diff --git "a/checkpoint-3511/trainer_state.json" "b/checkpoint-3511/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3511/trainer_state.json"
@@ -0,0 +1,21245 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997864312664626,
+  "eval_steps": 176,
+  "global_step": 3511,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 3.5718,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 5.830782413482666,
+      "eval_runtime": 2940.9278,
+      "eval_samples_per_second": 6.966,
+      "eval_steps_per_second": 2.322,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 6.0175,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 5.8622,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 5.8119,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 4.4185,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 4.4103,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 4.1269,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 3.848,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 3.7346,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8e-05,
+      "loss": 3.5968,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 3.5486,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 3.492,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4e-05,
+      "loss": 3.4527,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 3.4616,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 3.4919,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 3.3711,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 3.3979,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 3.2596,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6e-05,
+      "loss": 3.3727,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8e-05,
+      "loss": 3.2903,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 3.3427,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2e-05,
+      "loss": 3.3386,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 3.2772,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 3.2423,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8e-05,
+      "loss": 3.2333,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 3.2475,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 3.3114,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 3.2723,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 3.1906,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.8e-05,
+      "loss": 3.2488,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 3.2203,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.2e-05,
+      "loss": 3.2247,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 3.1383,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.6e-05,
+      "loss": 3.1227,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 3.2286,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7e-05,
+      "loss": 3.2415,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2e-05,
+      "loss": 3.0958,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.4e-05,
+      "loss": 3.1529,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.6e-05,
+      "loss": 3.1929,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 3.1599,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8e-05,
+      "loss": 3.1247,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.2e-05,
+      "loss": 3.1767,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.4e-05,
+      "loss": 3.213,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.6e-05,
+      "loss": 3.2521,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 3.1038,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9e-05,
+      "loss": 3.2639,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 3.1598,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.4e-05,
+      "loss": 3.202,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.6e-05,
+      "loss": 3.2283,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.8e-05,
+      "loss": 3.1866,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 3.1442,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 3.0943,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 3.1444,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 3.1593,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 3.1638,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 3.2374,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 3.2035,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 3.2014,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000116,
+      "loss": 3.2609,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000118,
+      "loss": 3.1562,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00012,
+      "loss": 3.2534,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000122,
+      "loss": 3.1461,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000124,
+      "loss": 3.2507,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000126,
+      "loss": 3.2205,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 3.2393,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 3.1658,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000132,
+      "loss": 3.1736,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000134,
+      "loss": 3.2191,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 3.2439,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000138,
+      "loss": 3.2145,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00014,
+      "loss": 3.317,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000142,
+      "loss": 3.1762,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000144,
+      "loss": 3.2242,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000146,
+      "loss": 3.2284,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000148,
+      "loss": 3.2573,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 3.1517,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000152,
+      "loss": 3.2148,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000154,
+      "loss": 3.3142,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 3.2461,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 3.2156,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00016,
+      "loss": 3.2101,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000162,
+      "loss": 3.2675,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000164,
+      "loss": 3.3065,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000166,
+      "loss": 3.3035,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000168,
+      "loss": 3.2975,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00017,
+      "loss": 3.3041,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000172,
+      "loss": 3.2666,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000174,
+      "loss": 3.2871,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 3.264,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 3.2833,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00018,
+      "loss": 3.3067,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000182,
+      "loss": 3.2589,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 3.3806,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 3.219,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000188,
+      "loss": 3.3217,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019,
+      "loss": 3.2479,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000192,
+      "loss": 3.288,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000194,
+      "loss": 3.2907,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000196,
+      "loss": 3.1544,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 3.3607,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 3.2381,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001999413661682791,
+      "loss": 3.3826,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001998827323365582,
+      "loss": 3.3419,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001998240985048373,
+      "loss": 3.2726,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001997654646731164,
+      "loss": 3.2404,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001997068308413955,
+      "loss": 3.267,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019964819700967458,
+      "loss": 3.3027,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001995895631779537,
+      "loss": 3.3157,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001995309293462328,
+      "loss": 3.2919,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019947229551451189,
+      "loss": 3.3454,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019941366168279098,
+      "loss": 3.3259,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019935502785107007,
+      "loss": 3.3948,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019929639401934917,
+      "loss": 3.2172,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019923776018762826,
+      "loss": 3.3734,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019917912635590736,
+      "loss": 3.3004,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019912049252418648,
+      "loss": 3.326,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019906185869246557,
+      "loss": 3.3342,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019900322486074464,
+      "loss": 3.3513,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019894459102902376,
+      "loss": 3.302,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019888595719730286,
+      "loss": 3.2558,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019882732336558195,
+      "loss": 3.2531,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019876868953386104,
+      "loss": 3.2707,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019871005570214014,
+      "loss": 3.3266,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019865142187041926,
+      "loss": 3.3349,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019859278803869833,
+      "loss": 3.3154,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019853415420697742,
+      "loss": 3.2647,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019847552037525654,
+      "loss": 3.3634,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019841688654353564,
+      "loss": 3.3126,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001983582527118147,
+      "loss": 3.3529,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019829961888009383,
+      "loss": 3.3262,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019824098504837292,
+      "loss": 3.3167,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019818235121665201,
+      "loss": 3.4109,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001981237173849311,
+      "loss": 3.1823,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001980650835532102,
+      "loss": 3.3014,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019800644972148932,
+      "loss": 3.3807,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019794781588976842,
+      "loss": 3.3407,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019788918205804749,
+      "loss": 3.4206,
+      "step": 137
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001978305482263266,
+      "loss": 3.4182,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001977719143946057,
+      "loss": 3.3931,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001977132805628848,
+      "loss": 3.3758,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001976546467311639,
+      "loss": 3.3941,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019759601289944299,
+      "loss": 3.3912,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001975373790677221,
+      "loss": 3.2704,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019747874523600117,
+      "loss": 3.4061,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019742011140428027,
+      "loss": 3.4493,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001973614775725594,
+      "loss": 3.346,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019730284374083848,
+      "loss": 3.3776,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019724420990911755,
+      "loss": 3.3023,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019718557607739667,
+      "loss": 3.3579,
+      "step": 149
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019712694224567577,
+      "loss": 3.2763,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019706830841395486,
+      "loss": 3.3203,
+      "step": 151
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019700967458223396,
+      "loss": 3.3814,
+      "step": 152
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019695104075051305,
+      "loss": 3.3263,
+      "step": 153
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019689240691879217,
+      "loss": 3.2977,
+      "step": 154
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019683377308707124,
+      "loss": 3.3911,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019677513925535033,
+      "loss": 3.3715,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019671650542362945,
+      "loss": 3.3647,
+      "step": 157
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019665787159190855,
+      "loss": 3.3792,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019659923776018762,
+      "loss": 3.4429,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019654060392846674,
+      "loss": 3.4239,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019648197009674583,
+      "loss": 3.4442,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019642333626502493,
+      "loss": 3.4245,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019636470243330402,
+      "loss": 3.2396,
+      "step": 163
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019630606860158311,
+      "loss": 3.375,
+      "step": 164
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019624743476986224,
+      "loss": 3.356,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001961888009381413,
+      "loss": 3.2876,
+      "step": 166
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001961301671064204,
+      "loss": 3.3384,
+      "step": 167
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019607153327469952,
+      "loss": 3.4072,
+      "step": 168
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001960128994429786,
+      "loss": 3.4314,
+      "step": 169
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001959542656112577,
+      "loss": 3.3795,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001958956317795368,
+      "loss": 3.2868,
+      "step": 171
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001958369979478159,
+      "loss": 3.3781,
+      "step": 172
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000195778364116095,
+      "loss": 3.4215,
+      "step": 173
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019571973028437408,
+      "loss": 3.4015,
+      "step": 174
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019566109645265318,
+      "loss": 3.3857,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001956024626209323,
+      "loss": 3.3115,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 3.4127612113952637,
+      "eval_runtime": 2941.7575,
+      "eval_samples_per_second": 6.964,
+      "eval_steps_per_second": 2.321,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019554382878921137,
+      "loss": 3.3804,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001954851949574905,
+      "loss": 3.3767,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019542656112576958,
+      "loss": 3.1557,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019536792729404868,
+      "loss": 3.4065,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019530929346232777,
+      "loss": 3.4646,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019525065963060687,
+      "loss": 3.2766,
+      "step": 182
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019519202579888596,
+      "loss": 3.3506,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019513339196716508,
+      "loss": 3.3775,
+      "step": 184
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019507475813544415,
+      "loss": 3.3716,
+      "step": 185
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019501612430372324,
+      "loss": 3.4035,
+      "step": 186
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019495749047200236,
+      "loss": 3.3617,
+      "step": 187
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019489885664028146,
+      "loss": 3.3539,
+      "step": 188
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019484022280856055,
+      "loss": 3.3618,
+      "step": 189
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019478158897683965,
+      "loss": 3.275,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019472295514511874,
+      "loss": 3.3481,
+      "step": 191
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019466432131339784,
+      "loss": 3.4303,
+      "step": 192
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019460568748167693,
+      "loss": 3.4039,
+      "step": 193
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019454705364995603,
+      "loss": 3.4618,
+      "step": 194
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019448841981823515,
+      "loss": 3.2109,
+      "step": 195
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019442978598651421,
+      "loss": 3.3467,
+      "step": 196
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019437115215479334,
+      "loss": 3.3039,
+      "step": 197
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019431251832307243,
+      "loss": 3.4182,
+      "step": 198
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019425388449135152,
+      "loss": 3.3874,
+      "step": 199
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019419525065963062,
+      "loss": 3.3259,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001941366168279097,
+      "loss": 3.4389,
+      "step": 201
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001940779829961888,
+      "loss": 3.311,
+      "step": 202
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001940193491644679,
+      "loss": 3.2309,
+      "step": 203
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000193960715332747,
+      "loss": 3.445,
+      "step": 204
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019390208150102612,
+      "loss": 3.4121,
+      "step": 205
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001938434476693052,
+      "loss": 3.4134,
+      "step": 206
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019378481383758428,
+      "loss": 3.3747,
+      "step": 207
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001937261800058634,
+      "loss": 3.2936,
+      "step": 208
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001936675461741425,
+      "loss": 3.2643,
+      "step": 209
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001936089123424216,
+      "loss": 3.4549,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019355027851070068,
+      "loss": 3.3649,
+      "step": 211
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019349164467897978,
+      "loss": 3.3066,
+      "step": 212
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001934330108472589,
+      "loss": 3.2665,
+      "step": 213
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019337437701553797,
+      "loss": 3.427,
+      "step": 214
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019331574318381706,
+      "loss": 3.4163,
+      "step": 215
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019325710935209618,
+      "loss": 3.4074,
+      "step": 216
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019319847552037528,
+      "loss": 3.3811,
+      "step": 217
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019313984168865434,
+      "loss": 3.4369,
+      "step": 218
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019308120785693346,
+      "loss": 3.3511,
+      "step": 219
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019302257402521256,
+      "loss": 3.3779,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019296394019349165,
+      "loss": 3.272,
+      "step": 221
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019290530636177075,
+      "loss": 3.3569,
+      "step": 222
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019284667253004984,
+      "loss": 3.4014,
+      "step": 223
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019278803869832896,
+      "loss": 3.3813,
+      "step": 224
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019272940486660803,
+      "loss": 3.3642,
+      "step": 225
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019267077103488712,
+      "loss": 3.384,
+      "step": 226
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019261213720316625,
+      "loss": 3.3342,
+      "step": 227
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019255350337144534,
+      "loss": 3.3221,
+      "step": 228
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001924948695397244,
+      "loss": 3.3293,
+      "step": 229
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019243623570800353,
+      "loss": 3.4013,
+      "step": 230
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019237760187628262,
+      "loss": 3.3182,
+      "step": 231
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019231896804456174,
+      "loss": 3.3558,
+      "step": 232
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001922603342128408,
+      "loss": 3.2817,
+      "step": 233
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001922017003811199,
+      "loss": 3.4694,
+      "step": 234
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019214306654939903,
+      "loss": 3.4166,
+      "step": 235
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019208443271767812,
+      "loss": 3.2853,
+      "step": 236
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001920257988859572,
+      "loss": 3.334,
+      "step": 237
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001919671650542363,
+      "loss": 3.2872,
+      "step": 238
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001919085312225154,
+      "loss": 3.2861,
+      "step": 239
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001918498973907945,
+      "loss": 3.3128,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001917912635590736,
+      "loss": 3.3114,
+      "step": 241
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001917326297273527,
+      "loss": 3.3369,
+      "step": 242
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001916739958956318,
+      "loss": 3.3872,
+      "step": 243
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019161536206391088,
+      "loss": 3.347,
+      "step": 244
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019155672823218997,
+      "loss": 3.4326,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001914980944004691,
+      "loss": 3.3617,
+      "step": 246
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001914394605687482,
+      "loss": 3.3735,
+      "step": 247
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019138082673702725,
+      "loss": 3.3881,
+      "step": 248
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019132219290530638,
+      "loss": 3.3635,
+      "step": 249
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019126355907358547,
+      "loss": 3.4983,
+      "step": 250
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019120492524186456,
+      "loss": 3.2744,
+      "step": 251
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019114629141014366,
+      "loss": 3.3445,
+      "step": 252
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019108765757842275,
+      "loss": 3.3678,
+      "step": 253
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019102902374670187,
+      "loss": 3.2684,
+      "step": 254
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019097038991498094,
+      "loss": 3.2846,
+      "step": 255
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019091175608326004,
+      "loss": 3.3552,
+      "step": 256
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019085312225153916,
+      "loss": 3.2922,
+      "step": 257
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019079448841981825,
+      "loss": 3.3105,
+      "step": 258
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019073585458809735,
+      "loss": 3.3673,
+      "step": 259
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019067722075637644,
+      "loss": 3.3258,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019061858692465553,
+      "loss": 3.232,
+      "step": 261
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019055995309293463,
+      "loss": 3.3009,
+      "step": 262
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019050131926121372,
+      "loss": 3.3539,
+      "step": 263
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019044268542949282,
+      "loss": 3.4128,
+      "step": 264
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019038405159777194,
+      "loss": 3.3834,
+      "step": 265
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000190325417766051,
+      "loss": 3.418,
+      "step": 266
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019026678393433013,
+      "loss": 3.2632,
+      "step": 267
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019020815010260922,
+      "loss": 3.398,
+      "step": 268
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019014951627088832,
+      "loss": 3.2799,
+      "step": 269
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001900908824391674,
+      "loss": 3.4037,
+      "step": 270
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001900322486074465,
+      "loss": 3.3112,
+      "step": 271
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001899736147757256,
+      "loss": 3.3791,
+      "step": 272
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001899149809440047,
+      "loss": 3.4785,
+      "step": 273
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001898563471122838,
+      "loss": 3.2971,
+      "step": 274
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018979771328056288,
+      "loss": 3.345,
+      "step": 275
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000189739079448842,
+      "loss": 3.3416,
+      "step": 276
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018968044561712107,
+      "loss": 3.4207,
+      "step": 277
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001896218117854002,
+      "loss": 3.3759,
+      "step": 278
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018956317795367929,
+      "loss": 3.4658,
+      "step": 279
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018950454412195838,
+      "loss": 3.3258,
+      "step": 280
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018944591029023747,
+      "loss": 3.2984,
+      "step": 281
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018938727645851657,
+      "loss": 3.3746,
+      "step": 282
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018932864262679566,
+      "loss": 3.3248,
+      "step": 283
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018927000879507478,
+      "loss": 3.368,
+      "step": 284
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018921137496335385,
+      "loss": 3.3685,
+      "step": 285
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018915274113163297,
+      "loss": 3.2488,
+      "step": 286
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018909410729991207,
+      "loss": 3.4339,
+      "step": 287
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018903547346819116,
+      "loss": 3.2997,
+      "step": 288
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018897683963647026,
+      "loss": 3.3752,
+      "step": 289
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018891820580474935,
+      "loss": 3.3072,
+      "step": 290
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018885957197302845,
+      "loss": 3.3456,
+      "step": 291
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018880093814130754,
+      "loss": 3.3443,
+      "step": 292
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018874230430958663,
+      "loss": 3.3228,
+      "step": 293
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018868367047786576,
+      "loss": 3.4276,
+      "step": 294
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018862503664614485,
+      "loss": 3.313,
+      "step": 295
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018856640281442392,
+      "loss": 3.3224,
+      "step": 296
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018850776898270304,
+      "loss": 3.3374,
+      "step": 297
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018844913515098213,
+      "loss": 3.4112,
+      "step": 298
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018839050131926123,
+      "loss": 3.3292,
+      "step": 299
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018833186748754032,
+      "loss": 3.2262,
+      "step": 300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018827323365581942,
+      "loss": 3.3758,
+      "step": 301
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001882145998240985,
+      "loss": 3.3297,
+      "step": 302
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001881559659923776,
+      "loss": 3.2446,
+      "step": 303
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001880973321606567,
+      "loss": 3.4397,
+      "step": 304
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018803869832893582,
+      "loss": 3.3154,
+      "step": 305
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018798006449721491,
+      "loss": 3.3331,
+      "step": 306
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018792143066549398,
+      "loss": 3.2873,
+      "step": 307
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001878627968337731,
+      "loss": 3.3713,
+      "step": 308
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001878041630020522,
+      "loss": 3.3679,
+      "step": 309
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001877455291703313,
+      "loss": 3.3151,
+      "step": 310
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018768689533861039,
+      "loss": 3.3328,
+      "step": 311
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018762826150688948,
+      "loss": 3.4282,
+      "step": 312
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001875696276751686,
+      "loss": 3.2908,
+      "step": 313
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018751099384344767,
+      "loss": 3.3646,
+      "step": 314
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018745236001172676,
+      "loss": 3.3616,
+      "step": 315
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018739372618000588,
+      "loss": 3.3221,
+      "step": 316
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018733509234828498,
+      "loss": 3.1692,
+      "step": 317
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018727645851656405,
+      "loss": 3.327,
+      "step": 318
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018721782468484317,
+      "loss": 3.3521,
+      "step": 319
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018715919085312226,
+      "loss": 3.2864,
+      "step": 320
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018710055702140136,
+      "loss": 3.4104,
+      "step": 321
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018704192318968045,
+      "loss": 3.3519,
+      "step": 322
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018698328935795954,
+      "loss": 3.3847,
+      "step": 323
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018692465552623867,
+      "loss": 3.3199,
+      "step": 324
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018686602169451773,
+      "loss": 3.3427,
+      "step": 325
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018680738786279683,
+      "loss": 3.3561,
+      "step": 326
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018674875403107595,
+      "loss": 3.2695,
+      "step": 327
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018669012019935504,
+      "loss": 3.3613,
+      "step": 328
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001866314863676341,
+      "loss": 3.3666,
+      "step": 329
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018657285253591323,
+      "loss": 3.356,
+      "step": 330
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018651421870419233,
+      "loss": 3.312,
+      "step": 331
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018645558487247142,
+      "loss": 3.3382,
+      "step": 332
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00018639695104075051,
+      "loss": 3.3648,
+      "step": 333
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001863383172090296,
+      "loss": 3.3673,
+      "step": 334
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018627968337730873,
+      "loss": 3.2083,
+      "step": 335
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001862210495455878,
+      "loss": 3.3504,
+      "step": 336
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001861624157138669,
+      "loss": 3.3457,
+      "step": 337
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018610378188214601,
+      "loss": 3.3261,
+      "step": 338
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001860451480504251,
+      "loss": 3.2873,
+      "step": 339
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001859865142187042,
+      "loss": 3.3141,
+      "step": 340
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001859278803869833,
+      "loss": 3.3467,
+      "step": 341
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001858692465552624,
+      "loss": 3.3692,
+      "step": 342
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001858106127235415,
+      "loss": 3.3813,
+      "step": 343
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018575197889182058,
+      "loss": 3.3228,
+      "step": 344
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018569334506009967,
+      "loss": 3.341,
+      "step": 345
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001856347112283788,
+      "loss": 3.3532,
+      "step": 346
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001855760773966579,
+      "loss": 3.3119,
+      "step": 347
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018551744356493698,
+      "loss": 3.3122,
+      "step": 348
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018545880973321608,
+      "loss": 3.3532,
+      "step": 349
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018540017590149517,
+      "loss": 3.3514,
+      "step": 350
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018534154206977427,
+      "loss": 3.3955,
+      "step": 351
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018528290823805336,
+      "loss": 3.3363,
+      "step": 352
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 3.4325549602508545,
+      "eval_runtime": 2940.8692,
+      "eval_samples_per_second": 6.966,
+      "eval_steps_per_second": 2.322,
+      "step": 352
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018522427440633246,
+      "loss": 3.2021,
+      "step": 353
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018516564057461158,
+      "loss": 3.2987,
+      "step": 354
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018510700674289064,
+      "loss": 3.3542,
+      "step": 355
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018504837291116974,
+      "loss": 3.3654,
+      "step": 356
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018498973907944886,
+      "loss": 3.4304,
+      "step": 357
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018493110524772795,
+      "loss": 3.4141,
+      "step": 358
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018487247141600705,
+      "loss": 3.3684,
+      "step": 359
