,
+      "learning_rate": 7.539108576140264e-09,
+      "loss": 0.6763,
+      "num_input_tokens_seen": 172475580,
+      "step": 8093
+    },
+    {
+      "epoch": 0.9732459568328022,
+      "grad_norm": 2.8911897303082528,
+      "learning_rate": 7.471686893661732e-09,
+      "loss": 0.6897,
+      "num_input_tokens_seen": 172493595,
+      "step": 8094
+    },
+    {
+      "epoch": 0.9733661997234414,
+      "grad_norm": 2.151796859380502,
+      "learning_rate": 7.4045674761442636e-09,
+      "loss": 0.6357,
+      "num_input_tokens_seen": 172510645,
+      "step": 8095
+    },
+    {
+      "epoch": 0.9734864426140805,
+      "grad_norm": 2.183968291740034,
+      "learning_rate": 7.337750333769488e-09,
+      "loss": 0.7258,
+      "num_input_tokens_seen": 172530170,
+      "step": 8096
+    },
+    {
+      "epoch": 0.9736066855047195,
+      "grad_norm": 1.7555191558854748,
+      "learning_rate": 7.2712354766737425e-09,
+      "loss": 0.7269,
+      "num_input_tokens_seen": 172550220,
+      "step": 8097
+    },
+    {
+      "epoch": 0.9737269283953586,
+      "grad_norm": 2.02609369359964,
+      "learning_rate": 7.2050229149469565e-09,
+      "loss": 0.7998,
+      "num_input_tokens_seen": 172569950,
+      "step": 8098
+    },
+    {
+      "epoch": 0.9738471712859977,
+      "grad_norm": 2.188116142239428,
+      "learning_rate": 7.139112658633984e-09,
+      "loss": 0.6264,
+      "num_input_tokens_seen": 172589820,
+      "step": 8099
+    },
+    {
+      "epoch": 0.9739674141766368,
+      "grad_norm": 2.079628933827158,
+      "learning_rate": 7.073504717733048e-09,
+      "loss": 0.6986,
+      "num_input_tokens_seen": 172609105,
+      "step": 8100
+    },
+    {
+      "epoch": 0.9740876570672758,
+      "grad_norm": 0.7790619474049189,
+      "learning_rate": 7.008199102196855e-09,
+      "loss": 0.5921,
+      "num_input_tokens_seen": 172670250,
+      "step": 8101
+    },
+    {
+      "epoch": 0.974207899957915,
+      "grad_norm": 0.8477342095568586,
+      "learning_rate": 6.9431958219321464e-09,
+      "loss": 0.6195,
+      "num_input_tokens_seen": 172726135,
+      "step": 8102
+    },
+    {
+      "epoch": 0.9743281428485541,
+      "grad_norm": 2.02706254065749,
+      "learning_rate": 6.878494886800146e-09,
+      "loss": 0.7777,
+      "num_input_tokens_seen": 172746630,
+      "step": 8103
+    },
+    {
+      "epoch": 0.9744483857391931,
+      "grad_norm": 2.753348006769035,
+      "learning_rate": 6.814096306615669e-09,
+      "loss": 0.7568,
+      "num_input_tokens_seen": 172764490,
+      "step": 8104
+    },
+    {
+      "epoch": 0.9745686286298323,
+      "grad_norm": 2.696879116593512,
+      "learning_rate": 6.750000091148234e-09,
+      "loss": 0.6456,
+      "num_input_tokens_seen": 172781505,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9746888715204713,
+      "grad_norm": 2.0069405232496274,
+      "learning_rate": 6.686206250120729e-09,
+      "loss": 0.733,
+      "num_input_tokens_seen": 172802720,
+      "step": 8106
+    },
+    {
+      "epoch": 0.9748091144111104,
+      "grad_norm": 2.1308811603440865,
+      "learning_rate": 6.622714793210749e-09,
+      "loss": 0.743,
+      "num_input_tokens_seen": 172821360,
+      "step": 8107
+    },
+    {
+      "epoch": 0.9749293573017496,
+      "grad_norm": 2.48559094604535,
+      "learning_rate": 6.559525730050364e-09,
+      "loss": 0.7749,
+      "num_input_tokens_seen": 172841180,
+      "step": 8108
+    },
+    {
+      "epoch": 0.9750496001923886,
+      "grad_norm": 1.8725943292313862,
+      "learning_rate": 6.496639070224796e-09,
+      "loss": 0.7528,
+      "num_input_tokens_seen": 172859385,
+      "step": 8109
+    },
+    {
+      "epoch": 0.9751698430830277,
+      "grad_norm": 3.3564167404108565,
+      "learning_rate": 6.4340548232739714e-09,
+      "loss": 0.824,
+      "num_input_tokens_seen": 172875305,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9752900859736668,
+      "grad_norm": 1.7237646268025921,
+      "learning_rate": 6.371772998692071e-09,
+      "loss": 0.7886,
+      "num_input_tokens_seen": 172894280,
+      "step": 8111
+    },
+    {
+      "epoch": 0.9754103288643059,
+      "grad_norm": 7.091303962462014,
+      "learning_rate": 6.309793605927094e-09,
+      "loss": 0.6467,
+      "num_input_tokens_seen": 172912320,
+      "step": 8112
+    },
+    {
+      "epoch": 0.975530571754945,
+      "grad_norm": 2.292543861897221,
+      "learning_rate": 6.248116654381297e-09,
+      "loss": 0.7944,
+      "num_input_tokens_seen": 172930510,
+      "step": 8113
+    },
+    {
+      "epoch": 0.9756508146455841,
+      "grad_norm": 2.511923641487094,
+      "learning_rate": 6.186742153410751e-09,
+      "loss": 0.7235,
+      "num_input_tokens_seen": 172949725,
+      "step": 8114
+    },
+    {
+      "epoch": 0.9757710575362232,
+      "grad_norm": 2.5089775571322877,
+      "learning_rate": 6.125670112326453e-09,
+      "loss": 0.8584,
+      "num_input_tokens_seen": 172968705,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9758913004268622,
+      "grad_norm": 2.0001391148349614,
+      "learning_rate": 6.064900540392548e-09,
+      "loss": 0.6976,
+      "num_input_tokens_seen": 172990520,
+      "step": 8116
+    },
+    {
+      "epoch": 0.9760115433175014,
+      "grad_norm": 2.852656165969772,
+      "learning_rate": 6.0044334468278835e-09,
+      "loss": 0.7909,
+      "num_input_tokens_seen": 173009585,
+      "step": 8117
+    },
+    {
+      "epoch": 0.9761317862081405,
+      "grad_norm": 2.2101290299613274,
+      "learning_rate": 5.944268840805345e-09,
+      "loss": 0.7169,
+      "num_input_tokens_seen": 173030050,
+      "step": 8118
+    },
+    {
+      "epoch": 0.9762520290987795,
+      "grad_norm": 2.8577039029166764,
+      "learning_rate": 5.88440673145163e-09,
+      "loss": 0.6348,
+      "num_input_tokens_seen": 173050820,
+      "step": 8119
+    },
+    {
+      "epoch": 0.9763722719894187,
+      "grad_norm": 2.5105576609568274,
+      "learning_rate": 5.824847127848142e-09,
+      "loss": 0.8205,
+      "num_input_tokens_seen": 173069065,
+      "step": 8120
+    },
+    {
+      "epoch": 0.9764925148800577,
+      "grad_norm": 2.2402784511274585,
+      "learning_rate": 5.765590039029433e-09,
+      "loss": 0.7782,
+      "num_input_tokens_seen": 173088105,
+      "step": 8121
+    },
+    {
+      "epoch": 0.9766127577706968,
+      "grad_norm": 1.9066163550843596,
+      "learning_rate": 5.706635473985422e-09,
+      "loss": 0.705,
+      "num_input_tokens_seen": 173111695,
+      "step": 8122
+    },
+    {
+      "epoch": 0.976733000661336,
+      "grad_norm": 2.2632997316209478,
+      "learning_rate": 5.647983441658955e-09,
+      "loss": 0.8472,
+      "num_input_tokens_seen": 173130775,
+      "step": 8123
+    },
+    {
+      "epoch": 0.976853243551975,
+      "grad_norm": 2.74453420957537,
+      "learning_rate": 5.589633950947803e-09,
+      "loss": 0.6699,
+      "num_input_tokens_seen": 173147995,
+      "step": 8124
+    },
+    {
+      "epoch": 0.9769734864426141,
+      "grad_norm": 2.2149294735564493,
+      "learning_rate": 5.5315870107035535e-09,
+      "loss": 0.6926,
+      "num_input_tokens_seen": 173165765,
+      "step": 8125
+    },
+    {
+      "epoch": 0.9770937293332532,
+      "grad_norm": 2.1642668845232995,
+      "learning_rate": 5.473842629731607e-09,
+      "loss": 0.7792,
+      "num_input_tokens_seen": 173183985,
+      "step": 8126
+    },
+    {
+      "epoch": 0.9772139722238923,
+      "grad_norm": 2.4800352386580244,
+      "learning_rate": 5.416400816792066e-09,
+      "loss": 0.7822,
+      "num_input_tokens_seen": 173201220,
+      "step": 8127
+    },
+    {
+      "epoch": 0.9773342151145313,
+      "grad_norm": 2.6603243783657478,
+      "learning_rate": 5.359261580598407e-09,
+      "loss": 0.7745,
+      "num_input_tokens_seen": 173216780,
+      "step": 8128
+    },
+    {
+      "epoch": 0.9774544580051704,
+      "grad_norm": 2.9479008075999946,
+      "learning_rate": 5.302424929819027e-09,
+      "loss": 0.7791,
+      "num_input_tokens_seen": 173230510,
+      "step": 8129
+    },
+    {
+      "epoch": 0.9775747008958096,
+      "grad_norm": 2.638121312000568,
+      "learning_rate": 5.24589087307592e-09,
+      "loss": 0.7194,
+      "num_input_tokens_seen": 173247850,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9776949437864486,
+      "grad_norm": 1.8325908802789879,
+      "learning_rate": 5.189659418944891e-09,
+      "loss": 0.6484,
+      "num_input_tokens_seen": 173277745,
+      "step": 8131
+    },
+    {
+      "epoch": 0.9778151866770877,
+      "grad_norm": 2.0465207692860234,
+      "learning_rate": 5.133730575956674e-09,
+      "loss": 0.7629,
+      "num_input_tokens_seen": 173297135,
+      "step": 8132
+    },
+    {
+      "epoch": 0.9779354295677268,
+      "grad_norm": 2.5631169101329028,
+      "learning_rate": 5.0781043525953696e-09,
+      "loss": 0.7163,
+      "num_input_tokens_seen": 173314920,
+      "step": 8133
+    },
+    {
+      "epoch": 0.9780556724583659,
+      "grad_norm": 2.387397494384778,
+      "learning_rate": 5.0227807572995605e-09,
+      "loss": 0.7244,
+      "num_input_tokens_seen": 173336615,
+      "step": 8134
+    },
+    {
+      "epoch": 0.9781759153490049,
+      "grad_norm": 2.1916600276962397,
+      "learning_rate": 4.967759798461646e-09,
+      "loss": 0.6678,
+      "num_input_tokens_seen": 173354680,
+      "step": 8135
+    },
+    {
+      "epoch": 0.9782961582396441,
+      "grad_norm": 2.9123066526660715,
+      "learning_rate": 4.913041484428282e-09,
+      "loss": 0.7351,
+      "num_input_tokens_seen": 173374875,
+      "step": 8136
+    },
+    {
+      "epoch": 0.9784164011302832,
+      "grad_norm": 2.3403729152970825,
+      "learning_rate": 4.858625823500384e-09,
+      "loss": 0.7366,
+      "num_input_tokens_seen": 173392295,
+      "step": 8137
+    },
+    {
+      "epoch": 0.9785366440209222,
+      "grad_norm": 2.5410462260144833,
+      "learning_rate": 4.80451282393246e-09,
+      "loss": 0.7344,
+      "num_input_tokens_seen": 173412000,
+      "step": 8138
+    },
+    {
+      "epoch": 0.9786568869115614,
+      "grad_norm": 2.6484215176247243,
+      "learning_rate": 4.750702493933722e-09,
+      "loss": 0.6724,
+      "num_input_tokens_seen": 173431605,
+      "step": 8139
+    },
+    {
+      "epoch": 0.9787771298022004,
+      "grad_norm": 2.4341760697815125,
+      "learning_rate": 4.697194841666974e-09,
+      "loss": 0.8481,
+      "num_input_tokens_seen": 173450250,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9788973726928395,
+      "grad_norm": 2.0426261787081046,
+      "learning_rate": 4.6439898752492764e-09,
+      "loss": 0.8124,
+      "num_input_tokens_seen": 173470110,
+      "step": 8141
+    },
+    {
+      "epoch": 0.9790176155834787,
+      "grad_norm": 0.7648630205831016,
+      "learning_rate": 4.591087602751731e-09,
+      "loss": 0.6391,
+      "num_input_tokens_seen": 173531690,
+      "step": 8142
+    },
+    {
+      "epoch": 0.9791378584741177,
+      "grad_norm": 1.777440161350849,
+      "learning_rate": 4.538488032199916e-09,
+      "loss": 0.7188,
+      "num_input_tokens_seen": 173549510,
+      "step": 8143
+    },
+    {
+      "epoch": 0.9792581013647568,
+      "grad_norm": 3.5043439688867357,
+      "learning_rate": 4.486191171572784e-09,
+      "loss": 0.6789,
+      "num_input_tokens_seen": 173566500,
+      "step": 8144
+    },
+    {
+      "epoch": 0.9793783442553959,
+      "grad_norm": 1.6145476880619811,
+      "learning_rate": 4.434197028803766e-09,
+      "loss": 0.7736,
+      "num_input_tokens_seen": 173585445,
+      "step": 8145
+    },
+    {
+      "epoch": 0.979498587146035,
+      "grad_norm": 2.2120591621396035,
+      "learning_rate": 4.38250561178033e-09,
+      "loss": 0.8134,
+      "num_input_tokens_seen": 173601050,
+      "step": 8146
+    },
+    {
+      "epoch": 0.979618830036674,
+      "grad_norm": 2.874269653901541,
+      "learning_rate": 4.331116928344425e-09,
+      "loss": 0.7989,
+      "num_input_tokens_seen": 173617085,
+      "step": 8147
+    },
+    {
+      "epoch": 0.9797390729273132,
+      "grad_norm": 2.920578558212431,
+      "learning_rate": 4.28003098629115e-09,
+      "loss": 0.6279,
+      "num_input_tokens_seen": 173632940,
+      "step": 8148
+    },
+    {
+      "epoch": 0.9798593158179523,
+      "grad_norm": 2.2739781890537714,
+      "learning_rate": 4.229247793370305e-09,
+      "loss": 0.79,
+      "num_input_tokens_seen": 173651785,
+      "step": 8149