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018481383758428614,
+      "loss": 3.2512,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018475520375256524,
+      "loss": 3.2606,
+      "step": 361
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018469656992084433,
+      "loss": 3.2782,
+      "step": 362
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018463793608912343,
+      "loss": 3.3394,
+      "step": 363
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018457930225740252,
+      "loss": 3.4092,
+      "step": 364
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018452066842568164,
+      "loss": 3.2721,
+      "step": 365
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001844620345939607,
+      "loss": 3.3498,
+      "step": 366
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018440340076223983,
+      "loss": 3.304,
+      "step": 367
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00018434476693051892,
+      "loss": 3.2888,
+      "step": 368
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018428613309879802,
+      "loss": 3.3503,
+      "step": 369
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001842274992670771,
+      "loss": 3.3369,
+      "step": 370
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001841688654353562,
+      "loss": 3.2832,
+      "step": 371
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001841102316036353,
+      "loss": 3.343,
+      "step": 372
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001840515977719144,
+      "loss": 3.3148,
+      "step": 373
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001839929639401935,
+      "loss": 3.3384,
+      "step": 374
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001839343301084726,
+      "loss": 3.2707,
+      "step": 375
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001838756962767517,
+      "loss": 3.38,
+      "step": 376
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018381706244503077,
+      "loss": 3.3358,
+      "step": 377
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001837584286133099,
+      "loss": 3.3641,
+      "step": 378
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000183699794781589,
+      "loss": 3.3345,
+      "step": 379
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018364116094986808,
+      "loss": 3.3788,
+      "step": 380
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018358252711814718,
+      "loss": 3.3313,
+      "step": 381
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018352389328642627,
+      "loss": 3.3056,
+      "step": 382
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018346525945470537,
+      "loss": 3.3711,
+      "step": 383
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018340662562298446,
+      "loss": 3.3281,
+      "step": 384
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018334799179126355,
+      "loss": 3.2252,
+      "step": 385
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018328935795954268,
+      "loss": 3.3881,
+      "step": 386
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018323072412782177,
+      "loss": 3.3027,
+      "step": 387
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018317209029610084,
+      "loss": 3.4018,
+      "step": 388
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018311345646437996,
+      "loss": 3.2718,
+      "step": 389
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018305482263265905,
+      "loss": 3.3853,
+      "step": 390
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018299618880093815,
+      "loss": 3.3667,
+      "step": 391
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018293755496921724,
+      "loss": 3.387,
+      "step": 392
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018287892113749634,
+      "loss": 3.294,
+      "step": 393
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018282028730577546,
+      "loss": 3.4073,
+      "step": 394
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018276165347405455,
+      "loss": 3.3658,
+      "step": 395
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018270301964233362,
+      "loss": 3.2588,
+      "step": 396
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018264438581061274,
+      "loss": 3.3282,
+      "step": 397
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018258575197889184,
+      "loss": 3.2229,
+      "step": 398
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018252711814717093,
+      "loss": 3.2904,
+      "step": 399
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018246848431545002,
+      "loss": 3.3482,
+      "step": 400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018240985048372912,
+      "loss": 3.3721,
+      "step": 401
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00018235121665200824,
+      "loss": 3.3032,
+      "step": 402
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001822925828202873,
+      "loss": 3.3469,
+      "step": 403
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001822339489885664,
+      "loss": 3.2639,
+      "step": 404
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018217531515684552,
+      "loss": 3.421,
+      "step": 405
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018211668132512462,
+      "loss": 3.4373,
+      "step": 406
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018205804749340368,
+      "loss": 3.3135,
+      "step": 407
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001819994136616828,
+      "loss": 3.2879,
+      "step": 408
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001819407798299619,
+      "loss": 3.3979,
+      "step": 409
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000181882145998241,
+      "loss": 3.2865,
+      "step": 410
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001818235121665201,
+      "loss": 3.3575,
+      "step": 411
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018176487833479918,
+      "loss": 3.3499,
+      "step": 412
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001817062445030783,
+      "loss": 3.2888,
+      "step": 413
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018164761067135737,
+      "loss": 3.3062,
+      "step": 414
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018158897683963647,
+      "loss": 3.4237,
+      "step": 415
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001815303430079156,
+      "loss": 3.2535,
+      "step": 416
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018147170917619468,
+      "loss": 3.3479,
+      "step": 417
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018141307534447375,
+      "loss": 3.324,
+      "step": 418
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018135444151275287,
+      "loss": 3.3864,
+      "step": 419
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018129580768103196,
+      "loss": 3.2319,
+      "step": 420
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018123717384931106,
+      "loss": 3.3216,
+      "step": 421
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018117854001759015,
+      "loss": 3.2657,
+      "step": 422
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018111990618586925,
+      "loss": 3.3169,
+      "step": 423
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018106127235414837,
+      "loss": 3.3759,
+      "step": 424
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018100263852242744,
+      "loss": 3.4322,
+      "step": 425
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018094400469070653,
+      "loss": 3.3721,
+      "step": 426
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018088537085898565,
+      "loss": 3.3239,
+      "step": 427
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018082673702726475,
+      "loss": 3.3027,
+      "step": 428
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018076810319554384,
+      "loss": 3.3012,
+      "step": 429
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018070946936382293,
+      "loss": 3.2466,
+      "step": 430
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018065083553210203,
+      "loss": 3.3689,
+      "step": 431
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018059220170038112,
+      "loss": 3.2777,
+      "step": 432
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018053356786866022,
+      "loss": 3.3685,
+      "step": 433
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001804749340369393,
+      "loss": 3.4021,
+      "step": 434
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018041630020521843,
+      "loss": 3.3168,
+      "step": 435
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001803576663734975,
+      "loss": 3.2871,
+      "step": 436
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018029903254177662,
+      "loss": 3.2451,
+      "step": 437
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00018024039871005572,
+      "loss": 3.3165,
+      "step": 438
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001801817648783348,
+      "loss": 3.3736,
+      "step": 439
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001801231310466139,
+      "loss": 3.2207,
+      "step": 440
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000180064497214893,
+      "loss": 3.3741,
+      "step": 441
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001800058633831721,
+      "loss": 3.34,
+      "step": 442
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017994722955145122,
+      "loss": 3.3197,
+      "step": 443
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017988859571973028,
+      "loss": 3.2689,
+      "step": 444
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017982996188800938,
+      "loss": 3.3944,
+      "step": 445
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001797713280562885,
+      "loss": 3.3577,
+      "step": 446
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001797126942245676,
+      "loss": 3.3384,
+      "step": 447
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001796540603928467,
+      "loss": 3.3514,
+      "step": 448
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017959542656112578,
+      "loss": 3.2715,
+      "step": 449
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017953679272940488,
+      "loss": 3.4112,
+      "step": 450
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017947815889768397,
+      "loss": 3.2883,
+      "step": 451
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017941952506596306,
+      "loss": 3.3312,
+      "step": 452
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017936089123424216,
+      "loss": 3.399,
+      "step": 453
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017930225740252128,
+      "loss": 3.2589,
+      "step": 454
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017924362357080035,
+      "loss": 3.3033,
+      "step": 455
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017918498973907947,
+      "loss": 3.2986,
+      "step": 456
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017912635590735856,
+      "loss": 3.1767,
+      "step": 457
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017906772207563766,
+      "loss": 3.3192,
+      "step": 458
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017900908824391675,
+      "loss": 3.3241,
+      "step": 459
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017895045441219585,
+      "loss": 3.201,
+      "step": 460
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017889182058047494,
+      "loss": 3.2929,
+      "step": 461
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017883318674875403,
+      "loss": 3.3327,
+      "step": 462
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017877455291703313,
+      "loss": 3.418,
+      "step": 463
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017871591908531225,
+      "loss": 3.3366,
+      "step": 464
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017865728525359134,
+      "loss": 3.2592,
+      "step": 465
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001785986514218704,
+      "loss": 3.3467,
+      "step": 466
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017854001759014953,
+      "loss": 3.3619,
+      "step": 467
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017848138375842863,
+      "loss": 3.3509,
+      "step": 468
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017842274992670772,
+      "loss": 3.1956,
+      "step": 469
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017836411609498682,
+      "loss": 3.3698,
+      "step": 470
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001783054822632659,
+      "loss": 3.321,
+      "step": 471
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000178246848431545,
+      "loss": 3.2352,
+      "step": 472
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001781882145998241,
+      "loss": 3.224,
+      "step": 473
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001781295807681032,
+      "loss": 3.3287,
+      "step": 474
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017807094693638231,
+      "loss": 3.2841,
+      "step": 475
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001780123131046614,
+      "loss": 3.3262,
+      "step": 476
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017795367927294048,
+      "loss": 3.3455,
+      "step": 477
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001778950454412196,
+      "loss": 3.2282,
+      "step": 478
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001778364116094987,
+      "loss": 3.3557,
+      "step": 479
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 3.2092,
+      "step": 480
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017771914394605688,
+      "loss": 3.3508,
+      "step": 481
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017766051011433597,
+      "loss": 3.3306,
+      "step": 482
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001776018762826151,
+      "loss": 3.1666,
+      "step": 483
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017754324245089416,
+      "loss": 3.3202,
+      "step": 484
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017748460861917326,
+      "loss": 3.3197,
+      "step": 485
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017742597478745238,
+      "loss": 3.2957,
+      "step": 486
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017736734095573147,
+      "loss": 3.3891,
+      "step": 487
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017730870712401054,
+      "loss": 3.4085,
+      "step": 488
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017725007329228966,
+      "loss": 3.3088,
+      "step": 489
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017719143946056876,
+      "loss": 3.3182,
+      "step": 490
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017713280562884785,
+      "loss": 3.2542,
+      "step": 491
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017707417179712695,
+      "loss": 3.2773,
+      "step": 492
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017701553796540604,
+      "loss": 3.2836,
+      "step": 493
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017695690413368516,
+      "loss": 3.2482,
+      "step": 494
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017689827030196426,
+      "loss": 3.3793,
+      "step": 495
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017683963647024332,
+      "loss": 3.405,
+      "step": 496
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017678100263852244,
+      "loss": 3.2463,
+      "step": 497
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017672236880680154,
+      "loss": 3.2425,
+      "step": 498
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017666373497508063,
+      "loss": 3.3272,
+      "step": 499
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017660510114335973,
+      "loss": 3.3286,
+      "step": 500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017654646731163882,
+      "loss": 3.2452,
+      "step": 501
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017648783347991794,
+      "loss": 3.2671,
+      "step": 502
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000176429199648197,
+      "loss": 3.3648,
+      "step": 503
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001763705658164761,
+      "loss": 3.3606,
+      "step": 504
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017631193198475523,
+      "loss": 3.2764,
+      "step": 505
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017625329815303432,
+      "loss": 3.2594,
+      "step": 506
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001761946643213134,
+      "loss": 3.317,
+      "step": 507
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001761360304895925,
+      "loss": 3.2618,
+      "step": 508
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001760773966578716,
+      "loss": 3.4423,
+      "step": 509
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001760187628261507,
+      "loss": 3.3306,
+      "step": 510
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001759601289944298,
+      "loss": 3.3913,
+      "step": 511
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017590149516270889,
+      "loss": 3.4276,
+      "step": 512
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000175842861330988,
+      "loss": 3.2858,
+      "step": 513
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017578422749926707,
+      "loss": 3.4223,
+      "step": 514
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017572559366754617,
+      "loss": 3.383,
+      "step": 515
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001756669598358253,
+      "loss": 3.1365,
+      "step": 516
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017560832600410438,
+      "loss": 3.2742,
+      "step": 517
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017554969217238348,
+      "loss": 3.2793,
+      "step": 518
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017549105834066257,
+      "loss": 3.2637,
+      "step": 519
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017543242450894167,
+      "loss": 3.3172,
+      "step": 520
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017537379067722076,
+      "loss": 3.2949,
+      "step": 521
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017531515684549986,
+      "loss": 3.3339,
+      "step": 522
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017525652301377895,
+      "loss": 3.1933,
+      "step": 523
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017519788918205807,
+      "loss": 3.3356,
+      "step": 524
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017513925535033714,
+      "loss": 3.2832,
+      "step": 525
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017508062151861623,
+      "loss": 3.2465,
+      "step": 526
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017502198768689535,
+      "loss": 3.2227,
+      "step": 527
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017496335385517445,
+      "loss": 3.2431,
+      "step": 528
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 3.4339988231658936,
+      "eval_runtime": 2941.2933,
+      "eval_samples_per_second": 6.965,
+      "eval_steps_per_second": 2.322,
+      "step": 528
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017490472002345354,
+      "loss": 3.2959,
+      "step": 529
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017484608619173264,
+      "loss": 3.3558,
+      "step": 530
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017478745236001173,
+      "loss": 3.3577,
+      "step": 531
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017472881852829083,
+      "loss": 3.3003,
+      "step": 532
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017467018469656992,
+      "loss": 3.2796,
+      "step": 533
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017461155086484902,
+      "loss": 3.3259,
+      "step": 534
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017455291703312814,
+      "loss": 3.3561,
+      "step": 535
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001744942832014072,
+      "loss": 3.3565,
+      "step": 536
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017443564936968633,
+      "loss": 3.3997,
+      "step": 537
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017437701553796542,
+      "loss": 3.276,
+      "step": 538
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017431838170624451,
+      "loss": 3.285,
+      "step": 539
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001742597478745236,
+      "loss": 3.2681,
+      "step": 540
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001742011140428027,
+      "loss": 3.3277,
+      "step": 541
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001741424802110818,
+      "loss": 3.3858,
+      "step": 542
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001740838463793609,
+      "loss": 3.337,
+      "step": 543
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00017402521254763999,
+      "loss": 3.2288,
+      "step": 544
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001739665787159191,
+      "loss": 3.3113,
+      "step": 545
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001739079448841982,
+      "loss": 3.36,
+      "step": 546
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017384931105247727,
+      "loss": 3.285,
+      "step": 547
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001737906772207564,
+      "loss": 3.2721,
+      "step": 548
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017373204338903548,
+      "loss": 3.2411,
+      "step": 549
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017367340955731458,
+      "loss": 3.2043,
+      "step": 550
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017361477572559367,
+      "loss": 3.2924,
+      "step": 551
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017355614189387277,
+      "loss": 3.2879,
+      "step": 552
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017349750806215186,
+      "loss": 3.292,
+      "step": 553
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017343887423043098,
+      "loss": 3.3447,
+      "step": 554
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017338024039871005,
+      "loss": 3.304,
+      "step": 555
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017332160656698917,
+      "loss": 3.298,
+      "step": 556
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017326297273526827,
+      "loss": 3.2038,
+      "step": 557
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017320433890354736,
+      "loss": 3.3378,
+      "step": 558
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017314570507182645,
+      "loss": 3.3532,
+      "step": 559
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017308707124010555,
+      "loss": 3.2387,
+      "step": 560
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017302843740838464,
+      "loss": 3.2829,
+      "step": 561
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017296980357666374,
+      "loss": 3.2958,
+      "step": 562
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017291116974494283,
+      "loss": 3.4214,
+      "step": 563
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017285253591322195,
+      "loss": 3.2844,
+      "step": 564
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017279390208150105,
+      "loss": 3.4155,
+      "step": 565
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017273526824978011,
+      "loss": 3.3632,
+      "step": 566
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017267663441805924,
+      "loss": 3.3174,
+      "step": 567
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017261800058633833,
+      "loss": 3.263,
+      "step": 568
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017255936675461742,
+      "loss": 3.3077,
+      "step": 569
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017250073292289652,
+      "loss": 3.3081,
+      "step": 570
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001724420990911756,
+      "loss": 3.2987,
+      "step": 571
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017238346525945473,
+      "loss": 3.3379,
+      "step": 572
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001723248314277338,
+      "loss": 3.324,
+      "step": 573
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001722661975960129,
+      "loss": 3.136,
+      "step": 574
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017220756376429202,
+      "loss": 3.261,
+      "step": 575
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001721489299325711,
+      "loss": 3.3249,
+      "step": 576
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00017209029610085018,
+      "loss": 3.2153,
+      "step": 577
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001720316622691293,
+      "loss": 3.2558,
+      "step": 578
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001719730284374084,
+      "loss": 3.2936,
+      "step": 579
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001719143946056875,
+      "loss": 3.3077,
+      "step": 580
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017185576077396658,
+      "loss": 3.3064,
+      "step": 581
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017179712694224568,
+      "loss": 3.3206,
+      "step": 582
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001717384931105248,
+      "loss": 3.3589,
+      "step": 583
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017167985927880387,
+      "loss": 3.2409,
+      "step": 584
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017162122544708296,
+      "loss": 3.3868,
+      "step": 585
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017156259161536208,
+      "loss": 3.3109,
+      "step": 586
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017150395778364118,
+      "loss": 3.3553,
+      "step": 587
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017144532395192024,
+      "loss": 3.4257,
+      "step": 588
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017138669012019937,
+      "loss": 3.3823,
+      "step": 589
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017132805628847846,
+      "loss": 3.3419,
+      "step": 590
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017126942245675755,
+      "loss": 3.3304,
+      "step": 591
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017121078862503665,
+      "loss": 3.3838,
+      "step": 592
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017115215479331574,
+      "loss": 3.1078,
+      "step": 593
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017109352096159486,
+      "loss": 3.351,
+      "step": 594
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017103488712987393,
+      "loss": 3.2375,
+      "step": 595
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017097625329815303,
+      "loss": 3.2834,
+      "step": 596
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017091761946643215,
+      "loss": 3.373,
+      "step": 597
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017085898563471124,
+      "loss": 3.3146,
+      "step": 598
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017080035180299034,
+      "loss": 3.3141,
+      "step": 599
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017074171797126943,
+      "loss": 3.2932,
+      "step": 600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017068308413954852,
+      "loss": 3.3591,
+      "step": 601
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017062445030782765,
+      "loss": 3.2531,
+      "step": 602
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001705658164761067,
+      "loss": 3.2469,
+      "step": 603
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001705071826443858,
+      "loss": 3.2581,
+      "step": 604
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017044854881266493,
+      "loss": 3.421,
+      "step": 605
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017038991498094402,
+      "loss": 3.2698,
+      "step": 606
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00017033128114922312,
+      "loss": 3.301,
+      "step": 607
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001702726473175022,
+      "loss": 3.2526,
+      "step": 608
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001702140134857813,
+      "loss": 3.318,
+      "step": 609
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001701553796540604,
+      "loss": 3.4293,
+      "step": 610
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001700967458223395,
+      "loss": 3.3258,
+      "step": 611
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001700381119906186,
+      "loss": 3.2636,
+      "step": 612
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001699794781588977,
+      "loss": 3.277,
+      "step": 613
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00016992084432717678,
+      "loss": 3.2717,
+      "step": 614
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016986221049545587,
+      "loss": 3.2445,
+      "step": 615
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000169803576663735,
+      "loss": 3.3164,
+      "step": 616
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001697449428320141,
+      "loss": 3.3404,
+      "step": 617
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016968630900029318,
+      "loss": 3.3986,
+      "step": 618
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016962767516857228,
+      "loss": 3.3836,
+      "step": 619
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016956904133685137,
+      "loss": 3.3015,
+      "step": 620
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016951040750513046,
+      "loss": 3.2088,
+      "step": 621
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016945177367340956,
+      "loss": 3.2471,
+      "step": 622
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016939313984168865,
+      "loss": 3.2617,
+      "step": 623
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016933450600996777,
+      "loss": 3.3185,
+      "step": 624
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016927587217824684,
+      "loss": 3.342,
+      "step": 625
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016921723834652596,
+      "loss": 3.3705,
+      "step": 626
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016915860451480506,
+      "loss": 3.2627,
+      "step": 627
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016909997068308415,
+      "loss": 3.2532,
+      "step": 628
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016904133685136325,
+      "loss": 3.2419,
+      "step": 629
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016898270301964234,
+      "loss": 3.2455,
+      "step": 630
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016892406918792143,
+      "loss": 3.1866,
+      "step": 631
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016886543535620053,
+      "loss": 3.3726,
+      "step": 632
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016880680152447962,
+      "loss": 3.2777,
+      "step": 633
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016874816769275875,
+      "loss": 3.3045,
+      "step": 634
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016868953386103784,
+      "loss": 3.2767,
+      "step": 635
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001686309000293169,
+      "loss": 3.2479,
+      "step": 636
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016857226619759603,
+      "loss": 3.2611,
+      "step": 637
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016851363236587512,
+      "loss": 3.1176,
+      "step": 638
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016845499853415422,
+      "loss": 3.1545,
+      "step": 639
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001683963647024333,
+      "loss": 3.3133,
+      "step": 640
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001683377308707124,
+      "loss": 3.4034,
+      "step": 641
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001682790970389915,
+      "loss": 3.2923,
+      "step": 642
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001682204632072706,
+      "loss": 3.2192,
+      "step": 643
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001681618293755497,
+      "loss": 3.3476,
+      "step": 644
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001681031955438288,
+      "loss": 3.2615,
+      "step": 645
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001680445617121079,
+      "loss": 3.2505,
+      "step": 646
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00016798592788038697,
+      "loss": 3.1758,
+      "step": 647
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001679272940486661,
+      "loss": 3.2939,
+      "step": 648
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001678686602169452,
+      "loss": 3.2557,
+      "step": 649
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016781002638522428,
+      "loss": 3.3448,
+      "step": 650
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016775139255350338,
+      "loss": 3.3577,
+      "step": 651
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016769275872178247,
+      "loss": 3.1448,
+      "step": 652
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001676341248900616,
+      "loss": 3.3233,
+      "step": 653
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016757549105834069,
+      "loss": 3.2937,
+      "step": 654
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016751685722661975,
+      "loss": 3.1965,
+      "step": 655
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016745822339489887,
+      "loss": 3.284,
+      "step": 656
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016739958956317797,
+      "loss": 3.2228,
+      "step": 657
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016734095573145706,
+      "loss": 3.2362,
+      "step": 658
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016728232189973616,
+      "loss": 3.2238,
+      "step": 659
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016722368806801525,
+      "loss": 3.3228,
+      "step": 660
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016716505423629437,
+      "loss": 3.2826,
+      "step": 661
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016710642040457344,
+      "loss": 3.3176,
+      "step": 662
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016704778657285253,
+      "loss": 3.2703,
+      "step": 663
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016698915274113166,
+      "loss": 3.2819,
+      "step": 664
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016693051890941075,
+      "loss": 3.3043,
+      "step": 665
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016687188507768982,
+      "loss": 3.344,
+      "step": 666
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016681325124596894,
+      "loss": 3.2764,
+      "step": 667
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016675461741424803,
+      "loss": 3.2562,
+      "step": 668
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016669598358252713,
+      "loss": 3.2733,
+      "step": 669
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016663734975080622,
+      "loss": 3.2605,
+      "step": 670
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016657871591908532,
+      "loss": 3.2495,
+      "step": 671
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016652008208736444,
+      "loss": 3.2638,
+      "step": 672
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001664614482556435,
+      "loss": 3.3926,
+      "step": 673
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001664028144239226,
+      "loss": 3.1857,
+      "step": 674
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016634418059220172,