+    },
+    {
+      "epoch": 0.9799795587085913,
+      "grad_norm": 2.022791279758367,
+      "learning_rate": 4.178767357285951e-09,
+      "loss": 0.7045,
+      "num_input_tokens_seen": 173673135,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9800998015992305,
+      "grad_norm": 2.222109900524572,
+      "learning_rate": 4.128589685695516e-09,
+      "loss": 0.7013,
+      "num_input_tokens_seen": 173693280,
+      "step": 8151
+    },
+    {
+      "epoch": 0.9802200444898695,
+      "grad_norm": 2.8337398786263743,
+      "learning_rate": 4.078714786211135e-09,
+      "loss": 0.8374,
+      "num_input_tokens_seen": 173708850,
+      "step": 8152
+    },
+    {
+      "epoch": 0.9803402873805086,
+      "grad_norm": 1.7054149618946184,
+      "learning_rate": 4.029142666398977e-09,
+      "loss": 0.7587,
+      "num_input_tokens_seen": 173728735,
+      "step": 8153
+    },
+    {
+      "epoch": 0.9804605302711478,
+      "grad_norm": 2.291803388693196,
+      "learning_rate": 3.979873333778805e-09,
+      "loss": 0.7918,
+      "num_input_tokens_seen": 173746630,
+      "step": 8154
+    },
+    {
+      "epoch": 0.9805807731617868,
+      "grad_norm": 2.5327666698827453,
+      "learning_rate": 3.930906795824862e-09,
+      "loss": 0.7396,
+      "num_input_tokens_seen": 173767025,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9807010160524259,
+      "grad_norm": 2.270367188752037,
+      "learning_rate": 3.882243059965207e-09,
+      "loss": 0.7688,
+      "num_input_tokens_seen": 173784460,
+      "step": 8156
+    },
+    {
+      "epoch": 0.980821258943065,
+      "grad_norm": 2.987200053417994,
+      "learning_rate": 3.833882133582156e-09,
+      "loss": 0.6633,
+      "num_input_tokens_seen": 173799840,
+      "step": 8157
+    },
+    {
+      "epoch": 0.9809415018337041,
+      "grad_norm": 1.750130174657618,
+      "learning_rate": 3.785824024012285e-09,
+      "loss": 0.7695,
+      "num_input_tokens_seen": 173818560,
+      "step": 8158
+    },
+    {
+      "epoch": 0.9810617447243432,
+      "grad_norm": 1.6644756653090327,
+      "learning_rate": 3.738068738545541e-09,
+      "loss": 0.7795,
+      "num_input_tokens_seen": 173837365,
+      "step": 8159
+    },
+    {
+      "epoch": 0.9811819876149822,
+      "grad_norm": 7.848257978691657,
+      "learning_rate": 3.6906162844265733e-09,
+      "loss": 0.7729,
+      "num_input_tokens_seen": 173854170,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9813022305056214,
+      "grad_norm": 2.160243322855058,
+      "learning_rate": 3.643466668853845e-09,
+      "loss": 0.7056,
+      "num_input_tokens_seen": 173871915,
+      "step": 8161
+    },
+    {
+      "epoch": 0.9814224733962604,
+      "grad_norm": 2.1169723195939536,
+      "learning_rate": 3.59661989898008e-09,
+      "loss": 0.7451,
+      "num_input_tokens_seen": 173892690,
+      "step": 8162
+    },
+    {
+      "epoch": 0.9815427162868995,
+      "grad_norm": 2.306559728903708,
+      "learning_rate": 3.5500759819115934e-09,
+      "loss": 0.7599,
+      "num_input_tokens_seen": 173912775,
+      "step": 8163
+    },
+    {
+      "epoch": 0.9816629591775387,
+      "grad_norm": 2.720169889083904,
+      "learning_rate": 3.5038349247094034e-09,
+      "loss": 0.811,
+      "num_input_tokens_seen": 173929755,
+      "step": 8164
+    },
+    {
+      "epoch": 0.9817832020681777,
+      "grad_norm": 2.4729755387539165,
+      "learning_rate": 3.4578967343878994e-09,
+      "loss": 0.7628,
+      "num_input_tokens_seen": 173945680,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9819034449588168,
+      "grad_norm": 2.0508586812392013,
+      "learning_rate": 3.4122614179161733e-09,
+      "loss": 0.7983,
+      "num_input_tokens_seen": 173965360,
+      "step": 8166
+    },
+    {
+      "epoch": 0.9820236878494559,
+      "grad_norm": 1.7245063401288443,
+      "learning_rate": 3.36692898221691e-09,
+      "loss": 0.7669,
+      "num_input_tokens_seen": 173983445,
+      "step": 8167
+    },
+    {
+      "epoch": 0.982143930740095,
+      "grad_norm": 2.0767389731463624,
+      "learning_rate": 3.3218994341668305e-09,
+      "loss": 0.7332,
+      "num_input_tokens_seen": 174002095,
+      "step": 8168
+    },
+    {
+      "epoch": 0.982264173630734,
+      "grad_norm": 1.7552384827444198,
+      "learning_rate": 3.2771727805971373e-09,
+      "loss": 0.75,
+      "num_input_tokens_seen": 174023200,
+      "step": 8169
+    },
+    {
+      "epoch": 0.9823844165213732,
+      "grad_norm": 1.9469663833398747,
+      "learning_rate": 3.232749028292847e-09,
+      "loss": 0.7666,
+      "num_input_tokens_seen": 174039885,
+      "step": 8170
+    },
+    {
+      "epoch": 0.9825046594120123,
+      "grad_norm": 1.8509515365385438,
+      "learning_rate": 3.188628183992792e-09,
+      "loss": 0.8757,
+      "num_input_tokens_seen": 174059870,
+      "step": 8171
+    },
+    {
+      "epoch": 0.9826249023026513,
+      "grad_norm": 0.7985777011380686,
+      "learning_rate": 3.1448102543902844e-09,
+      "loss": 0.655,
+      "num_input_tokens_seen": 174123505,
+      "step": 8172
+    },
+    {
+      "epoch": 0.9827451451932905,
+      "grad_norm": 2.207216559422279,
+      "learning_rate": 3.1012952461324515e-09,
+      "loss": 0.6682,
+      "num_input_tokens_seen": 174142200,
+      "step": 8173
+    },
+    {
+      "epoch": 0.9828653880839295,
+      "grad_norm": 2.136622668066744,
+      "learning_rate": 3.0580831658202354e-09,
+      "loss": 0.7395,
+      "num_input_tokens_seen": 174159500,
+      "step": 8174
+    },
+    {
+      "epoch": 0.9829856309745686,
+      "grad_norm": 2.009022276249234,
+      "learning_rate": 3.015174020009281e-09,
+      "loss": 0.7772,
+      "num_input_tokens_seen": 174178545,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9831058738652078,
+      "grad_norm": 2.930687565141807,
+      "learning_rate": 2.972567815208382e-09,
+      "loss": 0.747,
+      "num_input_tokens_seen": 174196835,
+      "step": 8176
+    },
+    {
+      "epoch": 0.9832261167558468,
+      "grad_norm": 3.464233973715276,
+      "learning_rate": 2.930264557881257e-09,
+      "loss": 0.8159,
+      "num_input_tokens_seen": 174211740,
+      "step": 8177
+    },
+    {
+      "epoch": 0.9833463596464859,
+      "grad_norm": 0.8292809613255099,
+      "learning_rate": 2.8882642544452163e-09,
+      "loss": 0.6299,
+      "num_input_tokens_seen": 174276185,
+      "step": 8178
+    },
+    {
+      "epoch": 0.983466602537125,
+      "grad_norm": 7.170591166620345,
+      "learning_rate": 2.8465669112716083e-09,
+      "loss": 0.7411,
+      "num_input_tokens_seen": 174293430,
+      "step": 8179
+    },
+    {
+      "epoch": 0.9835868454277641,
+      "grad_norm": 2.451753247009581,
+      "learning_rate": 2.8051725346858177e-09,
+      "loss": 0.7508,
+      "num_input_tokens_seen": 174313410,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9837070883184031,
+      "grad_norm": 2.3576880398151445,
+      "learning_rate": 2.7640811309674883e-09,
+      "loss": 0.7003,
+      "num_input_tokens_seen": 174332630,
+      "step": 8181
+    },
+    {
+      "epoch": 0.9838273312090423,
+      "grad_norm": 1.7012050999043942,
+      "learning_rate": 2.7232927063498557e-09,
+      "loss": 0.7945,
+      "num_input_tokens_seen": 174352725,
+      "step": 8182
+    },
+    {
+      "epoch": 0.9839475740996814,
+      "grad_norm": 2.2260371973360313,
+      "learning_rate": 2.682807267020859e-09,
+      "loss": 0.6782,
+      "num_input_tokens_seen": 174375205,
+      "step": 8183
+    },
+    {
+      "epoch": 0.9840678169903204,
+      "grad_norm": 1.6781314806344834,
+      "learning_rate": 2.642624819121808e-09,
+      "loss": 0.6214,
+      "num_input_tokens_seen": 174395075,
+      "step": 8184
+    },
+    {
+      "epoch": 0.9841880598809596,
+      "grad_norm": 2.1025618045242886,
+      "learning_rate": 2.6027453687487154e-09,
+      "loss": 0.6116,
+      "num_input_tokens_seen": 174411885,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9843083027715986,
+      "grad_norm": 3.023044858553157,
+      "learning_rate": 2.5631689219507422e-09,
+      "loss": 0.5248,
+      "num_input_tokens_seen": 174430285,
+      "step": 8186
+    },
+    {
+      "epoch": 0.9844285456622377,
+      "grad_norm": 1.8129161082976373,
+      "learning_rate": 2.523895484732197e-09,
+      "loss": 0.8294,
+      "num_input_tokens_seen": 174449460,
+      "step": 8187
+    },
+    {
+      "epoch": 0.9845487885528769,
+      "grad_norm": 2.169234175647031,
+      "learning_rate": 2.4849250630505357e-09,
+      "loss": 0.7429,
+      "num_input_tokens_seen": 174467425,
+      "step": 8188
+    },
+    {
+      "epoch": 0.9846690314435159,
+      "grad_norm": 1.959938241126848,
+      "learning_rate": 2.4462576628172528e-09,
+      "loss": 0.7305,
+      "num_input_tokens_seen": 174485775,
+      "step": 8189
+    },
+    {
+      "epoch": 0.984789274334155,
+      "grad_norm": 2.554926060490561,
+      "learning_rate": 2.407893289898544e-09,
+      "loss": 0.73,
+      "num_input_tokens_seen": 174504525,
+      "step": 8190
+    },
+    {
+      "epoch": 0.984909517224794,
+      "grad_norm": 2.1015829064329,
+      "learning_rate": 2.3698319501144202e-09,
+      "loss": 0.832,
+      "num_input_tokens_seen": 174525230,
+      "step": 8191
+    },
+    {
+      "epoch": 0.9850297601154332,
+      "grad_norm": 1.803213035332049,
+      "learning_rate": 2.3320736492382644e-09,
+      "loss": 0.7242,
+      "num_input_tokens_seen": 174543785,
+      "step": 8192
+    },
+    {
+      "epoch": 0.9851500030060723,
+      "grad_norm": 1.8538934348030842,
+      "learning_rate": 2.29461839299816e-09,
+      "loss": 0.6762,
+      "num_input_tokens_seen": 174563220,
+      "step": 8193
+    },
+    {
+      "epoch": 0.9852702458967113,
+      "grad_norm": 1.817544588049787,
+      "learning_rate": 2.257466187076229e-09,
+      "loss": 0.7956,
+      "num_input_tokens_seen": 174582145,
+      "step": 8194
+    },
+    {
+      "epoch": 0.9853904887873505,
+      "grad_norm": 2.2617389240460244,
+      "learning_rate": 2.2206170371081854e-09,
+      "loss": 0.702,
+      "num_input_tokens_seen": 174600450,
+      "step": 8195
+    },
+    {
+      "epoch": 0.9855107316779895,
+      "grad_norm": 1.6744999911213394,
+      "learning_rate": 2.1840709486842247e-09,
+      "loss": 0.8375,
+      "num_input_tokens_seen": 174619790,
+      "step": 8196
+    },
+    {
+      "epoch": 0.9856309745686286,
+      "grad_norm": 2.942772035092758,
+      "learning_rate": 2.1478279273481335e-09,
+      "loss": 0.7918,
+      "num_input_tokens_seen": 174637995,
+      "step": 8197
+    },
+    {
+      "epoch": 0.9857512174592677,
+      "grad_norm": 2.8771065626776515,
+      "learning_rate": 2.1118879785981815e-09,
+      "loss": 0.8024,
+      "num_input_tokens_seen": 174657855,
+      "step": 8198
+    },
+    {
+      "epoch": 0.9858714603499068,
+      "grad_norm": 1.7558341214960185,
+      "learning_rate": 2.0762511078862288e-09,
+      "loss": 0.7857,
+      "num_input_tokens_seen": 174677920,
+      "step": 8199
+    },
+    {
+      "epoch": 0.9859917032405459,
+      "grad_norm": 2.4751821131737866,
+      "learning_rate": 2.0409173206186183e-09,
+      "loss": 0.6524,
+      "num_input_tokens_seen": 174696880,
+      "step": 8200
+    },
+    {
+      "epoch": 0.986111946131185,
+      "grad_norm": 2.138024592105558,
+      "learning_rate": 2.0058866221550617e-09,
+      "loss": 0.8705,
+      "num_input_tokens_seen": 174714840,
+      "step": 8201
+    },
+    {
+      "epoch": 0.9862321890218241,
+      "grad_norm": 3.8660672270306793,
+      "learning_rate": 1.971159017809976e-09,
+      "loss": 0.7465,
+      "num_input_tokens_seen": 174732850,
+      "step": 8202
+    },
+    {
+      "epoch": 0.9863524319124631,
+      "grad_norm": 2.6995616277124594,
+      "learning_rate": 1.93673451285159e-09,
+      "loss": 0.7798,
+      "num_input_tokens_seen": 174751620,
+      "step": 8203
+    },
+    {
+      "epoch": 0.9864726748031023,
+      "grad_norm": 0.7724341943356106,
+      "learning_rate": 1.9026131125019495e-09,
+      "loss": 0.5964,
+      "num_input_tokens_seen": 174808710,
+      "step": 8204
+    },
+    {
+      "epoch": 0.9865929176937414,
+      "grad_norm": 1.8469702633006309,
+      "learning_rate": 1.8687948219371363e-09,
+      "loss": 0.8516,
+      "num_input_tokens_seen": 174827655,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9867131605843804,
+      "grad_norm": 3.293176856132798,
+      "learning_rate": 1.835279646287491e-09,