+      "loss": 3.2965,
+      "step": 675
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016628554676048081,
+      "loss": 3.3535,
+      "step": 676
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016622691292875988,
+      "loss": 3.3562,
+      "step": 677
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000166168279097039,
+      "loss": 3.2641,
+      "step": 678
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001661096452653181,
+      "loss": 3.2676,
+      "step": 679
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001660510114335972,
+      "loss": 3.2632,
+      "step": 680
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016599237760187629,
+      "loss": 3.2399,
+      "step": 681
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016593374377015538,
+      "loss": 3.3722,
+      "step": 682
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001658751099384345,
+      "loss": 3.1751,
+      "step": 683
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00016581647610671357,
+      "loss": 3.304,
+      "step": 684
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016575784227499266,
+      "loss": 3.3627,
+      "step": 685
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016569920844327179,
+      "loss": 3.2769,
+      "step": 686
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016564057461155088,
+      "loss": 3.333,
+      "step": 687
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016558194077982997,
+      "loss": 3.2288,
+      "step": 688
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016552330694810907,
+      "loss": 3.2726,
+      "step": 689
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016546467311638816,
+      "loss": 3.2776,
+      "step": 690
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016540603928466726,
+      "loss": 3.2747,
+      "step": 691
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016534740545294635,
+      "loss": 3.2204,
+      "step": 692
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016528877162122545,
+      "loss": 3.2548,
+      "step": 693
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016523013778950457,
+      "loss": 3.2743,
+      "step": 694
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016517150395778363,
+      "loss": 3.33,
+      "step": 695
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016511287012606273,
+      "loss": 3.3194,
+      "step": 696
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016505423629434185,
+      "loss": 3.2658,
+      "step": 697
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016499560246262094,
+      "loss": 3.1893,
+      "step": 698
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016493696863090004,
+      "loss": 3.3421,
+      "step": 699
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016487833479917913,
+      "loss": 3.2042,
+      "step": 700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016481970096745823,
+      "loss": 3.3802,
+      "step": 701
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016476106713573732,
+      "loss": 3.272,
+      "step": 702
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016470243330401642,
+      "loss": 3.3066,
+      "step": 703
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001646437994722955,
+      "loss": 3.3326,
+      "step": 704
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 3.4172792434692383,
+      "eval_runtime": 2941.3473,
+      "eval_samples_per_second": 6.965,
+      "eval_steps_per_second": 2.322,
+      "step": 704
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016458516564057463,
+      "loss": 3.2473,
+      "step": 705
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016452653180885373,
+      "loss": 3.2433,
+      "step": 706
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016446789797713282,
+      "loss": 3.2078,
+      "step": 707
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016440926414541191,
+      "loss": 3.3541,
+      "step": 708
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000164350630313691,
+      "loss": 3.394,
+      "step": 709
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001642919964819701,
+      "loss": 3.3283,
+      "step": 710
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001642333626502492,
+      "loss": 3.2292,
+      "step": 711
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001641747288185283,
+      "loss": 3.3035,
+      "step": 712
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001641160949868074,
+      "loss": 3.3012,
+      "step": 713
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016405746115508648,
+      "loss": 3.182,
+      "step": 714
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001639988273233656,
+      "loss": 3.349,
+      "step": 715
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001639401934916447,
+      "loss": 3.3429,
+      "step": 716
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001638815596599238,
+      "loss": 3.3554,
+      "step": 717
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016382292582820288,
+      "loss": 3.2935,
+      "step": 718
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00016376429199648198,
+      "loss": 3.3004,
+      "step": 719
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016370565816476107,
+      "loss": 3.2854,
+      "step": 720
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016364702433304017,
+      "loss": 3.3397,
+      "step": 721
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016358839050131926,
+      "loss": 3.2613,
+      "step": 722
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016352975666959836,
+      "loss": 3.3017,
+      "step": 723
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016347112283787748,
+      "loss": 3.1922,
+      "step": 724
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016341248900615654,
+      "loss": 3.3134,
+      "step": 725
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016335385517443567,
+      "loss": 3.2636,
+      "step": 726
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016329522134271476,
+      "loss": 3.2879,
+      "step": 727
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016323658751099385,
+      "loss": 3.2707,
+      "step": 728
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016317795367927295,
+      "loss": 3.2435,
+      "step": 729
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016311931984755204,
+      "loss": 3.2848,
+      "step": 730
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016306068601583114,
+      "loss": 3.4065,
+      "step": 731
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016300205218411023,
+      "loss": 3.194,
+      "step": 732
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016294341835238933,
+      "loss": 3.2039,
+      "step": 733
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016288478452066845,
+      "loss": 3.2882,
+      "step": 734
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016282615068894754,
+      "loss": 3.3171,
+      "step": 735
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001627675168572266,
+      "loss": 3.3337,
+      "step": 736
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016270888302550573,
+      "loss": 3.2181,
+      "step": 737
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016265024919378483,
+      "loss": 3.318,
+      "step": 738
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016259161536206392,
+      "loss": 3.2269,
+      "step": 739
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016253298153034301,
+      "loss": 3.3264,
+      "step": 740
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001624743476986221,
+      "loss": 3.2743,
+      "step": 741
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016241571386690123,
+      "loss": 3.2306,
+      "step": 742
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001623570800351803,
+      "loss": 3.3459,
+      "step": 743
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001622984462034594,
+      "loss": 3.3822,
+      "step": 744
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001622398123717385,
+      "loss": 3.3037,
+      "step": 745
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001621811785400176,
+      "loss": 3.2747,
+      "step": 746
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016212254470829667,
+      "loss": 3.2404,
+      "step": 747
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001620639108765758,
+      "loss": 3.1767,
+      "step": 748
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001620052770448549,
+      "loss": 3.3484,
+      "step": 749
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016194664321313398,
+      "loss": 3.2865,
+      "step": 750
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016188800938141308,
+      "loss": 3.224,
+      "step": 751
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016182937554969217,
+      "loss": 3.1685,
+      "step": 752
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001617707417179713,
+      "loss": 3.1998,
+      "step": 753
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016171210788625036,
+      "loss": 3.2195,
+      "step": 754
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00016165347405452946,
+      "loss": 3.3184,
+      "step": 755
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016159484022280858,
+      "loss": 3.2806,
+      "step": 756
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016153620639108767,
+      "loss": 3.2766,
+      "step": 757
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016147757255936674,
+      "loss": 3.2813,
+      "step": 758
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016141893872764586,
+      "loss": 3.247,
+      "step": 759
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016136030489592495,
+      "loss": 3.3013,
+      "step": 760
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016130167106420408,
+      "loss": 3.2898,
+      "step": 761
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016124303723248314,
+      "loss": 3.193,
+      "step": 762
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016118440340076224,
+      "loss": 3.2976,
+      "step": 763
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016112576956904136,
+      "loss": 3.3017,
+      "step": 764
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016106713573732045,
+      "loss": 3.3193,
+      "step": 765
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016100850190559952,
+      "loss": 3.2682,
+      "step": 766
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016094986807387864,
+      "loss": 3.2774,
+      "step": 767
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016089123424215774,
+      "loss": 3.3002,
+      "step": 768
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016083260041043683,
+      "loss": 3.1781,
+      "step": 769
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016077396657871592,
+      "loss": 3.2766,
+      "step": 770
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016071533274699502,
+      "loss": 3.2483,
+      "step": 771
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016065669891527414,
+      "loss": 3.2515,
+      "step": 772
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001605980650835532,
+      "loss": 3.2771,
+      "step": 773
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001605394312518323,
+      "loss": 3.2535,
+      "step": 774
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016048079742011142,
+      "loss": 3.3926,
+      "step": 775
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016042216358839052,
+      "loss": 3.2368,
+      "step": 776
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001603635297566696,
+      "loss": 3.1698,
+      "step": 777
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001603048959249487,
+      "loss": 3.2374,
+      "step": 778
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001602462620932278,
+      "loss": 3.3045,
+      "step": 779
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001601876282615069,
+      "loss": 3.229,
+      "step": 780
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.000160128994429786,
+      "loss": 3.278,
+      "step": 781
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00016007036059806508,
+      "loss": 3.2746,
+      "step": 782
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001600117267663442,
+      "loss": 3.2816,
+      "step": 783
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015995309293462327,
+      "loss": 3.2566,
+      "step": 784
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015989445910290237,
+      "loss": 3.3259,
+      "step": 785
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001598358252711815,
+      "loss": 3.2119,
+      "step": 786
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015977719143946058,
+      "loss": 3.0796,
+      "step": 787
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015971855760773968,
+      "loss": 3.2004,
+      "step": 788
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015965992377601877,
+      "loss": 3.3308,
+      "step": 789
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00015960128994429787,
+      "loss": 3.3194,
+      "step": 790
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015954265611257696,
+      "loss": 3.3085,
+      "step": 791
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015948402228085605,
+      "loss": 3.2244,
+      "step": 792
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015942538844913515,
+      "loss": 3.0929,
+      "step": 793
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015936675461741427,
+      "loss": 3.236,
+      "step": 794
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015930812078569334,
+      "loss": 3.2935,
+      "step": 795
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015924948695397246,
+      "loss": 3.2709,
+      "step": 796
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015919085312225155,
+      "loss": 3.2316,
+      "step": 797
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015913221929053065,
+      "loss": 3.2388,
+      "step": 798
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015907358545880974,
+      "loss": 3.2513,
+      "step": 799
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015901495162708884,
+      "loss": 3.1954,
+      "step": 800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015895631779536793,
+      "loss": 3.2512,
+      "step": 801
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015889768396364702,
+      "loss": 3.2425,
+      "step": 802
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015883905013192612,
+      "loss": 3.2533,
+      "step": 803
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015878041630020524,
+      "loss": 3.3006,
+      "step": 804
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015872178246848433,
+      "loss": 3.3069,
+      "step": 805
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001586631486367634,
+      "loss": 3.2092,
+      "step": 806
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015860451480504252,
+      "loss": 3.2734,
+      "step": 807
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015854588097332162,
+      "loss": 3.2091,
+      "step": 808
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001584872471416007,
+      "loss": 3.2802,
+      "step": 809
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001584286133098798,
+      "loss": 3.2291,
+      "step": 810
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001583699794781589,
+      "loss": 3.3642,
+      "step": 811
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000158311345646438,
+      "loss": 3.2719,
+      "step": 812
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015825271181471712,
+      "loss": 3.3143,
+      "step": 813
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015819407798299618,
+      "loss": 3.1932,
+      "step": 814
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001581354441512753,
+      "loss": 3.353,
+      "step": 815
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001580768103195544,
+      "loss": 3.1533,
+      "step": 816
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001580181764878335,
+      "loss": 3.2604,
+      "step": 817
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001579595426561126,
+      "loss": 3.1985,
+      "step": 818
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015790090882439168,
+      "loss": 3.2331,
+      "step": 819
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015784227499267078,
+      "loss": 3.1971,
+      "step": 820
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015778364116094987,
+      "loss": 3.276,
+      "step": 821
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015772500732922896,
+      "loss": 3.2724,
+      "step": 822
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015766637349750809,
+      "loss": 3.2373,
+      "step": 823
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015760773966578718,
+      "loss": 3.23,
+      "step": 824
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00015754910583406625,
+      "loss": 3.2432,
+      "step": 825
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015749047200234537,
+      "loss": 3.2401,
+      "step": 826
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015743183817062446,
+      "loss": 3.2634,
+      "step": 827
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015737320433890356,
+      "loss": 3.3252,
+      "step": 828
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015731457050718265,
+      "loss": 3.2793,
+      "step": 829
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015725593667546175,
+      "loss": 3.2755,
+      "step": 830
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015719730284374087,
+      "loss": 3.3025,
+      "step": 831
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015713866901201994,
+      "loss": 3.1987,
+      "step": 832
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015708003518029903,
+      "loss": 3.3599,
+      "step": 833
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015702140134857815,
+      "loss": 3.3191,
+      "step": 834
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015696276751685725,
+      "loss": 3.2095,
+      "step": 835
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001569041336851363,
+      "loss": 3.2534,
+      "step": 836
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015684549985341543,
+      "loss": 3.3055,
+      "step": 837
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015678686602169453,
+      "loss": 3.2446,
+      "step": 838
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015672823218997362,
+      "loss": 3.3143,
+      "step": 839
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015666959835825272,
+      "loss": 3.2018,
+      "step": 840
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001566109645265318,
+      "loss": 3.2666,
+      "step": 841
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015655233069481093,
+      "loss": 3.249,
+      "step": 842
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015649369686309,
+      "loss": 3.3042,
+      "step": 843
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001564350630313691,
+      "loss": 3.2625,
+      "step": 844
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015637642919964822,
+      "loss": 3.2064,
+      "step": 845
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001563177953679273,
+      "loss": 3.1942,
+      "step": 846
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015625916153620638,
+      "loss": 3.2523,
+      "step": 847
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001562005277044855,
+      "loss": 3.1982,
+      "step": 848
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001561418938727646,
+      "loss": 3.244,
+      "step": 849
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001560832600410437,
+      "loss": 3.2332,
+      "step": 850
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015602462620932278,
+      "loss": 3.2345,
+      "step": 851
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015596599237760188,
+      "loss": 3.1666,
+      "step": 852
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000155907358545881,
+      "loss": 3.2655,
+      "step": 853
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015584872471416006,
+      "loss": 3.211,
+      "step": 854
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015579009088243916,
+      "loss": 3.0659,
+      "step": 855
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015573145705071828,
+      "loss": 3.2308,
+      "step": 856
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015567282321899737,
+      "loss": 3.1882,
+      "step": 857
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015561418938727647,
+      "loss": 3.2903,
+      "step": 858
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 3.2234,
+      "step": 859
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00015549692172383466,
+      "loss": 3.268,
+      "step": 860
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015543828789211378,
+      "loss": 3.245,
+      "step": 861
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015537965406039285,
+      "loss": 3.2705,
+      "step": 862
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015532102022867194,
+      "loss": 3.2247,
+      "step": 863
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015526238639695106,
+      "loss": 3.2725,
+      "step": 864
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015520375256523016,
+      "loss": 3.2226,
+      "step": 865
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015514511873350922,
+      "loss": 3.325,
+      "step": 866
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015508648490178834,
+      "loss": 3.1678,
+      "step": 867
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015502785107006744,
+      "loss": 3.1907,
+      "step": 868
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015496921723834653,
+      "loss": 3.2839,
+      "step": 869
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015491058340662563,
+      "loss": 3.2195,
+      "step": 870
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015485194957490472,
+      "loss": 3.2793,
+      "step": 871
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015479331574318384,
+      "loss": 3.2944,
+      "step": 872
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001547346819114629,
+      "loss": 3.2232,
+      "step": 873
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000154676048079742,
+      "loss": 3.2663,
+      "step": 874
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015461741424802113,
+      "loss": 3.229,
+      "step": 875
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015455878041630022,
+      "loss": 3.2287,
+      "step": 876
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015450014658457931,
+      "loss": 3.184,
+      "step": 877
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001544415127528584,
+      "loss": 3.2551,
+      "step": 878
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001543828789211375,
+      "loss": 3.2443,
+      "step": 879
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001543242450894166,
+      "loss": 3.1881,
+      "step": 880
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 3.4091506004333496,
+      "eval_runtime": 2941.0944,
+      "eval_samples_per_second": 6.966,
+      "eval_steps_per_second": 2.322,
+      "step": 880
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001542656112576957,
+      "loss": 3.1824,
+      "step": 881
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001542069774259748,
+      "loss": 3.2434,
+      "step": 882
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001541483435942539,
+      "loss": 3.177,
+      "step": 883
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015408970976253298,
+      "loss": 3.2763,
+      "step": 884
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001540310759308121,
+      "loss": 3.2472,
+      "step": 885
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001539724420990912,
+      "loss": 3.1991,
+      "step": 886
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015391380826737029,
+      "loss": 3.2498,
+      "step": 887
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015385517443564938,
+      "loss": 3.2251,
+      "step": 888
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015379654060392847,
+      "loss": 3.2537,
+      "step": 889
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015373790677220757,
+      "loss": 3.244,
+      "step": 890
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015367927294048666,
+      "loss": 3.1972,
+      "step": 891
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015362063910876576,
+      "loss": 3.2767,
+      "step": 892
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015356200527704485,
+      "loss": 3.2709,
+      "step": 893
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015350337144532397,
+      "loss": 3.2772,
+      "step": 894
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00015344473761360304,
+      "loss": 3.3388,
+      "step": 895
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015338610378188216,
+      "loss": 3.21,
+      "step": 896
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015332746995016126,
+      "loss": 3.2313,
+      "step": 897
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015326883611844035,
+      "loss": 3.4058,
+      "step": 898
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015321020228671944,
+      "loss": 3.2922,
+      "step": 899
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015315156845499854,
+      "loss": 3.2429,
+      "step": 900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015309293462327763,
+      "loss": 3.3019,
+      "step": 901
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015303430079155673,
+      "loss": 3.2689,
+      "step": 902
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015297566695983582,
+      "loss": 3.2101,
+      "step": 903
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015291703312811494,
+      "loss": 3.2623,
+      "step": 904
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015285839929639404,
+      "loss": 3.2196,
+      "step": 905
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001527997654646731,
+      "loss": 3.2559,
+      "step": 906
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015274113163295223,
+      "loss": 3.2888,
+      "step": 907
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015268249780123132,
+      "loss": 3.1584,
+      "step": 908
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015262386396951041,
+      "loss": 3.1906,
+      "step": 909
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001525652301377895,
+      "loss": 3.2,
+      "step": 910
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001525065963060686,
+      "loss": 3.3503,
+      "step": 911
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015244796247434772,
+      "loss": 3.3084,
+      "step": 912
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001523893286426268,
+      "loss": 3.2494,
+      "step": 913
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015233069481090589,
+      "loss": 3.2489,
+      "step": 914
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000152272060979185,
+      "loss": 3.2155,
+      "step": 915
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001522134271474641,
+      "loss": 3.2881,
+      "step": 916
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001521547933157432,
+      "loss": 3.2205,
+      "step": 917
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001520961594840223,
+      "loss": 3.1929,
+      "step": 918
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015203752565230138,
+      "loss": 3.307,
+      "step": 919
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015197889182058048,
+      "loss": 3.2806,
+      "step": 920
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015192025798885957,
+      "loss": 3.2074,
+      "step": 921
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015186162415713867,
+      "loss": 3.176,
+      "step": 922
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001518029903254178,
+      "loss": 3.2676,
+      "step": 923
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015174435649369688,
+      "loss": 3.2985,
+      "step": 924
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015168572266197595,
+      "loss": 3.2657,
+      "step": 925
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015162708883025507,
+      "loss": 3.2334,
+      "step": 926
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015156845499853417,
+      "loss": 3.2857,
+      "step": 927
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015150982116681326,
+      "loss": 3.1691,
+      "step": 928
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015145118733509236,
+      "loss": 3.2631,
+      "step": 929
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00015139255350337145,
+      "loss": 3.2074,
+      "step": 930
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015133391967165057,
+      "loss": 3.1926,
+      "step": 931
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015127528583992964,
+      "loss": 3.2915,
+      "step": 932
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015121665200820873,
+      "loss": 3.3273,
+      "step": 933
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015115801817648785,
+      "loss": 3.2319,
+      "step": 934
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015109938434476695,
+      "loss": 3.2416,
+      "step": 935
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015104075051304602,
+      "loss": 3.2577,
+      "step": 936
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015098211668132514,
+      "loss": 3.1961,
+      "step": 937
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015092348284960423,
+      "loss": 3.2255,
+      "step": 938
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015086484901788333,
+      "loss": 3.2932,
+      "step": 939
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015080621518616242,
+      "loss": 3.2688,
+      "step": 940
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015074758135444151,
+      "loss": 3.1805,
+      "step": 941
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015068894752272064,
+      "loss": 3.2549,
+      "step": 942
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001506303136909997,
+      "loss": 3.2801,
+      "step": 943
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001505716798592788,
+      "loss": 3.2847,
+      "step": 944
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015051304602755792,
+      "loss": 3.2777,
+      "step": 945
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000150454412195837,
+      "loss": 3.2552,
+      "step": 946
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015039577836411608,
+      "loss": 3.3059,
+      "step": 947
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001503371445323952,
+      "loss": 3.2302,
+      "step": 948
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001502785107006743,
+      "loss": 3.3052,
+      "step": 949
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001502198768689534,
+      "loss": 3.2225,
+      "step": 950
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015016124303723248,
+      "loss": 3.3093,
+      "step": 951
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00015010260920551158,
+      "loss": 3.3046,
+      "step": 952
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001500439753737907,
+      "loss": 3.2808,
+      "step": 953
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014998534154206977,
+      "loss": 3.2288,
+      "step": 954
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014992670771034886,
+      "loss": 3.302,
+      "step": 955
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014986807387862798,
+      "loss": 3.2052,
+      "step": 956
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014980944004690708,
+      "loss": 3.2758,
+      "step": 957
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014975080621518617,
+      "loss": 3.2215,
+      "step": 958
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014969217238346527,
+      "loss": 3.2778,
+      "step": 959
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014963353855174436,
+      "loss": 3.2338,
+      "step": 960
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014957490472002345,
+      "loss": 3.2369,
+      "step": 961
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014951627088830255,
+      "loss": 3.1956,
+      "step": 962
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014945763705658164,
+      "loss": 3.188,
+      "step": 963
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014939900322486076,
+      "loss": 3.1952,
+      "step": 964
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014934036939313983,
+      "loss": 3.2734,
+      "step": 965
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014928173556141895,
+      "loss": 3.1608,
+      "step": 966
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014922310172969805,
+      "loss": 3.3127,
+      "step": 967
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014916446789797714,
+      "loss": 3.283,
+      "step": 968
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014910583406625624,
+      "loss": 3.2635,
+      "step": 969
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014904720023453533,
+      "loss": 3.2095,
+      "step": 970
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014898856640281442,
+      "loss": 3.2504,
+      "step": 971
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014892993257109355,
+      "loss": 3.2941,
+      "step": 972
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001488712987393726,
+      "loss": 3.3075,
+      "step": 973
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014881266490765173,
+      "loss": 3.2687,
+      "step": 974
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014875403107593083,
+      "loss": 3.1588,
+      "step": 975
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014869539724420992,
+      "loss": 3.2512,
+      "step": 976
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014863676341248902,
+      "loss": 3.3122,
+      "step": 977
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001485781295807681,
+      "loss": 3.1459,
+      "step": 978
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001485194957490472,
+      "loss": 3.2629,
+      "step": 979
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001484608619173263,
+      "loss": 3.1711,
+      "step": 980
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001484022280856054,
+      "loss": 3.1911,
+      "step": 981
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001483435942538845,
+      "loss": 3.2479,
+      "step": 982
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001482849604221636,
+      "loss": 3.2854,
+      "step": 983
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014822632659044268,
+      "loss": 3.2204,
+      "step": 984
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001481676927587218,
+      "loss": 3.33,
+      "step": 985
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001481090589270009,
+      "loss": 3.1618,
+      "step": 986
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014805042509528,
+      "loss": 3.2279,
+      "step": 987
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014799179126355908,
+      "loss": 3.1532,
+      "step": 988