+      "loss": 0.8766,
+      "num_input_tokens_seen": 174845385,
+      "step": 8206
+    },
+    {
+      "epoch": 0.9868334034750196,
+      "grad_norm": 3.5308385935089834,
+      "learning_rate": 1.8020675906371685e-09,
+      "loss": 0.766,
+      "num_input_tokens_seen": 174864500,
+      "step": 8207
+    },
+    {
+      "epoch": 0.9869536463656586,
+      "grad_norm": 3.0348022487946267,
+      "learning_rate": 1.7691586600243612e-09,
+      "loss": 0.7438,
+      "num_input_tokens_seen": 174883120,
+      "step": 8208
+    },
+    {
+      "epoch": 0.9870738892562977,
+      "grad_norm": 3.361978996228318,
+      "learning_rate": 1.7365528594415202e-09,
+      "loss": 0.8547,
+      "num_input_tokens_seen": 174896910,
+      "step": 8209
+    },
+    {
+      "epoch": 0.9871941321469369,
+      "grad_norm": 2.0596399155289107,
+      "learning_rate": 1.7042501938346888e-09,
+      "loss": 0.669,
+      "num_input_tokens_seen": 174919360,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9873143750375759,
+      "grad_norm": 2.3605871324865277,
+      "learning_rate": 1.6722506681043913e-09,
+      "loss": 0.7636,
+      "num_input_tokens_seen": 174938040,
+      "step": 8211
+    },
+    {
+      "epoch": 0.987434617928215,
+      "grad_norm": 3.4691202377151793,
+      "learning_rate": 1.640554287104745e-09,
+      "loss": 0.6839,
+      "num_input_tokens_seen": 174956035,
+      "step": 8212
+    },
+    {
+      "epoch": 0.9875548608188541,
+      "grad_norm": 2.6814377921700676,
+      "learning_rate": 1.609161055644348e-09,
+      "loss": 0.7933,
+      "num_input_tokens_seen": 174971680,
+      "step": 8213
+    },
+    {
+      "epoch": 0.9876751037094932,
+      "grad_norm": 2.0702336996413226,
+      "learning_rate": 1.5780709784849467e-09,
+      "loss": 0.6672,
+      "num_input_tokens_seen": 174988420,
+      "step": 8214
+    },
+    {
+      "epoch": 0.9877953466001322,
+      "grad_norm": 2.091158107277124,
+      "learning_rate": 1.5472840603436565e-09,
+      "loss": 0.812,
+      "num_input_tokens_seen": 175005370,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9879155894907714,
+      "grad_norm": 4.941459196577964,
+      "learning_rate": 1.5168003058900757e-09,
+      "loss": 0.7906,
+      "num_input_tokens_seen": 175023090,
+      "step": 8216
+    },
+    {
+      "epoch": 0.9880358323814105,
+      "grad_norm": 2.0764351282813474,
+      "learning_rate": 1.4866197197491715e-09,
+      "loss": 0.9155,
+      "num_input_tokens_seen": 175042170,
+      "step": 8217
+    },
+    {
+      "epoch": 0.9881560752720495,
+      "grad_norm": 4.152594873888854,
+      "learning_rate": 1.4567423064988371e-09,
+      "loss": 0.7841,
+      "num_input_tokens_seen": 175059240,
+      "step": 8218
+    },
+    {
+      "epoch": 0.9882763181626887,
+      "grad_norm": 2.4157356594834867,
+      "learning_rate": 1.4271680706718913e-09,
+      "loss": 0.7689,
+      "num_input_tokens_seen": 175076635,
+      "step": 8219
+    },
+    {
+      "epoch": 0.9883965610533277,
+      "grad_norm": 2.2641177734521936,
+      "learning_rate": 1.3978970167543013e-09,
+      "loss": 0.8187,
+      "num_input_tokens_seen": 175096535,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9885168039439668,
+      "grad_norm": 3.0761114086101187,
+      "learning_rate": 1.3689291491867372e-09,
+      "loss": 0.7672,
+      "num_input_tokens_seen": 175114570,
+      "step": 8221
+    },
+    {
+      "epoch": 0.988637046834606,
+      "grad_norm": 2.188723533124547,
+      "learning_rate": 1.3402644723636836e-09,
+      "loss": 0.7336,
+      "num_input_tokens_seen": 175136320,
+      "step": 8222
+    },
+    {
+      "epoch": 0.988757289725245,
+      "grad_norm": 2.349230568079875,
+      "learning_rate": 1.311902990633218e-09,
+      "loss": 0.8308,
+      "num_input_tokens_seen": 175155005,
+      "step": 8223
+    },
+    {
+      "epoch": 0.9888775326158841,
+      "grad_norm": 1.769861787822553,
+      "learning_rate": 1.2838447082978987e-09,
+      "loss": 0.7063,
+      "num_input_tokens_seen": 175175880,
+      "step": 8224
+    },
+    {
+      "epoch": 0.9889977755065231,
+      "grad_norm": 3.293334393032793,
+      "learning_rate": 1.2560896296143208e-09,
+      "loss": 0.8243,
+      "num_input_tokens_seen": 175194065,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9891180183971623,
+      "grad_norm": 2.7623080149193124,
+      "learning_rate": 1.2286377587926722e-09,
+      "loss": 0.8184,
+      "num_input_tokens_seen": 175210575,
+      "step": 8226
+    },
+    {
+      "epoch": 0.9892382612878013,
+      "grad_norm": 2.2055527264777157,
+      "learning_rate": 1.2014890999973992e-09,
+      "loss": 0.7467,
+      "num_input_tokens_seen": 175227215,
+      "step": 8227
+    },
+    {
+      "epoch": 0.9893585041784404,
+      "grad_norm": 1.7809228354404718,
+      "learning_rate": 1.1746436573472073e-09,
+      "loss": 0.7756,
+      "num_input_tokens_seen": 175248670,
+      "step": 8228
+    },
+    {
+      "epoch": 0.9894787470690796,
+      "grad_norm": 2.5171829428685375,
+      "learning_rate": 1.1481014349141726e-09,
+      "loss": 0.6869,
+      "num_input_tokens_seen": 175265610,
+      "step": 8229
+    },
+    {
+      "epoch": 0.9895989899597186,
+      "grad_norm": 2.383116077565131,
+      "learning_rate": 1.121862436724852e-09,
+      "loss": 0.8338,
+      "num_input_tokens_seen": 175284170,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9897192328503577,
+      "grad_norm": 1.9637235341596146,
+      "learning_rate": 1.0959266667598388e-09,
+      "loss": 0.7032,
+      "num_input_tokens_seen": 175302705,
+      "step": 8231
+    },
+    {
+      "epoch": 0.9898394757409968,
+      "grad_norm": 2.7769046920525757,
+      "learning_rate": 1.0702941289533196e-09,
+      "loss": 0.7402,
+      "num_input_tokens_seen": 175321100,
+      "step": 8232
+    },
+    {
+      "epoch": 0.9899597186316359,
+      "grad_norm": 2.12520476244948,
+      "learning_rate": 1.0449648271939615e-09,
+      "loss": 0.8761,
+      "num_input_tokens_seen": 175337165,
+      "step": 8233
+    },
+    {
+      "epoch": 0.990079961522275,
+      "grad_norm": 1.627505953714871,
+      "learning_rate": 1.0199387653240243e-09,
+      "loss": 0.7257,
+      "num_input_tokens_seen": 175356575,
+      "step": 8234
+    },
+    {
+      "epoch": 0.9902002044129141,
+      "grad_norm": 1.7087624079813422,
+      "learning_rate": 9.952159471400267e-10,
+      "loss": 0.6997,
+      "num_input_tokens_seen": 175373335,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9903204473035532,
+      "grad_norm": 1.9484826577151628,
+      "learning_rate": 9.707963763925241e-10,
+      "loss": 0.8267,
+      "num_input_tokens_seen": 175392105,
+      "step": 8236
+    },
+    {
+      "epoch": 0.9904406901941922,
+      "grad_norm": 2.0765197558588726,
+      "learning_rate": 9.466800567856648e-10,
+      "loss": 0.7802,
+      "num_input_tokens_seen": 175410425,
+      "step": 8237
+    },
+    {
+      "epoch": 0.9905609330848314,
+      "grad_norm": 2.3984590820603398,
+      "learning_rate": 9.228669919778553e-10,
+      "loss": 0.683,
+      "num_input_tokens_seen": 175429070,
+      "step": 8238
+    },
+    {
+      "epoch": 0.9906811759754705,
+      "grad_norm": 2.1221046028586144,
+      "learning_rate": 8.993571855817617e-10,
+      "loss": 0.7896,
+      "num_input_tokens_seen": 175447620,
+      "step": 8239
+    },
+    {
+      "epoch": 0.9908014188661095,
+      "grad_norm": 2.0415281793379165,
+      "learning_rate": 8.761506411638642e-10,
+      "loss": 0.7344,
+      "num_input_tokens_seen": 175466805,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9909216617567487,
+      "grad_norm": 2.2644088759270757,
+      "learning_rate": 8.53247362244236e-10,
+      "loss": 0.7372,
+      "num_input_tokens_seen": 175485335,
+      "step": 8241
+    },
+    {
+      "epoch": 0.9910419046473877,
+      "grad_norm": 4.391329814701583,
+      "learning_rate": 8.306473522976532e-10,
+      "loss": 0.6787,
+      "num_input_tokens_seen": 175504460,
+      "step": 8242
+    },
+    {
+      "epoch": 0.9911621475380268,
+      "grad_norm": 2.2949354290347594,
+      "learning_rate": 8.083506147522623e-10,
+      "loss": 0.7135,
+      "num_input_tokens_seen": 175523575,
+      "step": 8243
+    },
+    {
+      "epoch": 0.991282390428666,
+      "grad_norm": 2.321622358009521,
+      "learning_rate": 7.863571529906909e-10,
+      "loss": 0.8519,
+      "num_input_tokens_seen": 175538880,
+      "step": 8244
+    },
+    {
+      "epoch": 0.991402633319305,
+      "grad_norm": 0.8000434019367024,
+      "learning_rate": 7.646669703489372e-10,
+      "loss": 0.6527,
+      "num_input_tokens_seen": 175602910,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9915228762099441,
+      "grad_norm": 2.4792683851780657,
+      "learning_rate": 7.432800701177023e-10,
+      "loss": 0.578,
+      "num_input_tokens_seen": 175620630,
+      "step": 8246
+    },
+    {
+      "epoch": 0.9916431191005832,
+      "grad_norm": 0.8829869352441092,
+      "learning_rate": 7.221964555415017e-10,
+      "loss": 0.6037,
+      "num_input_tokens_seen": 175680010,
+      "step": 8247
+    },
+    {
+      "epoch": 0.9917633619912223,
+      "grad_norm": 3.7492249374019573,
+      "learning_rate": 7.01416129818222e-10,
+      "loss": 0.7428,
+      "num_input_tokens_seen": 175697350,
+      "step": 8248
+    },
+    {
+      "epoch": 0.9918836048818613,
+      "grad_norm": 2.415236024482971,
+      "learning_rate": 6.809390961006745e-10,
+      "loss": 0.5816,
+      "num_input_tokens_seen": 175717200,
+      "step": 8249
+    },
+    {
+      "epoch": 0.9920038477725005,
+      "grad_norm": 2.5684401977483278,
+      "learning_rate": 6.607653574948191e-10,
+      "loss": 0.6831,
+      "num_input_tokens_seen": 175737700,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9921240906631396,
+      "grad_norm": 2.0867190167079914,
+      "learning_rate": 6.408949170613187e-10,
+      "loss": 0.8155,
+      "num_input_tokens_seen": 175756685,
+      "step": 8251
+    },
+    {
+      "epoch": 0.9922443335537786,
+      "grad_norm": 1.810709188480728,
+      "learning_rate": 6.213277778144288e-10,
+      "loss": 0.8132,
+      "num_input_tokens_seen": 175778050,
+      "step": 8252
+    },
+    {
+      "epoch": 0.9923645764444178,
+      "grad_norm": 2.040245725371067,
+      "learning_rate": 6.020639427224416e-10,
+      "loss": 0.6673,
+      "num_input_tokens_seen": 175795415,
+      "step": 8253
+    },
+    {
+      "epoch": 0.9924848193350568,
+      "grad_norm": 2.8215961746092946,
+      "learning_rate": 5.831034147076864e-10,
+      "loss": 0.7175,
+      "num_input_tokens_seen": 175812385,
+      "step": 8254
+    },
+    {
+      "epoch": 0.9926050622256959,
+      "grad_norm": 0.7216869926770196,
+      "learning_rate": 5.644461966463065e-10,
+      "loss": 0.575,
+      "num_input_tokens_seen": 175879715,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9927253051163349,
+      "grad_norm": 1.9879932919293617,
+      "learning_rate": 5.460922913687049e-10,
+      "loss": 0.7509,
+      "num_input_tokens_seen": 175898525,
+      "step": 8256
+    },
+    {
+      "epoch": 0.9928455480069741,
+      "grad_norm": 2.3696605000102378,
+      "learning_rate": 5.280417016593208e-10,
+      "loss": 0.747,
+      "num_input_tokens_seen": 175918035,
+      "step": 8257
+    },
+    {
+      "epoch": 0.9929657908976132,
+      "grad_norm": 2.006487433174183,
+      "learning_rate": 5.102944302559642e-10,
+      "loss": 0.7475,
+      "num_input_tokens_seen": 175935250,
+      "step": 8258
+    },
+    {
+      "epoch": 0.9930860337882522,
+      "grad_norm": 2.091255332901187,
+      "learning_rate": 4.9285047985137e-10,
+      "loss": 0.7821,
+      "num_input_tokens_seen": 175954390,
+      "step": 8259
+    },
+    {
+      "epoch": 0.9932062766788914,
+      "grad_norm": 2.091468950178844,
+      "learning_rate": 4.757098530916436e-10,
+      "loss": 0.7424,
+      "num_input_tokens_seen": 175974555,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9933265195695304,
+      "grad_norm": 3.5271336042439265,
+      "learning_rate": 4.5887255257670563e-10,
+      "loss": 0.7754,
+      "num_input_tokens_seen": 175991315,
+      "step": 8261
+    },
+    {
+      "epoch": 0.9934467624601695,
+      "grad_norm": 2.5085280957515375,
+      "learning_rate": 4.4233858086117906e-10,
+      "loss": 0.7603,
+      "num_input_tokens_seen": 176009560,
+      "step": 8262
+    },
+    {