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014793315743183818,
+      "loss": 3.3308,
+      "step": 989
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014787452360011727,
+      "loss": 3.1687,
+      "step": 990
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014781588976839637,
+      "loss": 3.2199,
+      "step": 991
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014775725593667546,
+      "loss": 3.3156,
+      "step": 992
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014769862210495458,
+      "loss": 3.3221,
+      "step": 993
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014763998827323368,
+      "loss": 3.2839,
+      "step": 994
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014758135444151274,
+      "loss": 3.2538,
+      "step": 995
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014752272060979186,
+      "loss": 3.1649,
+      "step": 996
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014746408677807096,
+      "loss": 3.2941,
+      "step": 997
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014740545294635005,
+      "loss": 3.1584,
+      "step": 998
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014734681911462915,
+      "loss": 3.1711,
+      "step": 999
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014728818528290824,
+      "loss": 3.0869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014722955145118736,
+      "loss": 3.1919,
+      "step": 1001
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014717091761946643,
+      "loss": 3.2672,
+      "step": 1002
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014711228378774552,
+      "loss": 3.1987,
+      "step": 1003
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014705364995602465,
+      "loss": 3.341,
+      "step": 1004
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014699501612430374,
+      "loss": 3.1695,
+      "step": 1005
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001469363822925828,
+      "loss": 3.1899,
+      "step": 1006
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014687774846086193,
+      "loss": 3.2716,
+      "step": 1007
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014681911462914102,
+      "loss": 3.2247,
+      "step": 1008
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014676048079742012,
+      "loss": 3.2712,
+      "step": 1009
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001467018469656992,
+      "loss": 3.2421,
+      "step": 1010
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001466432131339783,
+      "loss": 3.2861,
+      "step": 1011
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014658457930225743,
+      "loss": 3.2033,
+      "step": 1012
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001465259454705365,
+      "loss": 3.195,
+      "step": 1013
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001464673116388156,
+      "loss": 3.1744,
+      "step": 1014
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001464086778070947,
+      "loss": 3.2552,
+      "step": 1015
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001463500439753738,
+      "loss": 3.2764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014629141014365287,
+      "loss": 3.2084,
+      "step": 1017
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000146232776311932,
+      "loss": 3.2608,
+      "step": 1018
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001461741424802111,
+      "loss": 3.2078,
+      "step": 1019
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001461155086484902,
+      "loss": 3.2861,
+      "step": 1020
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014605687481676928,
+      "loss": 3.1761,
+      "step": 1021
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014599824098504837,
+      "loss": 3.258,
+      "step": 1022
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001459396071533275,
+      "loss": 3.2378,
+      "step": 1023
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014588097332160659,
+      "loss": 3.2288,
+      "step": 1024
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014582233948988565,
+      "loss": 3.2112,
+      "step": 1025
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014576370565816477,
+      "loss": 3.1836,
+      "step": 1026
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014570507182644387,
+      "loss": 3.2554,
+      "step": 1027
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014564643799472296,
+      "loss": 3.1601,
+      "step": 1028
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014558780416300206,
+      "loss": 3.1421,
+      "step": 1029
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014552917033128115,
+      "loss": 3.2407,
+      "step": 1030
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014547053649956027,
+      "loss": 3.1458,
+      "step": 1031
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014541190266783934,
+      "loss": 3.115,
+      "step": 1032
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014535326883611844,
+      "loss": 3.2058,
+      "step": 1033
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014529463500439756,
+      "loss": 3.1415,
+      "step": 1034
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014523600117267665,
+      "loss": 3.231,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014517736734095572,
+      "loss": 3.2668,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014511873350923484,
+      "loss": 3.2877,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014506009967751393,
+      "loss": 3.2669,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014500146584579303,
+      "loss": 3.2344,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014494283201407212,
+      "loss": 3.2069,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014488419818235122,
+      "loss": 3.2648,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014482556435063034,
+      "loss": 3.1754,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001447669305189094,
+      "loss": 3.3066,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001447082966871885,
+      "loss": 3.2235,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014464966285546762,
+      "loss": 3.3046,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014459102902374672,
+      "loss": 3.1704,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001445323951920258,
+      "loss": 3.1684,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001444737613603049,
+      "loss": 3.2783,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000144415127528584,
+      "loss": 3.2234,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001443564936968631,
+      "loss": 3.2467,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001442978598651422,
+      "loss": 3.1604,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014423922603342128,
+      "loss": 3.1556,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001441805922017004,
+      "loss": 3.324,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014412195836997947,
+      "loss": 3.2424,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001440633245382586,
+      "loss": 3.2606,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014400469070653769,
+      "loss": 3.2596,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 3.3927457332611084,
+      "eval_runtime": 2940.1937,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014394605687481678,
+      "loss": 3.1789,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014388742304309587,
+      "loss": 3.1009,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014382878921137497,
+      "loss": 3.2032,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014377015537965406,
+      "loss": 3.1781,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014371152154793316,
+      "loss": 3.1923,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014365288771621225,
+      "loss": 3.2216,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014359425388449135,
+      "loss": 3.2706,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014353562005277047,
+      "loss": 3.1661,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014347698622104953,
+      "loss": 3.2143,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014341835238932866,
+      "loss": 3.1501,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014335971855760775,
+      "loss": 3.1273,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014330108472588684,
+      "loss": 3.1188,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014324245089416594,
+      "loss": 3.2799,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014318381706244503,
+      "loss": 3.2518,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014312518323072413,
+      "loss": 3.2237,
+      "step": 1071
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014306654939900325,
+      "loss": 3.2532,
+      "step": 1072
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014300791556728232,
+      "loss": 3.2757,
+      "step": 1073
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014294928173556144,
+      "loss": 3.3044,
+      "step": 1074
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014289064790384053,
+      "loss": 3.3563,
+      "step": 1075
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014283201407211963,
+      "loss": 3.2188,
+      "step": 1076
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014277338024039872,
+      "loss": 3.2744,
+      "step": 1077
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014271474640867782,
+      "loss": 3.2477,
+      "step": 1078
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001426561125769569,
+      "loss": 3.1948,
+      "step": 1079
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000142597478745236,
+      "loss": 3.2597,
+      "step": 1080
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001425388449135151,
+      "loss": 3.2332,
+      "step": 1081
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014248021108179422,
+      "loss": 3.2089,
+      "step": 1082
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014242157725007331,
+      "loss": 3.0902,
+      "step": 1083
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014236294341835238,
+      "loss": 3.2496,
+      "step": 1084
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001423043095866315,
+      "loss": 3.2572,
+      "step": 1085
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001422456757549106,
+      "loss": 3.1796,
+      "step": 1086
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001421870419231897,
+      "loss": 3.1108,
+      "step": 1087
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014212840809146879,
+      "loss": 3.2689,
+      "step": 1088
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014206977425974788,
+      "loss": 3.2664,
+      "step": 1089
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014201114042802697,
+      "loss": 3.2866,
+      "step": 1090
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014195250659630607,
+      "loss": 3.1632,
+      "step": 1091
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014189387276458516,
+      "loss": 3.1654,
+      "step": 1092
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014183523893286428,
+      "loss": 3.2031,
+      "step": 1093
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014177660510114338,
+      "loss": 3.1857,
+      "step": 1094
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014171797126942245,
+      "loss": 3.2165,
+      "step": 1095
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014165933743770157,
+      "loss": 3.1198,
+      "step": 1096
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014160070360598066,
+      "loss": 3.1494,
+      "step": 1097
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014154206977425976,
+      "loss": 3.1959,
+      "step": 1098
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014148343594253885,
+      "loss": 3.2452,
+      "step": 1099
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014142480211081794,
+      "loss": 3.1721,
+      "step": 1100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014136616827909707,
+      "loss": 3.1778,
+      "step": 1101
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014130753444737613,
+      "loss": 3.2613,
+      "step": 1102
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014124890061565523,
+      "loss": 3.2279,
+      "step": 1103
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014119026678393435,
+      "loss": 3.2708,
+      "step": 1104
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014113163295221344,
+      "loss": 3.2222,
+      "step": 1105
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001410729991204925,
+      "loss": 3.1533,
+      "step": 1106
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014101436528877163,
+      "loss": 3.1679,
+      "step": 1107
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014095573145705073,
+      "loss": 3.211,
+      "step": 1108
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014089709762532982,
+      "loss": 3.2857,
+      "step": 1109
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014083846379360891,
+      "loss": 3.2805,
+      "step": 1110
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000140779829961888,
+      "loss": 3.243,
+      "step": 1111
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014072119613016713,
+      "loss": 3.3112,
+      "step": 1112
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001406625622984462,
+      "loss": 3.215,
+      "step": 1113
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001406039284667253,
+      "loss": 3.2921,
+      "step": 1114
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001405452946350044,
+      "loss": 3.3166,
+      "step": 1115
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001404866608032835,
+      "loss": 3.1663,
+      "step": 1116
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014042802697156257,
+      "loss": 3.1809,
+      "step": 1117
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001403693931398417,
+      "loss": 3.2643,
+      "step": 1118
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001403107593081208,
+      "loss": 3.1126,
+      "step": 1119
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014025212547639988,
+      "loss": 3.2337,
+      "step": 1120
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014019349164467898,
+      "loss": 3.2727,
+      "step": 1121
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014013485781295807,
+      "loss": 3.2654,
+      "step": 1122
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001400762239812372,
+      "loss": 3.2244,
+      "step": 1123
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014001759014951626,
+      "loss": 3.1932,
+      "step": 1124
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013995895631779536,
+      "loss": 3.1855,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013990032248607448,
+      "loss": 3.1859,
+      "step": 1126
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013984168865435357,
+      "loss": 3.2393,
+      "step": 1127
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013978305482263267,
+      "loss": 3.2284,
+      "step": 1128
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013972442099091176,
+      "loss": 3.2553,
+      "step": 1129
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013966578715919086,
+      "loss": 3.1857,
+      "step": 1130
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013960715332746998,
+      "loss": 3.2055,
+      "step": 1131
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013954851949574904,
+      "loss": 3.2172,
+      "step": 1132
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013948988566402814,
+      "loss": 3.1507,
+      "step": 1133
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013943125183230726,
+      "loss": 3.2008,
+      "step": 1134
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013937261800058635,
+      "loss": 3.109,
+      "step": 1135
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013931398416886545,
+      "loss": 3.1814,
+      "step": 1136
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013925535033714454,
+      "loss": 3.1604,
+      "step": 1137
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013919671650542364,
+      "loss": 3.1832,
+      "step": 1138
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013913808267370273,
+      "loss": 3.1034,
+      "step": 1139
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013907944884198183,
+      "loss": 3.2224,
+      "step": 1140
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00013902081501026092,
+      "loss": 3.119,
+      "step": 1141
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013896218117854004,
+      "loss": 3.2043,
+      "step": 1142
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001389035473468191,
+      "loss": 3.2507,
+      "step": 1143
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013884491351509823,
+      "loss": 3.2858,
+      "step": 1144
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013878627968337732,
+      "loss": 3.1763,
+      "step": 1145
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013872764585165642,
+      "loss": 3.2726,
+      "step": 1146
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001386690120199355,
+      "loss": 3.2172,
+      "step": 1147
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001386103781882146,
+      "loss": 3.1881,
+      "step": 1148
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001385517443564937,
+      "loss": 3.1157,
+      "step": 1149
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001384931105247728,
+      "loss": 3.2874,
+      "step": 1150
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001384344766930519,
+      "loss": 3.2671,
+      "step": 1151
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013837584286133098,
+      "loss": 3.1038,
+      "step": 1152
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001383172090296101,
+      "loss": 3.2201,
+      "step": 1153
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013825857519788917,
+      "loss": 3.2126,
+      "step": 1154
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001381999413661683,
+      "loss": 3.2184,
+      "step": 1155
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001381413075344474,
+      "loss": 3.1785,
+      "step": 1156
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013808267370272648,
+      "loss": 3.1887,
+      "step": 1157
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013802403987100558,
+      "loss": 3.2579,
+      "step": 1158
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013796540603928467,
+      "loss": 3.0964,
+      "step": 1159
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013790677220756377,
+      "loss": 3.1758,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013784813837584286,
+      "loss": 3.1086,
+      "step": 1161
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013778950454412195,
+      "loss": 3.2121,
+      "step": 1162
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013773087071240108,
+      "loss": 3.0181,
+      "step": 1163
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013767223688068017,
+      "loss": 3.0773,
+      "step": 1164
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013761360304895924,
+      "loss": 3.1632,
+      "step": 1165
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013755496921723836,
+      "loss": 3.1465,
+      "step": 1166
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013749633538551745,
+      "loss": 3.1723,
+      "step": 1167
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013743770155379655,
+      "loss": 3.2446,
+      "step": 1168
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013737906772207564,
+      "loss": 3.2361,
+      "step": 1169
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013732043389035474,
+      "loss": 3.2172,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013726180005863386,
+      "loss": 3.2092,
+      "step": 1171
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013720316622691292,
+      "loss": 3.2105,
+      "step": 1172
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013714453239519202,
+      "loss": 3.2252,
+      "step": 1173
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013708589856347114,
+      "loss": 3.2547,
+      "step": 1174
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00013702726473175024,
+      "loss": 3.1852,
+      "step": 1175
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001369686309000293,
+      "loss": 3.1305,
+      "step": 1176
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013690999706830842,
+      "loss": 3.0818,
+      "step": 1177
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013685136323658752,
+      "loss": 3.1579,
+      "step": 1178
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001367927294048666,
+      "loss": 3.2461,
+      "step": 1179
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001367340955731457,
+      "loss": 3.1656,
+      "step": 1180
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001366754617414248,
+      "loss": 3.1799,
+      "step": 1181
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013661682790970392,
+      "loss": 3.2685,
+      "step": 1182
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013655819407798302,
+      "loss": 3.1792,
+      "step": 1183
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013649956024626208,
+      "loss": 3.2519,
+      "step": 1184
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001364409264145412,
+      "loss": 3.1943,
+      "step": 1185
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001363822925828203,
+      "loss": 3.1631,
+      "step": 1186
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001363236587510994,
+      "loss": 3.1836,
+      "step": 1187
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001362650249193785,
+      "loss": 3.215,
+      "step": 1188
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013620639108765758,
+      "loss": 3.1792,
+      "step": 1189
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001361477572559367,
+      "loss": 3.1073,
+      "step": 1190
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013608912342421577,
+      "loss": 3.1156,
+      "step": 1191
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013603048959249487,
+      "loss": 3.2696,
+      "step": 1192
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000135971855760774,
+      "loss": 3.1732,
+      "step": 1193
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013591322192905308,
+      "loss": 3.2555,
+      "step": 1194
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013585458809733215,
+      "loss": 3.2004,
+      "step": 1195
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013579595426561127,
+      "loss": 3.1993,
+      "step": 1196
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013573732043389036,
+      "loss": 3.2065,
+      "step": 1197
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013567868660216946,
+      "loss": 3.1537,
+      "step": 1198
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013562005277044855,
+      "loss": 3.2104,
+      "step": 1199
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013556141893872765,
+      "loss": 3.2029,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013550278510700677,
+      "loss": 3.2434,
+      "step": 1201
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013544415127528584,
+      "loss": 3.2309,
+      "step": 1202
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013538551744356493,
+      "loss": 3.2562,
+      "step": 1203
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013532688361184405,
+      "loss": 3.1681,
+      "step": 1204
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013526824978012315,
+      "loss": 3.1001,
+      "step": 1205
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001352096159484022,
+      "loss": 3.1597,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013515098211668133,
+      "loss": 3.1631,
+      "step": 1207
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013509234828496043,
+      "loss": 3.2069,
+      "step": 1208
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013503371445323952,
+      "loss": 3.0383,
+      "step": 1209
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00013497508062151862,
+      "loss": 3.1166,
+      "step": 1210
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001349164467897977,
+      "loss": 3.1958,
+      "step": 1211
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013485781295807683,
+      "loss": 3.1733,
+      "step": 1212
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001347991791263559,
+      "loss": 3.2266,
+      "step": 1213
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000134740545294635,
+      "loss": 3.0868,
+      "step": 1214
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013468191146291412,
+      "loss": 3.261,
+      "step": 1215
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001346232776311932,
+      "loss": 3.2114,
+      "step": 1216
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001345646437994723,
+      "loss": 3.1535,
+      "step": 1217
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001345060099677514,
+      "loss": 3.1547,
+      "step": 1218
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001344473761360305,
+      "loss": 3.0801,
+      "step": 1219
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001343887423043096,
+      "loss": 3.174,
+      "step": 1220
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013433010847258868,
+      "loss": 3.1559,
+      "step": 1221
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013427147464086778,
+      "loss": 3.1063,
+      "step": 1222
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001342128408091469,
+      "loss": 3.1266,
+      "step": 1223
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013415420697742597,
+      "loss": 3.2614,
+      "step": 1224
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001340955731457051,
+      "loss": 3.2669,
+      "step": 1225
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013403693931398418,
+      "loss": 3.1827,
+      "step": 1226
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013397830548226328,
+      "loss": 3.1157,
+      "step": 1227
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013391967165054237,
+      "loss": 3.139,
+      "step": 1228
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013386103781882146,
+      "loss": 3.1076,
+      "step": 1229
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013380240398710056,
+      "loss": 3.1338,
+      "step": 1230
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013374377015537968,
+      "loss": 3.1602,
+      "step": 1231
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013368513632365875,
+      "loss": 3.1357,
+      "step": 1232
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 3.382495641708374,
+      "eval_runtime": 2940.5137,
+      "eval_samples_per_second": 6.967,
+      "eval_steps_per_second": 2.322,
+      "step": 1232
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013362650249193784,
+      "loss": 3.2242,
+      "step": 1233
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013356786866021696,
+      "loss": 3.0442,
+      "step": 1234
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013350923482849606,
+      "loss": 3.1842,
+      "step": 1235
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013345060099677515,
+      "loss": 3.1555,
+      "step": 1236
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013339196716505425,
+      "loss": 3.2172,
+      "step": 1237
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 3.1764,
+      "step": 1238
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013327469950161243,
+      "loss": 3.1507,
+      "step": 1239
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013321606566989153,
+      "loss": 3.1659,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013315743183817062,
+      "loss": 3.1516,
+      "step": 1241
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013309879800644974,
+      "loss": 3.1561,
+      "step": 1242
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001330401641747288,
+      "loss": 3.2656,
+      "step": 1243
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013298153034300793,
+      "loss": 3.1399,
+      "step": 1244
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013292289651128703,
+      "loss": 3.1239,
+      "step": 1245
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00013286426267956612,
+      "loss": 3.1478,
+      "step": 1246
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013280562884784522,
+      "loss": 3.2921,
+      "step": 1247
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001327469950161243,
+      "loss": 3.2601,
+      "step": 1248
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001326883611844034,
+      "loss": 3.2197,
+      "step": 1249
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001326297273526825,
+      "loss": 3.1752,
+      "step": 1250
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001325710935209616,
+      "loss": 3.2316,
+      "step": 1251
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013251245968924071,
+      "loss": 3.2014,
+      "step": 1252
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001324538258575198,
+      "loss": 3.2569,
+      "step": 1253
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013239519202579888,
+      "loss": 3.0543,
+      "step": 1254
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000132336558194078,
+      "loss": 3.2453,
+      "step": 1255
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001322779243623571,
+      "loss": 3.0933,
+      "step": 1256
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013221929053063619,
+      "loss": 3.3076,
+      "step": 1257
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013216065669891528,
+      "loss": 3.1836,
+      "step": 1258
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013210202286719437,
+      "loss": 3.1859,
+      "step": 1259
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013204338903547347,
+      "loss": 3.0626,
+      "step": 1260
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013198475520375256,
+      "loss": 3.2315,
+      "step": 1261
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013192612137203166,
+      "loss": 3.1726,
+      "step": 1262
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013186748754031078,
+      "loss": 3.1264,
+      "step": 1263
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013180885370858987,
+      "loss": 3.111,
+      "step": 1264
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013175021987686894,
+      "loss": 3.2399,
+      "step": 1265
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013169158604514806,
+      "loss": 3.2369,
+      "step": 1266
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013163295221342716,
+      "loss": 3.186,
+      "step": 1267
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013157431838170625,
+      "loss": 3.2087,
+      "step": 1268
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013151568454998534,
+      "loss": 3.1764,
+      "step": 1269
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013145705071826444,
+      "loss": 3.2164,
+      "step": 1270
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013139841688654356,
+      "loss": 3.1762,
+      "step": 1271
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013133978305482263,
+      "loss": 3.1607,
+      "step": 1272
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013128114922310172,
+      "loss": 3.1759,
+      "step": 1273
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013122251539138084,
+      "loss": 3.2293,
+      "step": 1274
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013116388155965994,
+      "loss": 3.1406,
+      "step": 1275
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000131105247727939,
+      "loss": 3.2147,
+      "step": 1276
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013104661389621813,
+      "loss": 3.1534,
+      "step": 1277
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013098798006449722,
+      "loss": 3.2401,
+      "step": 1278
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00013092934623277632,
+      "loss": 3.1721,
+      "step": 1279
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001308707124010554,
+      "loss": 3.1573,
+      "step": 1280
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001308120785693345,
+      "loss": 3.1665,
+      "step": 1281
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013075344473761363,
+      "loss": 3.0761,
+      "step": 1282
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013069481090589272,
+      "loss": 3.0883,
+      "step": 1283
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001306361770741718,
+      "loss": 3.1714,
+      "step": 1284
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001305775432424509,
+      "loss": 3.1752,
+      "step": 1285
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013051890941073,
+      "loss": 3.2126,
+      "step": 1286
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001304602755790091,
+      "loss": 3.0492,
+      "step": 1287
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001304016417472882,
+      "loss": 3.0933,
+      "step": 1288
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013034300791556729,
+      "loss": 3.2208,
+      "step": 1289
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001302843740838464,
+      "loss": 3.1825,
+      "step": 1290
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013022574025212547,
+      "loss": 3.0963,
+      "step": 1291
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013016710642040457,
+      "loss": 3.1437,
+      "step": 1292
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001301084725886837,
+      "loss": 3.1687,
+      "step": 1293
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00013004983875696278,
+      "loss": 3.1651,
+      "step": 1294
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012999120492524185,
+      "loss": 3.2251,
+      "step": 1295
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012993257109352097,
+      "loss": 2.9805,
+      "step": 1296
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012987393726180007,
+      "loss": 3.1236,
+      "step": 1297
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012981530343007916,
+      "loss": 3.2004,
+      "step": 1298
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012975666959835826,
+      "loss": 3.2191,
+      "step": 1299
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012969803576663735,
+      "loss": 3.1347,
+      "step": 1300
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012963940193491647,
+      "loss": 3.1517,
+      "step": 1301
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012958076810319554,
+      "loss": 3.1624,
+      "step": 1302
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012952213427147463,
+      "loss": 3.2144,
+      "step": 1303
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012946350043975375,
+      "loss": 3.2031,
+      "step": 1304
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012940486660803285,
+      "loss": 3.1027,
+      "step": 1305
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012934623277631194,
+      "loss": 3.1753,
+      "step": 1306
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012928759894459104,
+      "loss": 3.1492,
+      "step": 1307
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012922896511287013,
+      "loss": 3.1104,
+      "step": 1308
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012917033128114923,
+      "loss": 3.0851,
+      "step": 1309
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012911169744942832,
+      "loss": 3.2225,
+      "step": 1310
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012905306361770741,
+      "loss": 3.1292,
+      "step": 1311
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012899442978598654,
+      "loss": 3.1407,
+      "step": 1312
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001289357959542656,
+      "loss": 3.1578,
+      "step": 1313
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001288771621225447,
+      "loss": 3.1143,
+      "step": 1314
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00012881852829082382,
+      "loss": 3.2208,
+      "step": 1315
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001287598944591029,
+      "loss": 3.1903,
+      "step": 1316
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000128701260627382,
+      "loss": 3.1242,
+      "step": 1317
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001286426267956611,
+      "loss": 3.1529,
+      "step": 1318
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001285839929639402,
+      "loss": 3.0988,
+      "step": 1319
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001285253591322193,
+      "loss": 3.2027,
+      "step": 1320
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012846672530049839,
+      "loss": 3.1763,
+      "step": 1321
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012840809146877748,
+      "loss": 3.1793,
+      "step": 1322
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001283494576370566,
+      "loss": 3.2243,
+      "step": 1323
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012829082380533567,
+      "loss": 3.1714,
+      "step": 1324
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001282321899736148,
+      "loss": 3.2676,
+      "step": 1325
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012817355614189388,
+      "loss": 3.2629,
+      "step": 1326
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012811492231017298,
+      "loss": 3.1628,
+      "step": 1327
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012805628847845207,
+      "loss": 3.1174,
+      "step": 1328
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012799765464673117,
+      "loss": 3.2498,
+      "step": 1329
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012793902081501026,
+      "loss": 3.2414,
+      "step": 1330
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012788038698328936,
+      "loss": 3.1061,
+      "step": 1331
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012782175315156845,
+      "loss": 3.1805,
+      "step": 1332
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012776311931984757,
+      "loss": 3.1126,
+      "step": 1333
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012770448548812667,
+      "loss": 3.2429,
+      "step": 1334
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012764585165640573,
+      "loss": 3.1792,
+      "step": 1335
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012758721782468485,
+      "loss": 3.088,
+      "step": 1336
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012752858399296395,
+      "loss": 3.1542,
+      "step": 1337
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012746995016124304,
+      "loss": 2.9935,
+      "step": 1338
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012741131632952214,
+      "loss": 3.2113,
+      "step": 1339
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012735268249780123,
+      "loss": 3.2206,
+      "step": 1340
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012729404866608035,
+      "loss": 3.2388,
+      "step": 1341
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012723541483435945,
+      "loss": 3.2296,
+      "step": 1342
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012717678100263851,
+      "loss": 3.2265,
+      "step": 1343
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012711814717091764,
+      "loss": 3.2736,
+      "step": 1344
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012705951333919673,
+      "loss": 3.2912,
+      "step": 1345
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012700087950747582,
+      "loss": 3.046,
+      "step": 1346
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012694224567575492,
+      "loss": 3.1895,
+      "step": 1347
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000126883611844034,
+      "loss": 3.2153,
+      "step": 1348
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001268249780123131,
+      "loss": 3.1984,
+      "step": 1349
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001267663441805922,
+      "loss": 3.2679,
+      "step": 1350
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001267077103488713,
+      "loss": 3.1259,
+      "step": 1351
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00012664907651715042,
+      "loss": 3.2098,
+      "step": 1352
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001265904426854295,
+      "loss": 3.285,
+      "step": 1353
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012653180885370858,
+      "loss": 3.213,
+      "step": 1354
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001264731750219877,
+      "loss": 3.188,
+      "step": 1355
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001264145411902668,
+      "loss": 3.2731,
+      "step": 1356
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001263559073585459,
+      "loss": 3.1551,
+      "step": 1357
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012629727352682498,
+      "loss": 3.2515,
+      "step": 1358
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012623863969510408,
+      "loss": 3.1407,
+      "step": 1359
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001261800058633832,
+      "loss": 3.1234,
+      "step": 1360
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012612137203166227,
+      "loss": 3.1077,
+      "step": 1361
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012606273819994136,
+      "loss": 3.1269,
+      "step": 1362
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012600410436822048,
+      "loss": 3.1977,
+      "step": 1363
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012594547053649958,
+      "loss": 3.0656,
+      "step": 1364
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012588683670477864,
+      "loss": 3.1758,
+      "step": 1365
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012582820287305776,
+      "loss": 3.0228,
+      "step": 1366
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012576956904133686,
+      "loss": 3.2206,
+      "step": 1367
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012571093520961595,
+      "loss": 3.1765,
+      "step": 1368
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012565230137789505,
+      "loss": 3.2076,
+      "step": 1369
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012559366754617414,
+      "loss": 3.2441,
+      "step": 1370
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012553503371445326,
+      "loss": 3.1674,
+      "step": 1371
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012547639988273233,
+      "loss": 3.1856,
+      "step": 1372
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012541776605101143,
+      "loss": 3.1736,
+      "step": 1373
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012535913221929055,
+      "loss": 3.2034,
+      "step": 1374
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012530049838756964,
+      "loss": 3.084,
+      "step": 1375
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001252418645558487,
+      "loss": 3.0943,
+      "step": 1376
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012518323072412783,
+      "loss": 3.2176,
+      "step": 1377
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012512459689240692,
+      "loss": 3.0948,
+      "step": 1378
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012506596306068602,
+      "loss": 3.0982,
+      "step": 1379
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001250073292289651,
+      "loss": 3.1181,
+      "step": 1380
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001249486953972442,
+      "loss": 3.1607,
+      "step": 1381
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00012489006156552333,
+      "loss": 3.1799,
+      "step": 1382
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001248314277338024,
+      "loss": 3.0335,
+      "step": 1383
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001247727939020815,
+      "loss": 3.2615,
+      "step": 1384
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001247141600703606,
+      "loss": 3.1527,
+      "step": 1385
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001246555262386397,
+      "loss": 3.2526,
+      "step": 1386
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001245968924069188,
+      "loss": 3.1236,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001245382585751979,
+      "loss": 3.1207,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000124479624743477,
+      "loss": 3.0895,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001244209909117561,
+      "loss": 3.0801,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012436235708003518,
+      "loss": 3.1893,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012430372324831427,
+      "loss": 3.1752,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001242450894165934,
+      "loss": 3.1808,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001241864555848725,
+      "loss": 3.1455,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012412782175315158,
+      "loss": 3.081,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012406918792143068,
+      "loss": 3.3061,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012401055408970977,
+      "loss": 3.1577,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012395192025798886,
+      "loss": 3.1186,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012389328642626796,
+      "loss": 3.1425,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012383465259454705,
+      "loss": 3.0414,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012377601876282617,
+      "loss": 3.0458,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012371738493110524,
+      "loss": 3.1533,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012365875109938434,
+      "loss": 3.1256,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012360011726766346,
+      "loss": 3.2095,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012354148343594255,
+      "loss": 3.0902,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012348284960422165,
+      "loss": 3.2279,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012342421577250074,
+      "loss": 3.2378,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012336558194077983,
+      "loss": 3.1117,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 3.358116388320923,
+      "eval_runtime": 2940.1995,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012330694810905893,
+      "loss": 3.0724,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012324831427733802,
+      "loss": 3.0971,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012318968044561712,
+      "loss": 3.1188,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012313104661389624,
+      "loss": 3.1674,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001230724127821753,
+      "loss": 3.1835,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012301377895045443,
+      "loss": 3.1091,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012295514511873352,
+      "loss": 3.1179,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00012289651128701262,
+      "loss": 3.0855,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001228378774552917,
+      "loss": 3.1328,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001227792436235708,
+      "loss": 3.1542,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001227206097918499,
+      "loss": 3.0795,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000122661975960129,
+      "loss": 3.2493,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001226033421284081,
+      "loss": 2.9858,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001225447082966872,
+      "loss": 3.1612,
+      "step": 1422
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001224860744649663,
+      "loss": 3.2002,
+      "step": 1423
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012242744063324537,
+      "loss": 3.092,
+      "step": 1424
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001223688068015245,
+      "loss": 3.1716,
+      "step": 1425
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001223101729698036,
+      "loss": 3.2572,
+      "step": 1426
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012225153913808268,
+      "loss": 3.1645,
+      "step": 1427
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012219290530636178,
+      "loss": 3.1283,
+      "step": 1428
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012213427147464087,
+      "loss": 3.1762,
+      "step": 1429
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012207563764291996,
+      "loss": 3.2059,
+      "step": 1430
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012201700381119907,
+      "loss": 3.1986,
+      "step": 1431
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012195836997947817,
+      "loss": 3.1574,
+      "step": 1432
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012189973614775727,
+      "loss": 3.1534,
+      "step": 1433
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012184110231603635,
+      "loss": 2.9625,
+      "step": 1434
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012178246848431545,
+      "loss": 3.153,
+      "step": 1435
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012172383465259456,
+      "loss": 3.2113,
+      "step": 1436
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012166520082087365,
+      "loss": 3.1937,
+      "step": 1437
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012160656698915273,
+      "loss": 3.2227,
+      "step": 1438
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012154793315743185,
+      "loss": 3.0936,
+      "step": 1439
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012148929932571093,
+      "loss": 3.1431,
+      "step": 1440
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012143066549399004,
+      "loss": 3.1674,
+      "step": 1441
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012137203166226914,
+      "loss": 3.1355,
+      "step": 1442
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012131339783054823,
+      "loss": 3.2486,
+      "step": 1443
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012125476399882734,
+      "loss": 3.2519,
+      "step": 1444
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012119613016710642,
+      "loss": 3.1181,
+      "step": 1445
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012113749633538551,
+      "loss": 3.1866,
+      "step": 1446
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012107886250366462,
+      "loss": 3.0154,
+      "step": 1447
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012102022867194372,
+      "loss": 3.1062,
+      "step": 1448
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012096159484022282,
+      "loss": 3.2628,
+      "step": 1449
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012090296100850192,
+      "loss": 3.207,
+      "step": 1450
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000120844327176781,
+      "loss": 3.1502,
+      "step": 1451
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012078569334506012,
+      "loss": 3.201,
+      "step": 1452
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001207270595133392,
+      "loss": 3.0811,
+      "step": 1453
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001206684256816183,
+      "loss": 3.1412,
+      "step": 1454
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001206097918498974,
+      "loss": 3.215,
+      "step": 1455
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001205511580181765,
+      "loss": 3.1313,
+      "step": 1456
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00012049252418645558,
+      "loss": 3.0466,
+      "step": 1457
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012043389035473469,
+      "loss": 2.945,
+      "step": 1458
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012037525652301378,
+      "loss": 3.191,
+      "step": 1459
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012031662269129289,
+      "loss": 3.1762,
+      "step": 1460
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012025798885957198,
+      "loss": 3.1753,
+      "step": 1461
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012019935502785106,
+      "loss": 3.118,
+      "step": 1462
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012014072119613018,
+      "loss": 3.1987,
+      "step": 1463
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012008208736440927,
+      "loss": 3.1258,
+      "step": 1464
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012002345353268836,
+      "loss": 3.0951,
+      "step": 1465
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011996481970096747,
+      "loss": 3.2152,
+      "step": 1466
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011990618586924656,
+      "loss": 3.0928,
+      "step": 1467
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011984755203752567,
+      "loss": 3.0311,
+      "step": 1468
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011978891820580475,
+      "loss": 3.0732,
+      "step": 1469
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011973028437408385,
+      "loss": 3.1958,
+      "step": 1470
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011967165054236295,
+      "loss": 3.1185,
+      "step": 1471
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011961301671064205,
+      "loss": 3.1083,
+      "step": 1472
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011955438287892113,
+      "loss": 3.1742,
+      "step": 1473
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011949574904720025,
+      "loss": 3.1251,
+      "step": 1474
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011943711521547933,
+      "loss": 3.2046,
+      "step": 1475
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011937848138375844,
+      "loss": 3.1094,
+      "step": 1476
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011931984755203753,
+      "loss": 3.0576,
+      "step": 1477
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011926121372031663,
+      "loss": 3.1764,
+      "step": 1478
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011920257988859573,
+      "loss": 3.1838,
+      "step": 1479
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011914394605687483,
+      "loss": 3.1026,
+      "step": 1480
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011908531222515391,
+      "loss": 3.2143,
+      "step": 1481
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011902667839343302,
+      "loss": 3.1547,
+      "step": 1482
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011896804456171211,
+      "loss": 3.1739,
+      "step": 1483
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001189094107299912,
+      "loss": 3.1513,
+      "step": 1484
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011885077689827031,
+      "loss": 3.1518,
+      "step": 1485
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001187921430665494,
+      "loss": 3.1859,
+      "step": 1486
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011873350923482852,
+      "loss": 3.1476,
+      "step": 1487
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001186748754031076,
+      "loss": 3.0416,
+      "step": 1488
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011861624157138669,
+      "loss": 3.1129,
+      "step": 1489
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001185576077396658,
+      "loss": 3.1618,
+      "step": 1490
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001184989739079449,
+      "loss": 3.2003,
+      "step": 1491
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00011844034007622397,
+      "loss": 3.0911,
+      "step": 1492
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011838170624450308,
+      "loss": 3.2046,
+      "step": 1493
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011832307241278218,
+      "loss": 3.1409,
+      "step": 1494
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011826443858106128,
+      "loss": 3.0772,
+      "step": 1495
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011820580474934038,
+      "loss": 3.1814,
+      "step": 1496
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011814717091761946,
+      "loss": 3.1657,
+      "step": 1497
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011808853708589858,
+      "loss": 3.1691,
+      "step": 1498
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011802990325417766,
+      "loss": 3.2188,
+      "step": 1499
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011797126942245676,
+      "loss": 3.0851,
+      "step": 1500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011791263559073586,
+      "loss": 3.206,
+      "step": 1501
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011785400175901496,
+      "loss": 3.0703,
+      "step": 1502
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011779536792729407,
+      "loss": 3.1974,
+      "step": 1503
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011773673409557315,
+      "loss": 3.109,
+      "step": 1504
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011767810026385224,
+      "loss": 3.1281,
+      "step": 1505
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011761946643213135,
+      "loss": 3.2016,
+      "step": 1506
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011756083260041044,
+      "loss": 3.1947,
+      "step": 1507
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011750219876868954,
+      "loss": 3.1425,
+      "step": 1508
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011744356493696865,
+      "loss": 3.2012,
+      "step": 1509
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011738493110524773,
+      "loss": 3.1649,
+      "step": 1510
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011732629727352682,
+      "loss": 3.113,
+      "step": 1511
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011726766344180593,
+      "loss": 3.1457,
+      "step": 1512
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011720902961008502,
+      "loss": 3.1941,
+      "step": 1513
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011715039577836413,
+      "loss": 3.1677,
+      "step": 1514
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011709176194664322,
+      "loss": 3.1692,
+      "step": 1515
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001170331281149223,
+      "loss": 3.2623,
+      "step": 1516
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011697449428320141,
+      "loss": 3.0789,
+      "step": 1517
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011691586045148051,
+      "loss": 3.0899,
+      "step": 1518
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001168572266197596,
+      "loss": 3.2451,
+      "step": 1519
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011679859278803871,
+      "loss": 3.1909,
+      "step": 1520
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011673995895631779,
+      "loss": 3.1379,
+      "step": 1521
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011668132512459691,
+      "loss": 3.1656,
+      "step": 1522
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011662269129287599,
+      "loss": 3.0479,
+      "step": 1523
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011656405746115509,
+      "loss": 3.2149,
+      "step": 1524
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001165054236294342,
+      "loss": 3.1593,
+      "step": 1525
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011644678979771329,
+      "loss": 3.0278,
+      "step": 1526
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00011638815596599237,
+      "loss": 3.1505,
+      "step": 1527
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011632952213427148,
+      "loss": 3.0834,
+      "step": 1528
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011627088830255057,
+      "loss": 3.1135,
+      "step": 1529
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011621225447082968,
+      "loss": 3.2249,
+      "step": 1530
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011615362063910877,
+      "loss": 3.1105,
+      "step": 1531
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011609498680738786,
+      "loss": 3.1778,
+      "step": 1532
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011603635297566698,
+      "loss": 3.0562,
+      "step": 1533
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011597771914394606,
+      "loss": 3.1572,
+      "step": 1534
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011591908531222515,
+      "loss": 3.2102,
+      "step": 1535
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011586045148050426,
+      "loss": 3.0796,
+      "step": 1536
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011580181764878335,
+      "loss": 3.2072,
+      "step": 1537
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011574318381706246,
+      "loss": 3.1993,
+      "step": 1538
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011568454998534156,
+      "loss": 3.1107,
+      "step": 1539
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011562591615362064,
+      "loss": 3.2597,
+      "step": 1540
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011556728232189974,
+      "loss": 3.0999,
+      "step": 1541
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011550864849017884,
+      "loss": 3.2182,
+      "step": 1542
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011545001465845793,
+      "loss": 3.1421,
+      "step": 1543
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011539138082673704,
+      "loss": 3.0629,
+      "step": 1544
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011533274699501612,
+      "loss": 3.1179,
+      "step": 1545
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011527411316329522,
+      "loss": 3.1367,
+      "step": 1546
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011521547933157432,
+      "loss": 3.1791,
+      "step": 1547
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011515684549985342,
+      "loss": 3.1518,
+      "step": 1548
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011509821166813253,
+      "loss": 3.1704,
+      "step": 1549
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011503957783641162,
+      "loss": 3.1042,
+      "step": 1550
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001149809440046907,
+      "loss": 3.2331,
+      "step": 1551
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011492231017296981,
+      "loss": 3.1015,
+      "step": 1552
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001148636763412489,
+      "loss": 3.0461,
+      "step": 1553
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000114805042509528,
+      "loss": 3.2285,
+      "step": 1554
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001147464086778071,
+      "loss": 3.1485,
+      "step": 1555
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011468777484608619,
+      "loss": 3.1799,
+      "step": 1556
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011462914101436531,
+      "loss": 3.1473,
+      "step": 1557
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011457050718264439,
+      "loss": 3.1539,
+      "step": 1558
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011451187335092348,
+      "loss": 3.1422,
+      "step": 1559
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011445323951920259,
+      "loss": 3.1082,
+      "step": 1560
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011439460568748169,
+      "loss": 3.1161,
+      "step": 1561
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00011433597185576077,
+      "loss": 3.2615,
+      "step": 1562
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011427733802403989,
+      "loss": 3.1079,
+      "step": 1563
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011421870419231897,
+      "loss": 3.1397,
+      "step": 1564
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011416007036059808,
+      "loss": 3.1363,
+      "step": 1565
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011410143652887717,
+      "loss": 3.1261,
+      "step": 1566
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011404280269715626,
+      "loss": 3.0705,
+      "step": 1567
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011398416886543537,
+      "loss": 3.1307,
+      "step": 1568
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011392553503371445,
+      "loss": 3.0949,
+      "step": 1569
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011386690120199355,
+      "loss": 3.2387,
+      "step": 1570
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011380826737027266,
+      "loss": 3.2034,
+      "step": 1571
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011374963353855175,
+      "loss": 3.1882,
+      "step": 1572
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011369099970683083,
+      "loss": 3.0792,
+      "step": 1573
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011363236587510995,
+      "loss": 3.2237,
+      "step": 1574
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011357373204338903,
+      "loss": 3.2187,
+      "step": 1575
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011351509821166814,
+      "loss": 3.11,
+      "step": 1576
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011345646437994724,
+      "loss": 3.1262,
+      "step": 1577
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011339783054822633,
+      "loss": 3.1073,
+      "step": 1578
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011333919671650544,
+      "loss": 3.0989,
+      "step": 1579
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011328056288478452,
+      "loss": 3.1451,
+      "step": 1580
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011322192905306361,
+      "loss": 3.1339,
+      "step": 1581
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011316329522134272,
+      "loss": 3.2096,
+      "step": 1582
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011310466138962181,
+      "loss": 3.2126,
+      "step": 1583
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011304602755790092,
+      "loss": 3.1355,
+      "step": 1584
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 3.33901309967041,
+      "eval_runtime": 2939.0345,
+      "eval_samples_per_second": 6.971,
+      "eval_steps_per_second": 2.324,
+      "step": 1584
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011298739372618002,
+      "loss": 3.2127,
+      "step": 1585
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001129287598944591,
+      "loss": 3.0874,
+      "step": 1586
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011287012606273822,
+      "loss": 3.1317,
+      "step": 1587
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001128114922310173,
+      "loss": 3.1545,
+      "step": 1588
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001127528583992964,
+      "loss": 3.0883,
+      "step": 1589
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001126942245675755,
+      "loss": 3.1379,
+      "step": 1590
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001126355907358546,
+      "loss": 3.1477,
+      "step": 1591
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001125769569041337,
+      "loss": 3.0437,
+      "step": 1592
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011251832307241279,
+      "loss": 3.092,
+      "step": 1593
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011245968924069188,
+      "loss": 3.1902,
+      "step": 1594
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011240105540897099,
+      "loss": 3.1561,
+      "step": 1595
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011234242157725008,
+      "loss": 3.1247,
+      "step": 1596
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011228378774552916,
+      "loss": 3.1098,
+      "step": 1597
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011222515391380828,
+      "loss": 3.178,
+      "step": 1598
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011216652008208736,
+      "loss": 3.0661,
+      "step": 1599
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011210788625036646,
+      "loss": 3.0487,
+      "step": 1600
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011204925241864557,
+      "loss": 3.1827,
+      "step": 1601
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011199061858692466,
+      "loss": 3.1593,
+      "step": 1602
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011193198475520377,
+      "loss": 3.0422,
+      "step": 1603
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011187335092348285,
+      "loss": 3.1839,
+      "step": 1604
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011181471709176194,
+      "loss": 3.0899,
+      "step": 1605
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011175608326004105,
+      "loss": 3.2206,
+      "step": 1606
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011169744942832015,
+      "loss": 3.0958,
+      "step": 1607
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011163881559659923,
+      "loss": 3.1919,
+      "step": 1608
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011158018176487835,
+      "loss": 3.2004,
+      "step": 1609
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011152154793315743,
+      "loss": 3.1084,
+      "step": 1610
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011146291410143655,
+      "loss": 3.0346,
+      "step": 1611
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011140428026971563,
+      "loss": 3.0911,
+      "step": 1612
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011134564643799473,
+      "loss": 2.992,
+      "step": 1613
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011128701260627383,
+      "loss": 3.0912,
+      "step": 1614
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011122837877455293,
+      "loss": 3.1363,
+      "step": 1615
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011116974494283201,
+      "loss": 3.0195,
+      "step": 1616
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011111111111111112,
+      "loss": 3.0398,
+      "step": 1617
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011105247727939021,
+      "loss": 2.9856,
+      "step": 1618
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011099384344766932,
+      "loss": 3.1327,
+      "step": 1619
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011093520961594841,
+      "loss": 3.1394,
+      "step": 1620
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001108765757842275,
+      "loss": 3.1844,
+      "step": 1621
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011081794195250662,
+      "loss": 3.0771,
+      "step": 1622
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001107593081207857,
+      "loss": 3.1131,
+      "step": 1623
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011070067428906479,
+      "loss": 3.2034,
+      "step": 1624
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001106420404573439,
+      "loss": 3.0252,
+      "step": 1625