+      "epoch": 0.9935670053508087,
+      "grad_norm": 2.7292772982230034,
+      "learning_rate": 4.261079404528356e-10,
+      "loss": 0.6809,
+      "num_input_tokens_seen": 176028760,
+      "step": 8263
+    },
+    {
+      "epoch": 0.9936872482414477,
+      "grad_norm": 2.4280748400474295,
+      "learning_rate": 4.1018063381437205e-10,
+      "loss": 0.6805,
+      "num_input_tokens_seen": 176048865,
+      "step": 8264
+    },
+    {
+      "epoch": 0.9938074911320868,
+      "grad_norm": 1.0118182579815589,
+      "learning_rate": 3.9455666336141167e-10,
+      "loss": 0.6633,
+      "num_input_tokens_seen": 176112365,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9939277340227259,
+      "grad_norm": 3.054157089563749,
+      "learning_rate": 3.7923603146450267e-10,
+      "loss": 0.8114,
+      "num_input_tokens_seen": 176128145,
+      "step": 8266
+    },
+    {
+      "epoch": 0.994047976913365,
+      "grad_norm": 2.0958864446894894,
+      "learning_rate": 3.642187404473418e-10,
+      "loss": 0.8026,
+      "num_input_tokens_seen": 176146025,
+      "step": 8267
+    },
+    {
+      "epoch": 0.994168219804004,
+      "grad_norm": 2.2062681475225663,
+      "learning_rate": 3.495047925885508e-10,
+      "loss": 0.8478,
+      "num_input_tokens_seen": 176164080,
+      "step": 8268
+    },
+    {
+      "epoch": 0.9942884626946432,
+      "grad_norm": 2.4037161592757816,
+      "learning_rate": 3.350941901199e-10,
+      "loss": 0.832,
+      "num_input_tokens_seen": 176180720,
+      "step": 8269
+    },
+    {
+      "epoch": 0.9944087055852823,
+      "grad_norm": 2.8931073750826344,
+      "learning_rate": 3.2098693522764066e-10,
+      "loss": 0.8236,
+      "num_input_tokens_seen": 176193640,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9945289484759213,
+      "grad_norm": 2.653038071448288,
+      "learning_rate": 3.071830300516165e-10,
+      "loss": 0.8127,
+      "num_input_tokens_seen": 176211190,
+      "step": 8271
+    },
+    {
+      "epoch": 0.9946491913665605,
+      "grad_norm": 2.675827430428048,
+      "learning_rate": 2.9368247668615234e-10,
+      "loss": 0.7031,
+      "num_input_tokens_seen": 176229500,
+      "step": 8272
+    },
+    {
+      "epoch": 0.9947694342571995,
+      "grad_norm": 4.744550491672129,
+      "learning_rate": 2.804852771789434e-10,
+      "loss": 0.6143,
+      "num_input_tokens_seen": 176242520,
+      "step": 8273
+    },
+    {
+      "epoch": 0.9948896771478386,
+      "grad_norm": 2.1101443353275293,
+      "learning_rate": 2.675914335321661e-10,
+      "loss": 0.5556,
+      "num_input_tokens_seen": 176260995,
+      "step": 8274
+    },
+    {
+      "epoch": 0.9950099200384778,
+      "grad_norm": 3.3603475336718596,
+      "learning_rate": 2.550009477018111e-10,
+      "loss": 0.7888,
+      "num_input_tokens_seen": 176279485,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9951301629291168,
+      "grad_norm": 2.76719365977582,
+      "learning_rate": 2.4271382159790634e-10,
+      "loss": 0.6217,
+      "num_input_tokens_seen": 176296635,
+      "step": 8276
+    },
+    {
+      "epoch": 0.9952504058197559,
+      "grad_norm": 1.740750725358956,
+      "learning_rate": 2.3073005708429406e-10,
+      "loss": 0.855,
+      "num_input_tokens_seen": 176316000,
+      "step": 8277
+    },
+    {
+      "epoch": 0.995370648710395,
+      "grad_norm": 1.908443667053566,
+      "learning_rate": 2.190496559788535e-10,
+      "loss": 0.7113,
+      "num_input_tokens_seen": 176334005,
+      "step": 8278
+    },
+    {
+      "epoch": 0.9954908916010341,
+      "grad_norm": 3.2127900523222133,
+      "learning_rate": 2.0767262005372265e-10,
+      "loss": 0.7617,
+      "num_input_tokens_seen": 176351240,
+      "step": 8279
+    },
+    {
+      "epoch": 0.9956111344916732,
+      "grad_norm": 1.8994265271107622,
+      "learning_rate": 1.965989510346322e-10,
+      "loss": 0.7444,
+      "num_input_tokens_seen": 176370080,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9957313773823123,
+      "grad_norm": 4.47759596491919,
+      "learning_rate": 1.8582865060134955e-10,
+      "loss": 0.6966,
+      "num_input_tokens_seen": 176387990,
+      "step": 8281
+    },
+    {
+      "epoch": 0.9958516202729514,
+      "grad_norm": 0.8166389173617845,
+      "learning_rate": 1.7536172038790098e-10,
+      "loss": 0.5902,
+      "num_input_tokens_seen": 176448020,
+      "step": 8282
+    },
+    {
+      "epoch": 0.9959718631635904,
+      "grad_norm": 2.2943255952352968,
+      "learning_rate": 1.651981619819054e-10,
+      "loss": 0.6954,
+      "num_input_tokens_seen": 176464890,
+      "step": 8283
+    },
+    {
+      "epoch": 0.9960921060542296,
+      "grad_norm": 2.5726579921884793,
+      "learning_rate": 1.5533797692546257e-10,
+      "loss": 0.6965,
+      "num_input_tokens_seen": 176483345,
+      "step": 8284
+    },
+    {
+      "epoch": 0.9962123489448687,
+      "grad_norm": 2.15599553412446,
+      "learning_rate": 1.4578116671404296e-10,
+      "loss": 0.8288,
+      "num_input_tokens_seen": 176501345,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9963325918355077,
+      "grad_norm": 2.583962588903221,
+      "learning_rate": 1.3652773279759777e-10,
+      "loss": 0.7118,
+      "num_input_tokens_seen": 176517715,
+      "step": 8286
+    },
+    {
+      "epoch": 0.9964528347261468,
+      "grad_norm": 2.343758824603163,
+      "learning_rate": 1.2757767657989305e-10,
+      "loss": 0.6193,
+      "num_input_tokens_seen": 176541225,
+      "step": 8287
+    },
+    {
+      "epoch": 0.9965730776167859,
+      "grad_norm": 2.7563916119811016,
+      "learning_rate": 1.1893099941850948e-10,
+      "loss": 0.8599,
+      "num_input_tokens_seen": 176559840,
+      "step": 8288
+    },
+    {
+      "epoch": 0.996693320507425,
+      "grad_norm": 2.567794781252191,
+      "learning_rate": 1.105877026252866e-10,
+      "loss": 0.7654,
+      "num_input_tokens_seen": 176577890,
+      "step": 8289
+    },
+    {
+      "epoch": 0.996813563398064,
+      "grad_norm": 2.312729168030679,
+      "learning_rate": 1.0254778746565663e-10,
+      "loss": 0.7162,
+      "num_input_tokens_seen": 176592885,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9969338062887032,
+      "grad_norm": 3.1137882306079594,