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011058340662562299,
+      "loss": 3.0887,
+      "step": 1626
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011052477279390207,
+      "loss": 3.0802,
+      "step": 1627
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011046613896218118,
+      "loss": 3.1181,
+      "step": 1628
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011040750513046028,
+      "loss": 3.1592,
+      "step": 1629
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011034887129873938,
+      "loss": 3.1655,
+      "step": 1630
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011029023746701848,
+      "loss": 3.143,
+      "step": 1631
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011023160363529756,
+      "loss": 3.1253,
+      "step": 1632
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011017296980357668,
+      "loss": 3.1357,
+      "step": 1633
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011011433597185576,
+      "loss": 3.0326,
+      "step": 1634
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011005570214013485,
+      "loss": 3.1027,
+      "step": 1635
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010999706830841396,
+      "loss": 3.0767,
+      "step": 1636
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010993843447669306,
+      "loss": 3.0888,
+      "step": 1637
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010987980064497216,
+      "loss": 3.0943,
+      "step": 1638
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010982116681325126,
+      "loss": 3.2003,
+      "step": 1639
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010976253298153034,
+      "loss": 3.1548,
+      "step": 1640
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010970389914980945,
+      "loss": 2.9824,
+      "step": 1641
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010964526531808854,
+      "loss": 3.144,
+      "step": 1642
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010958663148636764,
+      "loss": 3.0977,
+      "step": 1643
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010952799765464674,
+      "loss": 3.1912,
+      "step": 1644
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010946936382292583,
+      "loss": 3.0353,
+      "step": 1645
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010941072999120495,
+      "loss": 3.0586,
+      "step": 1646
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010935209615948403,
+      "loss": 3.0516,
+      "step": 1647
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010929346232776312,
+      "loss": 3.117,
+      "step": 1648
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010923482849604223,
+      "loss": 3.0475,
+      "step": 1649
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010917619466432132,
+      "loss": 3.0815,
+      "step": 1650
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001091175608326004,
+      "loss": 3.1132,
+      "step": 1651
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010905892700087951,
+      "loss": 3.0678,
+      "step": 1652
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001090002931691586,
+      "loss": 3.1199,
+      "step": 1653
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001089416593374377,
+      "loss": 3.1443,
+      "step": 1654
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010888302550571681,
+      "loss": 3.13,
+      "step": 1655
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010882439167399589,
+      "loss": 3.1086,
+      "step": 1656
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010876575784227501,
+      "loss": 3.1537,
+      "step": 1657
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010870712401055409,
+      "loss": 3.1161,
+      "step": 1658
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010864849017883319,
+      "loss": 3.0585,
+      "step": 1659
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001085898563471123,
+      "loss": 3.1445,
+      "step": 1660
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010853122251539139,
+      "loss": 3.0764,
+      "step": 1661
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010847258868367047,
+      "loss": 2.9982,
+      "step": 1662
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010841395485194959,
+      "loss": 3.1593,
+      "step": 1663
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010835532102022867,
+      "loss": 3.0856,
+      "step": 1664
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010829668718850778,
+      "loss": 3.1578,
+      "step": 1665
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010823805335678687,
+      "loss": 3.1565,
+      "step": 1666
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010817941952506597,
+      "loss": 2.9098,
+      "step": 1667
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00010812078569334508,
+      "loss": 3.1399,
+      "step": 1668
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010806215186162416,
+      "loss": 3.0203,
+      "step": 1669
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010800351802990325,
+      "loss": 3.1578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010794488419818236,
+      "loss": 3.1458,
+      "step": 1671
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010788625036646145,
+      "loss": 3.2191,
+      "step": 1672
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010782761653474056,
+      "loss": 3.0857,
+      "step": 1673
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010776898270301966,
+      "loss": 3.096,
+      "step": 1674
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010771034887129874,
+      "loss": 3.1596,
+      "step": 1675
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010765171503957784,
+      "loss": 3.0256,
+      "step": 1676
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010759308120785694,
+      "loss": 3.0621,
+      "step": 1677
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010753444737613603,
+      "loss": 3.117,
+      "step": 1678
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010747581354441514,
+      "loss": 3.1636,
+      "step": 1679
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010741717971269422,
+      "loss": 3.1441,
+      "step": 1680
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010735854588097332,
+      "loss": 3.1357,
+      "step": 1681
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010729991204925242,
+      "loss": 3.0222,
+      "step": 1682
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010724127821753152,
+      "loss": 3.1049,
+      "step": 1683
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010718264438581063,
+      "loss": 3.2128,
+      "step": 1684
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010712401055408972,
+      "loss": 3.1145,
+      "step": 1685
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001070653767223688,
+      "loss": 3.1629,
+      "step": 1686
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010700674289064791,
+      "loss": 3.0896,
+      "step": 1687
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.000106948109058927,
+      "loss": 2.9751,
+      "step": 1688
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001068894752272061,
+      "loss": 3.0981,
+      "step": 1689
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001068308413954852,
+      "loss": 3.055,
+      "step": 1690
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001067722075637643,
+      "loss": 3.1003,
+      "step": 1691
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010671357373204341,
+      "loss": 3.2445,
+      "step": 1692
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010665493990032249,
+      "loss": 3.0466,
+      "step": 1693
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010659630606860158,
+      "loss": 3.1444,
+      "step": 1694
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010653767223688069,
+      "loss": 3.0813,
+      "step": 1695
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010647903840515978,
+      "loss": 3.1947,
+      "step": 1696
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010642040457343887,
+      "loss": 3.05,
+      "step": 1697
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010636177074171799,
+      "loss": 3.0693,
+      "step": 1698
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010630313690999707,
+      "loss": 3.172,
+      "step": 1699
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010624450307827618,
+      "loss": 3.1232,
+      "step": 1700
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010618586924655527,
+      "loss": 3.1415,
+      "step": 1701
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010612723541483436,
+      "loss": 3.1177,
+      "step": 1702
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010606860158311347,
+      "loss": 3.0969,
+      "step": 1703
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010600996775139255,
+      "loss": 3.1011,
+      "step": 1704
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010595133391967165,
+      "loss": 3.0822,
+      "step": 1705
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010589270008795075,
+      "loss": 3.1034,
+      "step": 1706
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010583406625622985,
+      "loss": 3.1182,
+      "step": 1707
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010577543242450893,
+      "loss": 3.0735,
+      "step": 1708
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010571679859278805,
+      "loss": 3.158,
+      "step": 1709
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010565816476106713,
+      "loss": 3.1124,
+      "step": 1710
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010559953092934624,
+      "loss": 3.0835,
+      "step": 1711
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010554089709762533,
+      "loss": 3.1919,
+      "step": 1712
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010548226326590443,
+      "loss": 3.248,
+      "step": 1713
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010542362943418354,
+      "loss": 3.0471,
+      "step": 1714
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010536499560246262,
+      "loss": 3.1297,
+      "step": 1715
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010530636177074171,
+      "loss": 3.1031,
+      "step": 1716
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010524772793902082,
+      "loss": 3.0993,
+      "step": 1717
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010518909410729991,
+      "loss": 3.0457,
+      "step": 1718
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010513046027557902,
+      "loss": 3.1279,
+      "step": 1719
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010507182644385812,
+      "loss": 3.1209,
+      "step": 1720
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001050131926121372,
+      "loss": 2.9745,
+      "step": 1721
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010495455878041632,
+      "loss": 3.1604,
+      "step": 1722
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001048959249486954,
+      "loss": 3.1523,
+      "step": 1723
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010483729111697449,
+      "loss": 3.1549,
+      "step": 1724
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001047786572852536,
+      "loss": 3.1036,
+      "step": 1725
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001047200234535327,
+      "loss": 3.0448,
+      "step": 1726
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001046613896218118,
+      "loss": 3.102,
+      "step": 1727
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010460275579009088,
+      "loss": 3.0906,
+      "step": 1728
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010454412195836998,
+      "loss": 3.0744,
+      "step": 1729
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010448548812664909,
+      "loss": 3.0511,
+      "step": 1730
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010442685429492818,
+      "loss": 3.1741,
+      "step": 1731
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010436822046320726,
+      "loss": 3.0744,
+      "step": 1732
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010430958663148638,
+      "loss": 3.1373,
+      "step": 1733
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010425095279976546,
+      "loss": 3.1071,
+      "step": 1734
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010419231896804457,
+      "loss": 3.0417,
+      "step": 1735
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010413368513632367,
+      "loss": 3.0372,
+      "step": 1736
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010407505130460276,
+      "loss": 3.1596,
+      "step": 1737
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010401641747288187,
+      "loss": 3.0743,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010395778364116095,
+      "loss": 3.142,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010389914980944004,
+      "loss": 3.1523,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010384051597771915,
+      "loss": 3.1408,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010378188214599825,
+      "loss": 3.0652,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010372324831427733,
+      "loss": 3.1059,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010366461448255645,
+      "loss": 3.0498,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010360598065083553,
+      "loss": 3.2089,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010354734681911465,
+      "loss": 3.0765,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010348871298739373,
+      "loss": 3.2044,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010343007915567282,
+      "loss": 3.1166,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010337144532395193,
+      "loss": 3.0821,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010331281149223103,
+      "loss": 3.0803,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010325417766051011,
+      "loss": 3.1026,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010319554382878922,
+      "loss": 3.1071,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010313690999706831,
+      "loss": 3.1106,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010307827616534742,
+      "loss": 3.0305,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010301964233362651,
+      "loss": 3.1255,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010296100850190559,
+      "loss": 3.0228,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010290237467018471,
+      "loss": 2.9448,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001028437408384638,
+      "loss": 3.1852,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010278510700674289,
+      "loss": 3.0626,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000102726473175022,
+      "loss": 3.1446,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 3.320050001144409,
+      "eval_runtime": 2939.4559,
+      "eval_samples_per_second": 6.97,
+      "eval_steps_per_second": 2.323,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010266783934330109,
+      "loss": 3.1573,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001026092055115802,
+      "loss": 3.0945,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010255057167985928,
+      "loss": 3.028,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010249193784813837,
+      "loss": 3.1876,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010243330401641748,
+      "loss": 3.0141,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010237467018469658,
+      "loss": 3.0834,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010231603635297566,
+      "loss": 3.1375,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010225740252125478,
+      "loss": 3.0324,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010219876868953386,
+      "loss": 3.1365,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010214013485781295,
+      "loss": 3.0708,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010208150102609206,
+      "loss": 3.0796,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010202286719437116,
+      "loss": 3.1592,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010196423336265026,
+      "loss": 3.0714,
+      "step": 1773
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010190559953092936,
+      "loss": 3.1238,
+      "step": 1774
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010184696569920844,
+      "loss": 3.1243,
+      "step": 1775
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010178833186748755,
+      "loss": 3.1121,
+      "step": 1776
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010172969803576664,
+      "loss": 3.1047,
+      "step": 1777
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010167106420404574,
+      "loss": 3.066,
+      "step": 1778
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010161243037232484,
+      "loss": 3.1417,
+      "step": 1779
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010155379654060392,
+      "loss": 3.0592,
+      "step": 1780
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010149516270888305,
+      "loss": 3.1251,
+      "step": 1781
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010143652887716213,
+      "loss": 3.1002,
+      "step": 1782
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010137789504544122,
+      "loss": 3.1436,
+      "step": 1783
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010131926121372033,
+      "loss": 3.0973,
+      "step": 1784
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010126062738199942,
+      "loss": 3.1076,
+      "step": 1785
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001012019935502785,
+      "loss": 3.2234,
+      "step": 1786
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010114335971855761,
+      "loss": 3.1533,
+      "step": 1787
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001010847258868367,
+      "loss": 3.0841,
+      "step": 1788
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010102609205511581,
+      "loss": 3.0895,
+      "step": 1789
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010096745822339491,
+      "loss": 3.2488,
+      "step": 1790
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010090882439167399,
+      "loss": 3.1007,
+      "step": 1791
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010085019055995311,
+      "loss": 3.0292,
+      "step": 1792
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010079155672823219,
+      "loss": 3.0973,
+      "step": 1793
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010073292289651129,
+      "loss": 3.1285,
+      "step": 1794
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010067428906479039,
+      "loss": 3.0221,
+      "step": 1795
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010061565523306949,
+      "loss": 3.0917,
+      "step": 1796
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010055702140134857,
+      "loss": 3.0951,
+      "step": 1797
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010049838756962769,
+      "loss": 3.1532,
+      "step": 1798
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010043975373790677,
+      "loss": 3.1203,
+      "step": 1799
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010038111990618588,
+      "loss": 3.2073,
+      "step": 1800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010032248607446497,
+      "loss": 3.1434,
+      "step": 1801
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010026385224274407,
+      "loss": 3.1396,
+      "step": 1802
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010020521841102317,
+      "loss": 3.2,
+      "step": 1803
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010014658457930226,
+      "loss": 3.1205,
+      "step": 1804
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010008795074758135,
+      "loss": 3.2149,
+      "step": 1805
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010002931691586046,
+      "loss": 3.0374,
+      "step": 1806
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.997068308413955e-05,
+      "loss": 3.0511,
+      "step": 1807
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.991204925241865e-05,
+      "loss": 3.1071,
+      "step": 1808
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.985341542069775e-05,
+      "loss": 3.055,
+      "step": 1809
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.979478158897685e-05,
+      "loss": 3.0356,
+      "step": 1810
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.973614775725594e-05,
+      "loss": 3.1747,
+      "step": 1811
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.967751392553504e-05,
+      "loss": 3.1509,
+      "step": 1812
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.961888009381413e-05,
+      "loss": 3.1017,
+      "step": 1813
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.956024626209324e-05,
+      "loss": 3.1541,
+      "step": 1814
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.950161243037232e-05,
+      "loss": 3.0134,
+      "step": 1815
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.944297859865143e-05,
+      "loss": 3.0249,
+      "step": 1816
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.938434476693052e-05,
+      "loss": 3.0088,
+      "step": 1817
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.932571093520963e-05,
+      "loss": 3.0087,
+      "step": 1818
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.926707710348871e-05,
+      "loss": 3.1616,
+      "step": 1819
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.920844327176782e-05,
+      "loss": 3.0938,
+      "step": 1820
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.914980944004691e-05,
+      "loss": 3.0721,
+      "step": 1821
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.909117560832601e-05,
+      "loss": 3.2352,
+      "step": 1822
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.90325417766051e-05,
+      "loss": 3.0524,
+      "step": 1823
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.897390794488421e-05,
+      "loss": 3.1045,
+      "step": 1824
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.89152741131633e-05,
+      "loss": 3.0853,
+      "step": 1825
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.88566402814424e-05,
+      "loss": 3.121,
+      "step": 1826
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.879800644972149e-05,
+      "loss": 3.0032,
+      "step": 1827
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.873937261800059e-05,
+      "loss": 3.1874,
+      "step": 1828
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.86807387862797e-05,
+      "loss": 3.0471,
+      "step": 1829
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.862210495455878e-05,
+      "loss": 3.1055,
+      "step": 1830
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.856347112283788e-05,
+      "loss": 3.0707,
+      "step": 1831
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.850483729111698e-05,
+      "loss": 3.1021,
+      "step": 1832
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.844620345939609e-05,
+      "loss": 3.1917,
+      "step": 1833
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.838756962767517e-05,
+      "loss": 3.0467,
+      "step": 1834
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.832893579595427e-05,
+      "loss": 3.1434,
+      "step": 1835
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.827030196423337e-05,
+      "loss": 3.1226,
+      "step": 1836
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.821166813251246e-05,
+      "loss": 3.0731,
+      "step": 1837
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.815303430079156e-05,
+      "loss": 3.1525,
+      "step": 1838
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.809440046907065e-05,
+      "loss": 3.0906,
+      "step": 1839
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.803576663734976e-05,
+      "loss": 3.0288,
+      "step": 1840
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.797713280562885e-05,
+      "loss": 3.0371,
+      "step": 1841
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.791849897390795e-05,
+      "loss": 3.0029,
+      "step": 1842
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.785986514218704e-05,
+      "loss": 3.158,
+      "step": 1843
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.780123131046615e-05,
+      "loss": 3.1321,
+      "step": 1844
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.774259747874524e-05,
+      "loss": 3.1156,
+      "step": 1845
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.768396364702434e-05,
+      "loss": 3.0089,
+      "step": 1846
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.762532981530343e-05,
+      "loss": 3.1628,
+      "step": 1847
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.756669598358254e-05,
+      "loss": 3.1002,
+      "step": 1848
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.750806215186162e-05,
+      "loss": 3.0235,
+      "step": 1849
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.744942832014073e-05,
+      "loss": 2.957,
+      "step": 1850
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.739079448841982e-05,
+      "loss": 3.0432,
+      "step": 1851
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.733216065669892e-05,
+      "loss": 3.1521,
+      "step": 1852
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.727352682497801e-05,
+      "loss": 3.0779,
+      "step": 1853
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.721489299325711e-05,
+      "loss": 3.1235,
+      "step": 1854
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.715625916153621e-05,
+      "loss": 2.9702,
+      "step": 1855
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.709762532981531e-05,
+      "loss": 3.136,
+      "step": 1856
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.70389914980944e-05,
+      "loss": 3.0957,
+      "step": 1857
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.69803576663735e-05,
+      "loss": 3.055,
+      "step": 1858
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.69217238346526e-05,
+      "loss": 3.0824,
+      "step": 1859
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.68630900029317e-05,
+      "loss": 3.0799,
+      "step": 1860
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.68044561712108e-05,
+      "loss": 3.0447,
+      "step": 1861
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.674582233948989e-05,
+      "loss": 3.1054,
+      "step": 1862
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.668718850776898e-05,
+      "loss": 3.0275,
+      "step": 1863
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.662855467604809e-05,
+      "loss": 3.1579,
+      "step": 1864
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.656992084432717e-05,
+      "loss": 3.0837,
+      "step": 1865
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.651128701260628e-05,
+      "loss": 3.0448,
+      "step": 1866
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.645265318088537e-05,
+      "loss": 2.9995,
+      "step": 1867
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.639401934916448e-05,
+      "loss": 3.0649,
+      "step": 1868
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.633538551744356e-05,
+      "loss": 3.0695,
+      "step": 1869
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.627675168572267e-05,
+      "loss": 3.0222,
+      "step": 1870
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.621811785400176e-05,
+      "loss": 3.1588,
+      "step": 1871
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.615948402228087e-05,
+      "loss": 3.09,
+      "step": 1872
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.610085019055995e-05,
+      "loss": 3.1339,
+      "step": 1873
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.604221635883906e-05,
+      "loss": 3.0358,
+      "step": 1874
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.598358252711816e-05,
+      "loss": 3.0608,
+      "step": 1875
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.592494869539725e-05,
+      "loss": 3.1485,
+      "step": 1876
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.586631486367634e-05,
+      "loss": 3.0677,
+      "step": 1877
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.580768103195544e-05,
+      "loss": 3.0062,
+      "step": 1878
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.574904720023455e-05,
+      "loss": 3.0928,
+      "step": 1879
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.569041336851363e-05,
+      "loss": 3.1299,
+      "step": 1880
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.563177953679273e-05,
+      "loss": 3.1471,
+      "step": 1881
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.557314570507183e-05,
+      "loss": 3.1515,
+      "step": 1882
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.551451187335094e-05,
+      "loss": 3.077,
+      "step": 1883
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.545587804163002e-05,
+      "loss": 3.1111,
+      "step": 1884
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.539724420990913e-05,
+      "loss": 3.1064,
+      "step": 1885
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.533861037818822e-05,
+      "loss": 3.1031,
+      "step": 1886
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.527997654646731e-05,
+      "loss": 3.0807,
+      "step": 1887
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.522134271474641e-05,
+      "loss": 3.1281,
+      "step": 1888
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.51627088830255e-05,
+      "loss": 3.1272,
+      "step": 1889
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.510407505130461e-05,
+      "loss": 3.1405,
+      "step": 1890
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.50454412195837e-05,
+      "loss": 3.1068,
+      "step": 1891
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.49868073878628e-05,
+      "loss": 2.9872,
+      "step": 1892
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.49281735561419e-05,
+      "loss": 3.1021,
+      "step": 1893
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.4869539724421e-05,
+      "loss": 3.053,
+      "step": 1894
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.48109058927001e-05,
+      "loss": 3.1373,
+      "step": 1895
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.475227206097919e-05,
+      "loss": 3.1356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.469363822925828e-05,
+      "loss": 3.1121,
+      "step": 1897
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.463500439753739e-05,
+      "loss": 3.0237,
+      "step": 1898
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.457637056581649e-05,
+      "loss": 3.0802,
+      "step": 1899
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.451773673409558e-05,
+      "loss": 2.99,
+      "step": 1900
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.445910290237468e-05,
+      "loss": 3.1088,
+      "step": 1901
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.440046907065377e-05,
+      "loss": 3.0627,
+      "step": 1902
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.434183523893288e-05,
+      "loss": 3.0744,
+      "step": 1903
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.428320140721196e-05,
+      "loss": 3.0008,
+      "step": 1904
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.422456757549107e-05,
+      "loss": 3.11,
+      "step": 1905
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.416593374377016e-05,
+      "loss": 3.0389,
+      "step": 1906
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.410729991204925e-05,
+      "loss": 3.0924,
+      "step": 1907
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.404866608032835e-05,
+      "loss": 3.0821,
+      "step": 1908
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.399003224860746e-05,
+      "loss": 3.0144,
+      "step": 1909
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.393139841688655e-05,
+      "loss": 3.0703,
+      "step": 1910
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.387276458516565e-05,
+      "loss": 3.1178,
+      "step": 1911
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.381413075344474e-05,
+      "loss": 3.077,
+      "step": 1912
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.375549692172383e-05,
+      "loss": 2.9368,
+      "step": 1913
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.369686309000294e-05,
+      "loss": 3.1405,
+      "step": 1914
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.363822925828202e-05,
+      "loss": 2.9568,
+      "step": 1915
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.357959542656113e-05,
+      "loss": 3.1213,
+      "step": 1916
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.352096159484023e-05,
+      "loss": 3.1674,
+      "step": 1917
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.346232776311933e-05,
+      "loss": 3.0357,
+      "step": 1918
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.340369393139841e-05,
+      "loss": 3.048,
+      "step": 1919
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.334506009967752e-05,
+      "loss": 2.9384,
+      "step": 1920
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.328642626795662e-05,
+      "loss": 3.0337,
+      "step": 1921
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.322779243623571e-05,
+      "loss": 3.1666,
+      "step": 1922
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.31691586045148e-05,
+      "loss": 3.0372,
+      "step": 1923
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.31105247727939e-05,
+      "loss": 3.1386,
+      "step": 1924
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.305189094107301e-05,
+      "loss": 3.1946,
+      "step": 1925
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.29932571093521e-05,
+      "loss": 3.1404,
+      "step": 1926
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.29346232776312e-05,
+      "loss": 2.9996,
+      "step": 1927
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.287598944591029e-05,
+      "loss": 3.123,
+      "step": 1928
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.28173556141894e-05,
+      "loss": 3.1337,
+      "step": 1929
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.275872178246849e-05,
+      "loss": 3.0988,
+      "step": 1930
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.270008795074759e-05,
+      "loss": 3.132,
+      "step": 1931
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.264145411902668e-05,
+      "loss": 2.9339,
+      "step": 1932
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.258282028730579e-05,
+      "loss": 3.0116,
+      "step": 1933
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.252418645558487e-05,
+      "loss": 3.1258,
+      "step": 1934
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.246555262386398e-05,
+      "loss": 3.0558,
+      "step": 1935
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.240691879214307e-05,
+      "loss": 2.9847,
+      "step": 1936
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 3.3032548427581787,
+      "eval_runtime": 2939.4432,
+      "eval_samples_per_second": 6.97,
+      "eval_steps_per_second": 2.323,
+      "step": 1936
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.234828496042217e-05,
+      "loss": 3.1278,
+      "step": 1937
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.228965112870126e-05,
+      "loss": 3.0167,
+      "step": 1938
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.223101729698035e-05,
+      "loss": 3.0658,
+      "step": 1939
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.217238346525946e-05,
+      "loss": 3.1698,
+      "step": 1940