+      "learning_rate": 9.481125515953259e-11,
+      "loss": 0.726,
+      "num_input_tokens_seen": 176610665,
+      "step": 8291
+    },
+    {
+      "epoch": 0.9970540491793423,
+      "grad_norm": 3.103135461880461,
+      "learning_rate": 8.737810688064228e-11,
+      "loss": 0.7931,
+      "num_input_tokens_seen": 176630220,
+      "step": 8292
+    },
+    {
+      "epoch": 0.9971742920699813,
+      "grad_norm": 4.054154418615418,
+      "learning_rate": 8.024834375608414e-11,
+      "loss": 0.7865,
+      "num_input_tokens_seen": 176648530,
+      "step": 8293
+    },
+    {
+      "epoch": 0.9972945349606205,
+      "grad_norm": 0.8539698343127142,
+      "learning_rate": 7.342196686788149e-11,
+      "loss": 0.6648,
+      "num_input_tokens_seen": 176701415,
+      "step": 8294
+    },
+    {
+      "epoch": 0.9974147778512595,
+      "grad_norm": 4.617238943334054,
+      "learning_rate": 6.689897725142834e-11,
+      "loss": 0.6789,
+      "num_input_tokens_seen": 176720610,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9975350207418986,
+      "grad_norm": 2.4320007278956117,
+      "learning_rate": 6.067937589615545e-11,
+      "loss": 0.8744,
+      "num_input_tokens_seen": 176738405,
+      "step": 8296
+    },
+    {
+      "epoch": 0.9976552636325378,
+      "grad_norm": 0.8037430908240526,
+      "learning_rate": 5.476316374575241e-11,
+      "loss": 0.5846,
+      "num_input_tokens_seen": 176801610,
+      "step": 8297
+    },
+    {
+      "epoch": 0.9977755065231768,
+      "grad_norm": 3.0512985585844783,
+      "learning_rate": 4.9150341697723476e-11,
+      "loss": 0.7263,
+      "num_input_tokens_seen": 176821220,
+      "step": 8298
+    },
+    {
+      "epoch": 0.9978957494138159,
+      "grad_norm": 1.6951311378268001,
+      "learning_rate": 4.384091060338768e-11,
+      "loss": 0.6563,
+      "num_input_tokens_seen": 176841410,
+      "step": 8299
+    },
+    {
+      "epoch": 0.998015992304455,
+      "grad_norm": 2.514276611531191,
+      "learning_rate": 3.883487126810081e-11,
+      "loss": 0.7359,
+      "num_input_tokens_seen": 176860390,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9981362351950941,
+      "grad_norm": 2.1396989651389244,
+      "learning_rate": 3.41322244516995e-11,
+      "loss": 0.7934,
+      "num_input_tokens_seen": 176878055,
+      "step": 8301
+    },
+    {
+      "epoch": 0.9982564780857331,
+      "grad_norm": 1.7317634566852855,
+      "learning_rate": 2.9732970866946925e-11,
+      "loss": 0.6223,
+      "num_input_tokens_seen": 176897655,
+      "step": 8302
+    },
+    {
+      "epoch": 0.9983767209763723,
+      "grad_norm": 3.3719932002171005,
+      "learning_rate": 2.563711118175327e-11,
+      "loss": 0.7808,
+      "num_input_tokens_seen": 176914260,
+      "step": 8303
+    },
+    {
+      "epoch": 0.9984969638670114,
+      "grad_norm": 1.8855250152951066,
+      "learning_rate": 2.184464601717728e-11,
+      "loss": 0.8316,
+      "num_input_tokens_seen": 176932295,
+      "step": 8304
+    },
+    {
+      "epoch": 0.9986172067576504,
+      "grad_norm": 2.6088639848298802,
+      "learning_rate": 1.8355575948758585e-11,
+      "loss": 0.771,
+      "num_input_tokens_seen": 176950000,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9987374496482896,
+      "grad_norm": 2.4449209720783576,
+      "learning_rate": 1.5169901505407424e-11,
+      "loss": 0.7396,
+      "num_input_tokens_seen": 176966785,
+      "step": 8306
+    },
+    {
+      "epoch": 0.9988576925389286,
+      "grad_norm": 1.8933684917611164,
+      "learning_rate": 1.228762317073695e-11,
+      "loss": 0.7304,
+      "num_input_tokens_seen": 176985335,
+      "step": 8307
+    },
+    {
+      "epoch": 0.9989779354295677,
+      "grad_norm": 2.2195442228715208,
+      "learning_rate": 9.70874138195299e-12,
+      "loss": 0.781,
+      "num_input_tokens_seen": 177006965,
+      "step": 8308
+    },
+    {
+      "epoch": 0.9990981783202069,
+      "grad_norm": 2.4100321606138677,
+      "learning_rate": 7.433256530076093e-12,
+      "loss": 0.7375,
+      "num_input_tokens_seen": 177026640,
+      "step": 8309
+    },
+    {
+      "epoch": 0.9992184212108459,
+      "grad_norm": 2.3830769448293965,
+      "learning_rate": 5.46116896038562e-12,
+      "loss": 0.742,
+      "num_input_tokens_seen": 177040770,
+      "step": 8310
+    },
+    {
+      "epoch": 0.999338664101485,
+      "grad_norm": 2.2288285164145263,
+      "learning_rate": 3.792478972197699e-12,
+      "loss": 0.6167,
+      "num_input_tokens_seen": 177061075,
+      "step": 8311
+    },
+    {
+      "epoch": 0.9994589069921241,
+      "grad_norm": 3.4773882566335352,
+      "learning_rate": 2.4271868181990895e-12,
+      "loss": 0.6946,
+      "num_input_tokens_seen": 177077960,
+      "step": 8312
+    },
+    {
+      "epoch": 0.9995791498827632,
+      "grad_norm": 2.313151740493287,
+      "learning_rate": 1.3652927060014973e-12,
+      "loss": 0.796,
+      "num_input_tokens_seen": 177093275,
+      "step": 8313
+    },
+    {
+      "epoch": 0.9996993927734023,
+      "grad_norm": 3.5257222933850607,
+      "learning_rate": 6.067967965872612e-13,
+      "loss": 0.6378,
+      "num_input_tokens_seen": 177112605,
+      "step": 8314
+    },
+    {
+      "epoch": 0.9998196356640414,
+      "grad_norm": 1.6849569618192852,
+      "learning_rate": 1.5169920497548615e-13,
+      "loss": 0.7645,
+      "num_input_tokens_seen": 177136945,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9999398785546805,
+      "grad_norm": 1.1960316289490573,
+      "learning_rate": 0.0,
+      "loss": 0.5808,
+      "num_input_tokens_seen": 177185545,
+      "step": 8316
+    },
+    {
+      "epoch": 0.9999398785546805,
+      "num_input_tokens_seen": 177185545,
+      "step": 8316,
+      "total_flos": 6.901864404405453e+17,
+      "train_loss": 0.7678336634006335,
+      "train_runtime": 60017.2753,
+      "train_samples_per_second": 5.543,
+      "train_steps_per_second": 0.139
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8316,
+  "num_input_tokens_seen": 177185545,
+  "num_train_epochs": 1,
+  "save_steps": 832,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.901864404405453e+17,
+  "train_batch_size": 5,
+  "trial_name": null,
+  "trial_params": null
+}