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.211374963353856e-05,
+      "loss": 3.0073,
+      "step": 1941
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.205511580181765e-05,
+      "loss": 2.965,
+      "step": 1942
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.199648197009675e-05,
+      "loss": 2.9975,
+      "step": 1943
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.193784813837585e-05,
+      "loss": 3.1208,
+      "step": 1944
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.187921430665495e-05,
+      "loss": 3.0442,
+      "step": 1945
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.182058047493404e-05,
+      "loss": 3.0306,
+      "step": 1946
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.176194664321314e-05,
+      "loss": 3.0539,
+      "step": 1947
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.170331281149223e-05,
+      "loss": 2.979,
+      "step": 1948
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.164467897977134e-05,
+      "loss": 3.0677,
+      "step": 1949
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.158604514805042e-05,
+      "loss": 2.9672,
+      "step": 1950
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.152741131632953e-05,
+      "loss": 3.1164,
+      "step": 1951
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.146877748460862e-05,
+      "loss": 3.0753,
+      "step": 1952
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.141014365288773e-05,
+      "loss": 3.1209,
+      "step": 1953
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.135150982116681e-05,
+      "loss": 3.1353,
+      "step": 1954
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.129287598944592e-05,
+      "loss": 3.1298,
+      "step": 1955
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.123424215772501e-05,
+      "loss": 3.0803,
+      "step": 1956
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.117560832600412e-05,
+      "loss": 3.1678,
+      "step": 1957
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.11169744942832e-05,
+      "loss": 3.0652,
+      "step": 1958
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.105834066256231e-05,
+      "loss": 3.1459,
+      "step": 1959
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.09997068308414e-05,
+      "loss": 2.9366,
+      "step": 1960
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.09410729991205e-05,
+      "loss": 3.0726,
+      "step": 1961
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.088243916739959e-05,
+      "loss": 3.0109,
+      "step": 1962
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.082380533567869e-05,
+      "loss": 3.0312,
+      "step": 1963
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.07651715039578e-05,
+      "loss": 3.0686,
+      "step": 1964
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.070653767223687e-05,
+      "loss": 3.0977,
+      "step": 1965
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.064790384051598e-05,
+      "loss": 3.079,
+      "step": 1966
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.058927000879508e-05,
+      "loss": 3.0986,
+      "step": 1967
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.053063617707418e-05,
+      "loss": 3.1546,
+      "step": 1968
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.047200234535327e-05,
+      "loss": 2.9669,
+      "step": 1969
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.041336851363237e-05,
+      "loss": 3.1267,
+      "step": 1970
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.035473468191147e-05,
+      "loss": 3.09,
+      "step": 1971
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.029610085019056e-05,
+      "loss": 3.0128,
+      "step": 1972
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.023746701846966e-05,
+      "loss": 3.0213,
+      "step": 1973
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.017883318674875e-05,
+      "loss": 2.9842,
+      "step": 1974
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.012019935502786e-05,
+      "loss": 3.0716,
+      "step": 1975
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.006156552330695e-05,
+      "loss": 3.1559,
+      "step": 1976
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.000293169158605e-05,
+      "loss": 2.9197,
+      "step": 1977
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.994429785986514e-05,
+      "loss": 2.9311,
+      "step": 1978
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.988566402814425e-05,
+      "loss": 3.086,
+      "step": 1979
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.982703019642334e-05,
+      "loss": 3.0,
+      "step": 1980
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.976839636470244e-05,
+      "loss": 3.0534,
+      "step": 1981
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.970976253298153e-05,
+      "loss": 2.9953,
+      "step": 1982
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.965112870126064e-05,
+      "loss": 3.1116,
+      "step": 1983
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.959249486953973e-05,
+      "loss": 3.1197,
+      "step": 1984
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.953386103781883e-05,
+      "loss": 3.0178,
+      "step": 1985
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.947522720609792e-05,
+      "loss": 3.0182,
+      "step": 1986
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.941659337437702e-05,
+      "loss": 2.9657,
+      "step": 1987
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.935795954265613e-05,
+      "loss": 3.1071,
+      "step": 1988
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.92993257109352e-05,
+      "loss": 3.1459,
+      "step": 1989
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.924069187921431e-05,
+      "loss": 3.0574,
+      "step": 1990
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.918205804749341e-05,
+      "loss": 2.9549,
+      "step": 1991
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.91234242157725e-05,
+      "loss": 3.0443,
+      "step": 1992
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.90647903840516e-05,
+      "loss": 2.9937,
+      "step": 1993
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.90061565523307e-05,
+      "loss": 3.0391,
+      "step": 1994
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.89475227206098e-05,
+      "loss": 2.9892,
+      "step": 1995
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 3.0484,
+      "step": 1996
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.883025505716799e-05,
+      "loss": 3.0741,
+      "step": 1997
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.877162122544708e-05,
+      "loss": 3.1785,
+      "step": 1998
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.871298739372619e-05,
+      "loss": 3.0739,
+      "step": 1999
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.865435356200527e-05,
+      "loss": 3.0987,
+      "step": 2000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.859571973028438e-05,
+      "loss": 3.1262,
+      "step": 2001
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.853708589856347e-05,
+      "loss": 3.077,
+      "step": 2002
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.847845206684258e-05,
+      "loss": 2.9417,
+      "step": 2003
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.841981823512166e-05,
+      "loss": 2.9296,
+      "step": 2004
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.836118440340077e-05,
+      "loss": 3.0343,
+      "step": 2005
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.830255057167986e-05,
+      "loss": 3.1408,
+      "step": 2006
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.824391673995897e-05,
+      "loss": 3.0642,
+      "step": 2007
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.818528290823805e-05,
+      "loss": 3.0765,
+      "step": 2008
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.812664907651716e-05,
+      "loss": 3.0626,
+      "step": 2009
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.806801524479625e-05,
+      "loss": 3.0625,
+      "step": 2010
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.800938141307535e-05,
+      "loss": 2.9488,
+      "step": 2011
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.795074758135444e-05,
+      "loss": 3.0291,
+      "step": 2012
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.789211374963354e-05,
+      "loss": 3.0251,
+      "step": 2013
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.783347991791265e-05,
+      "loss": 3.0409,
+      "step": 2014
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.777484608619174e-05,
+      "loss": 3.0749,
+      "step": 2015
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.771621225447083e-05,
+      "loss": 3.1053,
+      "step": 2016
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.765757842274993e-05,
+      "loss": 3.1473,
+      "step": 2017
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.759894459102904e-05,
+      "loss": 3.0467,
+      "step": 2018
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.754031075930812e-05,
+      "loss": 2.9854,
+      "step": 2019
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.748167692758722e-05,
+      "loss": 2.9706,
+      "step": 2020
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.742304309586632e-05,
+      "loss": 3.0775,
+      "step": 2021
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.736440926414541e-05,
+      "loss": 3.0419,
+      "step": 2022
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.730577543242451e-05,
+      "loss": 3.107,
+      "step": 2023
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.72471416007036e-05,
+      "loss": 2.9794,
+      "step": 2024
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.718850776898271e-05,
+      "loss": 3.0311,
+      "step": 2025
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.71298739372618e-05,
+      "loss": 3.1402,
+      "step": 2026
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.70712401055409e-05,
+      "loss": 2.9644,
+      "step": 2027
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.701260627381999e-05,
+      "loss": 2.997,
+      "step": 2028
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.69539724420991e-05,
+      "loss": 2.9759,
+      "step": 2029
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.68953386103782e-05,
+      "loss": 3.0223,
+      "step": 2030
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.683670477865729e-05,
+      "loss": 3.1275,
+      "step": 2031
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.677807094693638e-05,
+      "loss": 3.071,
+      "step": 2032
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.671943711521549e-05,
+      "loss": 3.1222,
+      "step": 2033
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.666080328349459e-05,
+      "loss": 2.9666,
+      "step": 2034
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.660216945177368e-05,
+      "loss": 3.034,
+      "step": 2035
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.654353562005277e-05,
+      "loss": 3.04,
+      "step": 2036
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.648490178833187e-05,
+      "loss": 3.0028,
+      "step": 2037
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.642626795661098e-05,
+      "loss": 3.087,
+      "step": 2038
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.636763412489006e-05,
+      "loss": 3.164,
+      "step": 2039
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.630900029316917e-05,
+      "loss": 3.0221,
+      "step": 2040
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.625036646144826e-05,
+      "loss": 3.0615,
+      "step": 2041
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.619173262972737e-05,
+      "loss": 3.0341,
+      "step": 2042
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.613309879800645e-05,
+      "loss": 3.0749,
+      "step": 2043
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.607446496628556e-05,
+      "loss": 2.9954,
+      "step": 2044
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.601583113456465e-05,
+      "loss": 3.0631,
+      "step": 2045
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.595719730284374e-05,
+      "loss": 3.1342,
+      "step": 2046
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.589856347112284e-05,
+      "loss": 3.076,
+      "step": 2047
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.583992963940193e-05,
+      "loss": 3.1095,
+      "step": 2048
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.578129580768104e-05,
+      "loss": 3.0005,
+      "step": 2049
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.572266197596012e-05,
+      "loss": 3.033,
+      "step": 2050
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.566402814423923e-05,
+      "loss": 3.0772,
+      "step": 2051
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.560539431251832e-05,
+      "loss": 3.0078,
+      "step": 2052
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.554676048079743e-05,
+      "loss": 3.1205,
+      "step": 2053
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.548812664907651e-05,
+      "loss": 2.9833,
+      "step": 2054
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.542949281735562e-05,
+      "loss": 2.9453,
+      "step": 2055
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.537085898563471e-05,
+      "loss": 3.0687,
+      "step": 2056
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.531222515391382e-05,
+      "loss": 2.9526,
+      "step": 2057
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.52535913221929e-05,
+      "loss": 2.9666,
+      "step": 2058
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.519495749047201e-05,
+      "loss": 3.0944,
+      "step": 2059
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.51363236587511e-05,
+      "loss": 3.0659,
+      "step": 2060
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.50776898270302e-05,
+      "loss": 3.1344,
+      "step": 2061
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.50190559953093e-05,
+      "loss": 2.9884,
+      "step": 2062
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.496042216358839e-05,
+      "loss": 3.1274,
+      "step": 2063
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.49017883318675e-05,
+      "loss": 2.9231,
+      "step": 2064
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.484315450014659e-05,
+      "loss": 2.9281,
+      "step": 2065
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.478452066842569e-05,
+      "loss": 3.0655,
+      "step": 2066
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.472588683670478e-05,
+      "loss": 2.9997,
+      "step": 2067
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.466725300498389e-05,
+      "loss": 3.0595,
+      "step": 2068
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.460861917326298e-05,
+      "loss": 3.0622,
+      "step": 2069
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.454998534154208e-05,
+      "loss": 3.0297,
+      "step": 2070
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.449135150982117e-05,
+      "loss": 3.0824,
+      "step": 2071
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.443271767810026e-05,
+      "loss": 3.0544,
+      "step": 2072
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.437408384637937e-05,
+      "loss": 3.0849,
+      "step": 2073
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.431545001465845e-05,
+      "loss": 3.0525,
+      "step": 2074
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.425681618293756e-05,
+      "loss": 3.0667,
+      "step": 2075
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.419818235121666e-05,
+      "loss": 3.0996,
+      "step": 2076
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.413954851949575e-05,
+      "loss": 3.0324,
+      "step": 2077
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.408091468777484e-05,
+      "loss": 3.086,
+      "step": 2078
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.402228085605395e-05,
+      "loss": 3.0223,
+      "step": 2079
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.396364702433305e-05,
+      "loss": 3.0605,
+      "step": 2080
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.390501319261214e-05,
+      "loss": 3.0081,
+      "step": 2081
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.384637936089123e-05,
+      "loss": 2.9988,
+      "step": 2082
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.378774552917034e-05,
+      "loss": 3.0181,
+      "step": 2083
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.372911169744944e-05,
+      "loss": 3.0363,
+      "step": 2084
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.367047786572853e-05,
+      "loss": 3.0154,
+      "step": 2085
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.361184403400763e-05,
+      "loss": 3.0487,
+      "step": 2086
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.355321020228672e-05,
+      "loss": 3.0229,
+      "step": 2087
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.349457637056583e-05,
+      "loss": 3.0661,
+      "step": 2088
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.343594253884491e-05,
+      "loss": 2.9821,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.337730870712402e-05,
+      "loss": 3.0254,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.331867487540311e-05,
+      "loss": 3.0446,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.326004104368222e-05,
+      "loss": 2.99,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.32014072119613e-05,
+      "loss": 2.9763,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.314277338024041e-05,
+      "loss": 2.9533,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.30841395485195e-05,
+      "loss": 3.0655,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.30255057167986e-05,
+      "loss": 3.0108,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.296687188507769e-05,
+      "loss": 2.9899,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.290823805335678e-05,
+      "loss": 3.0429,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.284960422163589e-05,
+      "loss": 3.0015,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.279097038991499e-05,
+      "loss": 2.9944,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.273233655819408e-05,
+      "loss": 3.0209,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.267370272647318e-05,
+      "loss": 3.132,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.261506889475228e-05,
+      "loss": 3.0258,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.255643506303136e-05,
+      "loss": 3.0699,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.249780123131047e-05,
+      "loss": 2.9658,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.243916739958957e-05,
+      "loss": 2.9796,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.238053356786866e-05,
+      "loss": 3.0962,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.232189973614775e-05,
+      "loss": 3.032,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.226326590442686e-05,
+      "loss": 3.0286,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.220463207270596e-05,
+      "loss": 2.9863,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.214599824098505e-05,
+      "loss": 3.0452,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.208736440926415e-05,
+      "loss": 3.1286,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 3.279003858566284,
+      "eval_runtime": 2938.7929,
+      "eval_samples_per_second": 6.971,
+      "eval_steps_per_second": 2.324,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.202873057754324e-05,
+      "loss": 2.9913,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.197009674582235e-05,
+      "loss": 3.0838,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.191146291410144e-05,
+      "loss": 2.9803,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.185282908238054e-05,
+      "loss": 3.0087,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.179419525065963e-05,
+      "loss": 3.0088,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.173556141893874e-05,
+      "loss": 3.0678,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.167692758721783e-05,
+      "loss": 2.9735,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.161829375549693e-05,
+      "loss": 3.0428,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.155965992377602e-05,
+      "loss": 2.979,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.150102609205512e-05,
+      "loss": 3.0504,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.144239226033422e-05,
+      "loss": 2.8659,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.13837584286133e-05,
+      "loss": 2.92,
+      "step": 2124
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.132512459689241e-05,
+      "loss": 3.0536,
+      "step": 2125
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.126649076517151e-05,
+      "loss": 3.0977,
+      "step": 2126
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.120785693345061e-05,
+      "loss": 2.9551,
+      "step": 2127
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.11492231017297e-05,
+      "loss": 3.1118,
+      "step": 2128
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.10905892700088e-05,
+      "loss": 3.0925,
+      "step": 2129
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.10319554382879e-05,
+      "loss": 3.0587,
+      "step": 2130
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.097332160656699e-05,
+      "loss": 2.9647,
+      "step": 2131
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.091468777484609e-05,
+      "loss": 2.8933,
+      "step": 2132
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.085605394312518e-05,
+      "loss": 3.0305,
+      "step": 2133
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.079742011140429e-05,
+      "loss": 3.0834,
+      "step": 2134
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.073878627968337e-05,
+      "loss": 3.0384,
+      "step": 2135
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.068015244796248e-05,
+      "loss": 3.1248,
+      "step": 2136
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.062151861624157e-05,
+      "loss": 2.9678,
+      "step": 2137
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.056288478452068e-05,
+      "loss": 3.0379,
+      "step": 2138
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.050425095279976e-05,
+      "loss": 2.9091,
+      "step": 2139
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.044561712107887e-05,
+      "loss": 3.0026,
+      "step": 2140
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.038698328935796e-05,
+      "loss": 3.0794,
+      "step": 2141
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.032834945763707e-05,
+      "loss": 3.1296,
+      "step": 2142
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.026971562591615e-05,
+      "loss": 3.1056,
+      "step": 2143
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.021108179419526e-05,
+      "loss": 3.0213,
+      "step": 2144
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.015244796247435e-05,
+      "loss": 3.1048,
+      "step": 2145
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.009381413075345e-05,
+      "loss": 3.0592,
+      "step": 2146
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 8.003518029903254e-05,
+      "loss": 3.0547,
+      "step": 2147
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.997654646731164e-05,
+      "loss": 3.1033,
+      "step": 2148
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.991791263559074e-05,
+      "loss": 3.0537,
+      "step": 2149
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.985927880386984e-05,
+      "loss": 3.1241,
+      "step": 2150
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.980064497214893e-05,
+      "loss": 3.0887,
+      "step": 2151
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.974201114042803e-05,
+      "loss": 3.0157,
+      "step": 2152
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.968337730870713e-05,
+      "loss": 2.8817,
+      "step": 2153
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.962474347698623e-05,
+      "loss": 3.1184,
+      "step": 2154
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.956610964526532e-05,
+      "loss": 2.9394,
+      "step": 2155
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.950747581354442e-05,
+      "loss": 3.0529,
+      "step": 2156
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.944884198182351e-05,
+      "loss": 3.0572,
+      "step": 2157
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.939020815010262e-05,
+      "loss": 2.9754,
+      "step": 2158
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.93315743183817e-05,
+      "loss": 3.0794,
+      "step": 2159
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.927294048666081e-05,
+      "loss": 3.0672,
+      "step": 2160
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.92143066549399e-05,
+      "loss": 3.1011,
+      "step": 2161
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.9155672823219e-05,
+      "loss": 3.0374,
+      "step": 2162
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.909703899149809e-05,
+      "loss": 3.0082,
+      "step": 2163
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.90384051597772e-05,
+      "loss": 3.1259,
+      "step": 2164
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.89797713280563e-05,
+      "loss": 3.0714,
+      "step": 2165
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.892113749633539e-05,
+      "loss": 3.0764,
+      "step": 2166
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.886250366461448e-05,
+      "loss": 3.0549,
+      "step": 2167
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.880386983289359e-05,
+      "loss": 3.0911,
+      "step": 2168
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.874523600117268e-05,
+      "loss": 3.0779,
+      "step": 2169
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.868660216945178e-05,
+      "loss": 2.9613,
+      "step": 2170
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.862796833773087e-05,
+      "loss": 3.1114,
+      "step": 2171
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.856933450600997e-05,
+      "loss": 3.0424,
+      "step": 2172
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.851070067428908e-05,
+      "loss": 3.0557,
+      "step": 2173
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.845206684256816e-05,
+      "loss": 2.9939,
+      "step": 2174
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.839343301084726e-05,
+      "loss": 3.0031,
+      "step": 2175
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.833479917912636e-05,
+      "loss": 2.9967,
+      "step": 2176
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.827616534740547e-05,
+      "loss": 2.9714,
+      "step": 2177
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.821753151568455e-05,
+      "loss": 2.9984,
+      "step": 2178
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.815889768396365e-05,
+      "loss": 3.1013,
+      "step": 2179
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.810026385224275e-05,
+      "loss": 3.0813,
+      "step": 2180
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.804163002052184e-05,
+      "loss": 2.9323,
+      "step": 2181
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.798299618880094e-05,
+      "loss": 2.9601,
+      "step": 2182
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.792436235708003e-05,
+      "loss": 2.9847,
+      "step": 2183
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.786572852535914e-05,
+      "loss": 3.081,
+      "step": 2184
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.780709469363823e-05,
+      "loss": 3.0644,
+      "step": 2185
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.774846086191733e-05,
+      "loss": 3.0474,
+      "step": 2186
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.768982703019642e-05,
+      "loss": 3.0382,
+      "step": 2187
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.763119319847553e-05,
+      "loss": 3.0454,
+      "step": 2188
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.757255936675461e-05,
+      "loss": 3.0269,
+      "step": 2189
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.751392553503372e-05,
+      "loss": 2.9757,
+      "step": 2190
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.745529170331281e-05,
+      "loss": 2.9173,
+      "step": 2191
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.739665787159192e-05,
+      "loss": 2.992,
+      "step": 2192
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.7338024039871e-05,
+      "loss": 2.976,
+      "step": 2193
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.727939020815011e-05,
+      "loss": 3.0376,
+      "step": 2194
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.72207563764292e-05,
+      "loss": 3.0334,
+      "step": 2195
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.71621225447083e-05,
+      "loss": 3.0348,
+      "step": 2196
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.71034887129874e-05,
+      "loss": 3.0085,
+      "step": 2197
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.704485488126649e-05,
+      "loss": 2.9804,
+      "step": 2198
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.69862210495456e-05,
+      "loss": 2.9899,
+      "step": 2199
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.692758721782469e-05,
+      "loss": 2.9572,
+      "step": 2200
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.686895338610378e-05,
+      "loss": 3.043,
+      "step": 2201
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.681031955438288e-05,
+      "loss": 3.0281,
+      "step": 2202
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.675168572266199e-05,
+      "loss": 2.8836,
+      "step": 2203
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.669305189094108e-05,
+      "loss": 3.0611,
+      "step": 2204
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.663441805922017e-05,
+      "loss": 2.9676,
+      "step": 2205
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.657578422749927e-05,
+      "loss": 3.1005,
+      "step": 2206
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.651715039577836e-05,
+      "loss": 2.9966,
+      "step": 2207
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.645851656405747e-05,
+      "loss": 3.0027,
+      "step": 2208
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.639988273233655e-05,
+      "loss": 3.0368,
+      "step": 2209
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.634124890061566e-05,
+      "loss": 3.0812,
+      "step": 2210
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.628261506889475e-05,
+      "loss": 3.1381,
+      "step": 2211
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.622398123717386e-05,
+      "loss": 3.1242,
+      "step": 2212
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.616534740545294e-05,
+      "loss": 3.0857,
+      "step": 2213
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.610671357373205e-05,
+      "loss": 3.0792,
+      "step": 2214
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.604807974201115e-05,
+      "loss": 3.1115,
+      "step": 2215
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.598944591029024e-05,
+      "loss": 3.0161,
+      "step": 2216
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.593081207856933e-05,
+      "loss": 3.1468,
+      "step": 2217
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.587217824684844e-05,
+      "loss": 2.8137,
+      "step": 2218
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.581354441512754e-05,
+      "loss": 3.0643,
+      "step": 2219
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.575491058340663e-05,
+      "loss": 3.114,
+      "step": 2220
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.569627675168572e-05,
+      "loss": 2.95,
+      "step": 2221
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.563764291996482e-05,
+      "loss": 2.9803,
+      "step": 2222
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.557900908824393e-05,
+      "loss": 3.0316,
+      "step": 2223
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.552037525652301e-05,
+      "loss": 2.9819,
+      "step": 2224
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.546174142480212e-05,
+      "loss": 2.9961,
+      "step": 2225
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.540310759308121e-05,
+      "loss": 3.0465,
+      "step": 2226
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.534447376136032e-05,
+      "loss": 2.9261,
+      "step": 2227
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.52858399296394e-05,
+      "loss": 2.9324,
+      "step": 2228
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.52272060979185e-05,
+      "loss": 2.9709,
+      "step": 2229
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.51685722661976e-05,
+      "loss": 3.0211,
+      "step": 2230
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.51099384344767e-05,
+      "loss": 2.9417,
+      "step": 2231
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.505130460275579e-05,
+      "loss": 3.1206,
+      "step": 2232
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.499267077103488e-05,
+      "loss": 3.0087,
+      "step": 2233
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.493403693931399e-05,
+      "loss": 3.0394,
+      "step": 2234
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.487540310759309e-05,
+      "loss": 2.9766,
+      "step": 2235
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.481676927587218e-05,
+      "loss": 3.0468,
+      "step": 2236
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.475813544415127e-05,
+      "loss": 3.0013,
+      "step": 2237
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.469950161243038e-05,
+      "loss": 3.077,
+      "step": 2238
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.464086778070948e-05,
+      "loss": 3.0647,
+      "step": 2239
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.458223394898857e-05,
+      "loss": 2.9289,
+      "step": 2240
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.452360011726767e-05,
+      "loss": 3.1249,
+      "step": 2241
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.446496628554677e-05,
+      "loss": 3.096,
+      "step": 2242
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.440633245382587e-05,
+      "loss": 2.9995,
+      "step": 2243
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.434769862210496e-05,
+      "loss": 3.1667,
+      "step": 2244
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.428906479038406e-05,
+      "loss": 3.0341,
+      "step": 2245
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.423043095866315e-05,
+      "loss": 3.0475,
+      "step": 2246
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.417179712694224e-05,
+      "loss": 3.0424,
+      "step": 2247
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.411316329522134e-05,
+      "loss": 3.0059,
+      "step": 2248
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.405452946350045e-05,
+      "loss": 3.0318,
+      "step": 2249
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.399589563177954e-05,
+      "loss": 3.0965,
+      "step": 2250
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.393726180005864e-05,
+      "loss": 3.0951,
+      "step": 2251
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.387862796833773e-05,
+      "loss": 3.024,
+      "step": 2252
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.381999413661684e-05,
+      "loss": 3.028,
+      "step": 2253
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.376136030489593e-05,
+      "loss": 3.0371,
+      "step": 2254
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.370272647317503e-05,
+      "loss": 2.8833,
+      "step": 2255
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.364409264145412e-05,
+      "loss": 2.9803,
+      "step": 2256
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.358545880973322e-05,
+      "loss": 3.1169,
+      "step": 2257
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.352682497801232e-05,
+      "loss": 3.0181,
+      "step": 2258
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.34681911462914e-05,
+      "loss": 2.974,
+      "step": 2259
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.340955731457051e-05,
+      "loss": 3.0522,
+      "step": 2260
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.33509234828496e-05,
+      "loss": 2.9721,
+      "step": 2261
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.329228965112871e-05,
+      "loss": 3.0173,
+      "step": 2262
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.32336558194078e-05,
+      "loss": 3.1451,
+      "step": 2263
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.31750219876869e-05,
+      "loss": 3.0079,
+      "step": 2264
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.3116388155966e-05,
+      "loss": 3.0314,
+      "step": 2265
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.30577543242451e-05,
+      "loss": 2.9993,
+      "step": 2266
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.299912049252419e-05,
+      "loss": 2.9569,
+      "step": 2267
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.294048666080329e-05,
+      "loss": 2.911,
+      "step": 2268
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.288185282908239e-05,
+      "loss": 3.0419,
+      "step": 2269
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.282321899736148e-05,
+      "loss": 3.0791,
+      "step": 2270
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.276458516564058e-05,
+      "loss": 3.0664,
+      "step": 2271
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.270595133391967e-05,
+      "loss": 3.0604,
+      "step": 2272
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.264731750219878e-05,
+      "loss": 3.0102,
+      "step": 2273
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.258868367047786e-05,
+      "loss": 2.9504,
+      "step": 2274
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.253004983875697e-05,
+      "loss": 3.0398,
+      "step": 2275
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.247141600703606e-05,
+      "loss": 3.0571,
+      "step": 2276
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.241278217531517e-05,
+      "loss": 3.0104,
+      "step": 2277
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.235414834359425e-05,
+      "loss": 3.1231,
+      "step": 2278
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.229551451187336e-05,
+      "loss": 2.9585,
+      "step": 2279
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.223688068015245e-05,
+      "loss": 3.1243,
+      "step": 2280
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.217824684843155e-05,
+      "loss": 2.92,
+      "step": 2281
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.211961301671064e-05,
+      "loss": 2.9059,
+      "step": 2282
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.206097918498974e-05,
+      "loss": 3.1515,
+      "step": 2283
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.200234535326884e-05,
+      "loss": 3.0166,
+      "step": 2284
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.194371152154794e-05,
+      "loss": 3.1146,
+      "step": 2285
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.188507768982703e-05,
+      "loss": 2.9681,
+      "step": 2286
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.182644385810613e-05,
+      "loss": 2.9228,
+      "step": 2287
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.176781002638523e-05,
+      "loss": 3.0504,
+      "step": 2288
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 3.2625436782836914,
+      "eval_runtime": 2940.0782,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 2288
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.170917619466433e-05,
+      "loss": 2.9851,
+      "step": 2289
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.165054236294342e-05,
+      "loss": 2.9923,
+      "step": 2290
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.159190853122252e-05,
+      "loss": 2.9812,
+      "step": 2291
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.153327469950162e-05,
+      "loss": 2.928,
+      "step": 2292
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.147464086778072e-05,
+      "loss": 3.0387,
+      "step": 2293
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.141600703605981e-05,
+      "loss": 3.0022,
+      "step": 2294
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.135737320433891e-05,
+      "loss": 3.096,
+      "step": 2295
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.1298739372618e-05,
+      "loss": 3.0206,
+      "step": 2296
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.124010554089711e-05,
+      "loss": 3.0001,
+      "step": 2297
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.118147170917619e-05,
+      "loss": 2.9541,
+      "step": 2298
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.11228378774553e-05,
+      "loss": 3.0121,
+      "step": 2299
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.106420404573439e-05,
+      "loss": 2.9971,
+      "step": 2300
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.100557021401349e-05,
+      "loss": 3.0368,
+      "step": 2301
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.094693638229258e-05,
+      "loss": 2.9971,
+      "step": 2302
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.088830255057169e-05,
+      "loss": 2.9714,
+      "step": 2303
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.082966871885078e-05,
+      "loss": 2.9774,
+      "step": 2304
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.077103488712988e-05,
+      "loss": 2.9699,
+      "step": 2305
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.071240105540897e-05,
+      "loss": 3.0251,
+      "step": 2306
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.065376722368807e-05,
+      "loss": 3.0374,
+      "step": 2307
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.059513339196717e-05,
+      "loss": 3.0141,
+      "step": 2308
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.053649956024626e-05,
+      "loss": 2.9218,
+      "step": 2309
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.047786572852536e-05,
+      "loss": 2.9611,
+      "step": 2310
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.041923189680446e-05,
+      "loss": 3.0331,
+      "step": 2311
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.036059806508357e-05,
+      "loss": 3.0187,
+      "step": 2312
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.030196423336265e-05,
+      "loss": 2.9778,
+      "step": 2313
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.024333040164175e-05,
+      "loss": 3.031,
+      "step": 2314
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.018469656992085e-05,
+      "loss": 3.0012,
+      "step": 2315
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.012606273819994e-05,
+      "loss": 2.9672,
+      "step": 2316
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.006742890647904e-05,
+      "loss": 2.8362,
+      "step": 2317
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.000879507475813e-05,
+      "loss": 3.0038,
+      "step": 2318
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.995016124303724e-05,
+      "loss": 2.9375,
+      "step": 2319
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.989152741131633e-05,
+      "loss": 3.0321,
+      "step": 2320
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.983289357959543e-05,
+      "loss": 3.0003,
+      "step": 2321
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.977425974787452e-05,
+      "loss": 2.9768,
+      "step": 2322
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.971562591615363e-05,
+      "loss": 3.0315,
+      "step": 2323
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.965699208443272e-05,
+      "loss": 3.0134,
+      "step": 2324
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.959835825271182e-05,
+      "loss": 3.0509,
+      "step": 2325
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.953972442099091e-05,
+      "loss": 3.008,
+      "step": 2326
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.948109058927002e-05,
+      "loss": 2.9663,
+      "step": 2327
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.942245675754911e-05,
+      "loss": 2.8545,
+      "step": 2328
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.936382292582821e-05,
+      "loss": 2.982,
+      "step": 2329
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.93051890941073e-05,
+      "loss": 3.0049,
+      "step": 2330
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.92465552623864e-05,
+      "loss": 3.0179,
+      "step": 2331
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.918792143066549e-05,
+      "loss": 2.9824,
+      "step": 2332
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.912928759894459e-05,
+      "loss": 2.9842,
+      "step": 2333
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.90706537672237e-05,
+      "loss": 3.0041,
+      "step": 2334
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.901201993550279e-05,
+      "loss": 2.9466,
+      "step": 2335
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.895338610378188e-05,
+      "loss": 3.0105,
+      "step": 2336
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.889475227206098e-05,
+      "loss": 2.9508,
+      "step": 2337
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.883611844034009e-05,
+      "loss": 2.911,
+      "step": 2338
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.877748460861918e-05,
+      "loss": 3.0504,
+      "step": 2339
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.871885077689827e-05,
+      "loss": 3.1059,
+      "step": 2340
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.866021694517737e-05,
+      "loss": 3.1035,
+      "step": 2341
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.860158311345646e-05,
+      "loss": 3.0156,
+      "step": 2342
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.854294928173557e-05,
+      "loss": 3.033,
+      "step": 2343
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.848431545001465e-05,
+      "loss": 3.0288,
+      "step": 2344
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.842568161829376e-05,
+      "loss": 2.9321,
+      "step": 2345
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.836704778657285e-05,
+      "loss": 3.012,
+      "step": 2346
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.830841395485196e-05,
+      "loss": 2.9936,
+      "step": 2347
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.824978012313104e-05,
+      "loss": 2.868,
+      "step": 2348
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.819114629141015e-05,
+      "loss": 3.0964,
+      "step": 2349
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.813251245968924e-05,
+      "loss": 3.035,
+      "step": 2350
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.807387862796835e-05,
+      "loss": 3.0535,
+      "step": 2351
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.801524479624743e-05,
+      "loss": 3.0401,
+      "step": 2352
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.795661096452654e-05,
+      "loss": 3.0324,
+      "step": 2353
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.789797713280563e-05,
+      "loss": 3.0138,
+      "step": 2354
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.783934330108473e-05,
+      "loss": 3.0568,
+      "step": 2355
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.778070946936382e-05,
+      "loss": 3.1124,
+      "step": 2356
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.772207563764292e-05,
+      "loss": 2.9387,
+      "step": 2357
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.766344180592203e-05,
+      "loss": 3.0234,
+      "step": 2358
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.76048079742011e-05,
+      "loss": 2.8785,
+      "step": 2359
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.754617414248021e-05,
+      "loss": 3.0862,
+      "step": 2360
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.748754031075931e-05,
+      "loss": 2.997,
+      "step": 2361
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.742890647903842e-05,
+      "loss": 2.9434,
+      "step": 2362
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.73702726473175e-05,
+      "loss": 3.0576,
+      "step": 2363
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.73116388155966e-05,
+      "loss": 2.9594,
+      "step": 2364
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.72530049838757e-05,
+      "loss": 3.0256,
+      "step": 2365
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.71943711521548e-05,
+      "loss": 3.0172,
+      "step": 2366
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.713573732043389e-05,
+      "loss": 3.0303,
+      "step": 2367
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.707710348871298e-05,
+      "loss": 3.0042,
+      "step": 2368
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.701846965699209e-05,
+      "loss": 3.0087,
+      "step": 2369
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.695983582527118e-05,
+      "loss": 3.0963,
+      "step": 2370
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.690120199355028e-05,
+      "loss": 2.9385,
+      "step": 2371
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.684256816182937e-05,
+      "loss": 3.026,
+      "step": 2372
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.678393433010848e-05,
+      "loss": 2.9379,
+      "step": 2373
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.672530049838758e-05,
+      "loss": 3.0521,
+      "step": 2374
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 3.0137,
+      "step": 2375
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.660803283494576e-05,
+      "loss": 3.0204,
+      "step": 2376
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.654939900322487e-05,
+      "loss": 3.004,
+      "step": 2377
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.649076517150397e-05,
+      "loss": 2.9018,
+      "step": 2378
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.643213133978306e-05,
+      "loss": 2.8955,
+      "step": 2379
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.637349750806216e-05,
+      "loss": 3.0233,
+      "step": 2380
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.631486367634125e-05,
+      "loss": 3.008,
+      "step": 2381
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.625622984462036e-05,
+      "loss": 2.9618,
+      "step": 2382
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.619759601289944e-05,
+      "loss": 2.8988,
+      "step": 2383
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.613896218117855e-05,
+      "loss": 3.0136,
+      "step": 2384
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.608032834945764e-05,
+      "loss": 2.952,
+      "step": 2385
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.602169451773673e-05,
+      "loss": 2.9119,
+      "step": 2386
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.596306068601583e-05,
+      "loss": 2.9193,
+      "step": 2387
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.590442685429494e-05,
+      "loss": 3.0116,
+      "step": 2388
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.584579302257403e-05,
+      "loss": 2.9762,
+      "step": 2389
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.578715919085313e-05,
+      "loss": 3.0246,
+      "step": 2390
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.572852535913222e-05,
+      "loss": 3.1,
+      "step": 2391
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.566989152741131e-05,
+      "loss": 2.9515,
+      "step": 2392
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.561125769569042e-05,
+      "loss": 3.0162,
+      "step": 2393
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.55526238639695e-05,
+      "loss": 2.9888,
+      "step": 2394
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.549399003224861e-05,
+      "loss": 2.9035,
+      "step": 2395
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.54353562005277e-05,
+      "loss": 2.9091,
+      "step": 2396
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.537672236880681e-05,
+      "loss": 2.9632,
+      "step": 2397
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.53180885370859e-05,
+      "loss": 2.9739,
+      "step": 2398
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.5259454705365e-05,
+      "loss": 3.0176,
+      "step": 2399
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.52008208736441e-05,
+      "loss": 3.131,
+      "step": 2400
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.51421870419232e-05,
+      "loss": 2.9436,
+      "step": 2401
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.508355321020228e-05,
+      "loss": 2.9454,
+      "step": 2402
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.502491937848139e-05,
+      "loss": 2.8289,
+      "step": 2403
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.496628554676049e-05,
+      "loss": 3.1042,
+      "step": 2404
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.490765171503958e-05,
+      "loss": 2.8591,
+      "step": 2405
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.484901788331868e-05,
+      "loss": 3.0201,
+      "step": 2406
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.479038405159777e-05,
+      "loss": 2.9499,
+      "step": 2407
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.473175021987688e-05,
+      "loss": 2.9972,
+      "step": 2408
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.467311638815597e-05,
+      "loss": 3.1087,
+      "step": 2409
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.461448255643507e-05,
+      "loss": 3.0046,
+      "step": 2410
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.455584872471416e-05,
+      "loss": 3.0949,
+      "step": 2411
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.449721489299327e-05,
+      "loss": 3.0089,
+      "step": 2412
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.443858106127235e-05,
+      "loss": 3.0895,
+      "step": 2413
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.437994722955146e-05,
+      "loss": 3.0326,
+      "step": 2414
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.432131339783055e-05,
+      "loss": 2.9201,
+      "step": 2415
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.426267956610965e-05,
+      "loss": 2.9691,
+      "step": 2416
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.420404573438874e-05,
+      "loss": 2.9328,
+      "step": 2417
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.414541190266783e-05,
+      "loss": 2.9624,
+      "step": 2418
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.408677807094694e-05,
+      "loss": 2.9547,
+      "step": 2419
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.402814423922604e-05,
+      "loss": 2.9956,
+      "step": 2420
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.396951040750513e-05,
+      "loss": 2.9702,
+      "step": 2421
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.391087657578422e-05,
+      "loss": 3.004,
+      "step": 2422
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.385224274406333e-05,
+      "loss": 2.9308,
+      "step": 2423
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.379360891234243e-05,
+      "loss": 2.9878,
+      "step": 2424
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.373497508062152e-05,
+      "loss": 3.0244,
+      "step": 2425
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.367634124890062e-05,
+      "loss": 2.9162,
+      "step": 2426
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.361770741717972e-05,
+      "loss": 2.9459,
+      "step": 2427
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.355907358545882e-05,
+      "loss": 2.9241,
+      "step": 2428
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.350043975373791e-05,
+      "loss": 2.9893,
+      "step": 2429
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.3441805922017e-05,
+      "loss": 3.0918,
+      "step": 2430
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.33831720902961e-05,
+      "loss": 2.9185,
+      "step": 2431
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.332453825857521e-05,
+      "loss": 3.0776,
+      "step": 2432
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.326590442685429e-05,
+      "loss": 3.0014,
+      "step": 2433
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.32072705951334e-05,
+      "loss": 2.9948,
+      "step": 2434
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.314863676341249e-05,
+      "loss": 2.8847,
+      "step": 2435
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.30900029316916e-05,
+      "loss": 2.9817,
+      "step": 2436
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.303136909997068e-05,
+      "loss": 2.8848,
+      "step": 2437
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.297273526824979e-05,
+      "loss": 3.1049,
+      "step": 2438
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.291410143652888e-05,
+      "loss": 3.028,
+      "step": 2439
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.285546760480798e-05,
+      "loss": 3.0241,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.279683377308707e-05,
+      "loss": 2.9669,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.273819994136617e-05,
+      "loss": 2.9753,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.267956610964527e-05,
+      "loss": 3.0769,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.262093227792435e-05,
+      "loss": 3.095,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.256229844620346e-05,
+      "loss": 3.0358,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.250366461448256e-05,
+      "loss": 3.009,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.244503078276166e-05,
+      "loss": 2.9533,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.238639695104074e-05,
+      "loss": 2.9614,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.232776311931985e-05,
+      "loss": 2.9125,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.226912928759895e-05,
+      "loss": 3.0652,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.221049545587805e-05,
+      "loss": 3.0921,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.215186162415714e-05,
+      "loss": 2.9856,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.209322779243624e-05,
+      "loss": 2.9609,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.203459396071534e-05,
+      "loss": 3.035,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.197596012899443e-05,
+      "loss": 2.9847,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.191732629727353e-05,
+      "loss": 3.0316,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.185869246555262e-05,
+      "loss": 2.9794,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.180005863383173e-05,
+      "loss": 2.9552,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.174142480211082e-05,
+      "loss": 2.9213,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.168279097038992e-05,
+      "loss": 3.0125,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.162415713866901e-05,
+      "loss": 2.9652,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.156552330694812e-05,
+      "loss": 3.0152,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.150688947522721e-05,
+      "loss": 3.0937,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.144825564350631e-05,
+      "loss": 2.9314,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 3.2398521900177,
+      "eval_runtime": 2939.9576,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.13896218117854e-05,
+      "loss": 2.972,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.13309879800645e-05,
+      "loss": 3.0006,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.12723541483436e-05,
+      "loss": 2.9709,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.121372031662269e-05,
+      "loss": 3.0649,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.11550864849018e-05,
+      "loss": 2.8472,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.109645265318089e-05,
+      "loss": 2.9919,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.103781882145998e-05,
+      "loss": 2.9293,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.097918498973908e-05,
+      "loss": 2.9208,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.092055115801818e-05,
+      "loss": 2.9485,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.086191732629728e-05,
+      "loss": 2.8891,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.0803283494576366e-05,
+      "loss": 2.9678,
+      "step": 2475
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.074464966285547e-05,
+      "loss": 2.9454,
+      "step": 2476
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.068601583113457e-05,
+      "loss": 2.9956,
+      "step": 2477
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.062738199941367e-05,
+      "loss": 2.9863,
+      "step": 2478
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.056874816769276e-05,
+      "loss": 2.9921,
+      "step": 2479
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.051011433597186e-05,
+      "loss": 3.0104,
+      "step": 2480
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.045148050425096e-05,
+      "loss": 2.8295,
+      "step": 2481
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.039284667253006e-05,
+      "loss": 2.97,
+      "step": 2482
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.033421284080915e-05,
+      "loss": 3.0534,
+      "step": 2483
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.027557900908825e-05,
+      "loss": 2.9503,
+      "step": 2484
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.021694517736734e-05,
+      "loss": 2.9844,
+      "step": 2485
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.0158311345646444e-05,
+      "loss": 2.8801,
+      "step": 2486
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.009967751392553e-05,
+      "loss": 2.9881,
+      "step": 2487
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.004104368220463e-05,
+      "loss": 2.8921,
+      "step": 2488
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.9982409850483734e-05,
+      "loss": 2.9311,
+      "step": 2489
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.9923776018762835e-05,
+      "loss": 2.9632,
+      "step": 2490
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.986514218704192e-05,
+      "loss": 3.0163,
+      "step": 2491
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.9806508355321024e-05,
+      "loss": 3.016,
+      "step": 2492
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.9747874523600125e-05,
+      "loss": 2.9402,
+      "step": 2493
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.968924069187922e-05,
+      "loss": 2.962,
+      "step": 2494
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.963060686015831e-05,
+      "loss": 2.9942,
+      "step": 2495
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.9571973028437414e-05,
+      "loss": 2.9702,
+      "step": 2496
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.951333919671651e-05,
+      "loss": 3.0159,
+      "step": 2497
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.94547053649956e-05,
+      "loss": 2.9585,
+      "step": 2498
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.93960715332747e-05,
+      "loss": 2.919,
+      "step": 2499
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.93374377015538e-05,
+      "loss": 2.9407,
+      "step": 2500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.92788038698329e-05,
+      "loss": 2.973,
+      "step": 2501
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.922017003811199e-05,
+      "loss": 2.9339,
+      "step": 2502
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.916153620639109e-05,
+      "loss": 3.0156,
+      "step": 2503
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.910290237467019e-05,
+      "loss": 2.9989,
+      "step": 2504
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.904426854294929e-05,
+      "loss": 2.9483,
+      "step": 2505
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.898563471122838e-05,
+      "loss": 2.9201,
+      "step": 2506
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.892700087950748e-05,
+      "loss": 3.0595,
+      "step": 2507
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.886836704778657e-05,
+      "loss": 3.0191,
+      "step": 2508
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.8809733216065674e-05,
+      "loss": 2.9978,
+      "step": 2509
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.875109938434477e-05,
+      "loss": 3.0281,
+      "step": 2510
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.869246555262386e-05,
+      "loss": 3.0379,
+      "step": 2511
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8633831720902964e-05,
+      "loss": 2.922,
+      "step": 2512
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8575197889182065e-05,
+      "loss": 2.9158,
+      "step": 2513
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.851656405746115e-05,
+      "loss": 3.0778,
+      "step": 2514
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8457930225740254e-05,
+      "loss": 3.0146,
+      "step": 2515
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8399296394019355e-05,
+      "loss": 2.9351,
+      "step": 2516
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8340662562298456e-05,
+      "loss": 2.8747,
+      "step": 2517
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8282028730577544e-05,
+      "loss": 2.9747,
+      "step": 2518
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8223394898856645e-05,
+      "loss": 3.0039,
+      "step": 2519
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.816476106713574e-05,
+      "loss": 2.9437,
+      "step": 2520
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.810612723541484e-05,
+      "loss": 2.8976,
+      "step": 2521
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.804749340369393e-05,
+      "loss": 2.9928,
+      "step": 2522
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.798885957197303e-05,
+      "loss": 2.9574,
+      "step": 2523
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.793022574025213e-05,
+      "loss": 2.9724,
+      "step": 2524
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.787159190853123e-05,
+      "loss": 2.9679,
+      "step": 2525
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.781295807681032e-05,
+      "loss": 3.0835,
+      "step": 2526
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.775432424508942e-05,
+      "loss": 3.0269,
+      "step": 2527
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.769569041336852e-05,
+      "loss": 2.991,
+      "step": 2528
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.763705658164761e-05,
+      "loss": 3.0316,
+      "step": 2529
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.757842274992671e-05,
+      "loss": 3.0771,
+      "step": 2530
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.751978891820581e-05,
+      "loss": 2.996,
+      "step": 2531
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7461155086484905e-05,
+      "loss": 3.0538,
+      "step": 2532
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7402521254764e-05,
+      "loss": 2.9899,
+      "step": 2533
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7343887423043093e-05,
+      "loss": 2.9144,
+      "step": 2534
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7285253591322195e-05,
+      "loss": 2.973,
+      "step": 2535
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7226619759601296e-05,
+      "loss": 2.9375,
+      "step": 2536
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.716798592788038e-05,
+      "loss": 3.0128,
+      "step": 2537
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7109352096159484e-05,
+      "loss": 3.0526,
+      "step": 2538
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7050718264438585e-05,
+      "loss": 2.9862,
+      "step": 2539
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6992084432717686e-05,
+      "loss": 2.9249,
+      "step": 2540
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6933450600996774e-05,
+      "loss": 2.9432,
+      "step": 2541
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6874816769275875e-05,
+      "loss": 2.945,
+      "step": 2542
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6816182937554976e-05,
+      "loss": 2.9054,
+      "step": 2543
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.675754910583407e-05,
+      "loss": 3.0417,
+      "step": 2544
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6698915274113165e-05,
+      "loss": 2.93,
+      "step": 2545
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.664028144239226e-05,
+      "loss": 2.9802,
+      "step": 2546
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.658164761067136e-05,
+      "loss": 2.9032,
+      "step": 2547
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.652301377895046e-05,
+      "loss": 3.063,
+      "step": 2548
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.646437994722955e-05,
+      "loss": 2.9094,
+      "step": 2549
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.640574611550865e-05,
+      "loss": 2.8761,
+      "step": 2550
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.634711228378775e-05,
+      "loss": 2.8374,
+      "step": 2551
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.628847845206685e-05,
+      "loss": 2.9685,
+      "step": 2552
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.622984462034594e-05,
+      "loss": 2.9527,
+      "step": 2553
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.617121078862504e-05,
+      "loss": 2.9442,
+      "step": 2554
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.611257695690414e-05,
+      "loss": 3.0894,
+      "step": 2555
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.605394312518323e-05,
+      "loss": 3.0,
+      "step": 2556
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.599530929346233e-05,
+      "loss": 2.9207,
+      "step": 2557
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5936675461741425e-05,
+      "loss": 2.918,
+      "step": 2558
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5878041630020526e-05,
+      "loss": 3.0187,
+      "step": 2559
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5819407798299613e-05,
+      "loss": 2.9735,
+      "step": 2560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5760773966578715e-05,
+      "loss": 3.0096,
+      "step": 2561
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5702140134857816e-05,
+      "loss": 2.9933,
+      "step": 2562
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.564350630313692e-05,
+      "loss": 2.9946,
+      "step": 2563
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5584872471416004e-05,
+      "loss": 3.0202,
+      "step": 2564
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5526238639695105e-05,
+      "loss": 2.8782,
+      "step": 2565
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5467604807974206e-05,
+      "loss": 3.0155,
+      "step": 2566
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.540897097625331e-05,
+      "loss": 3.0636,
+      "step": 2567
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5350337144532395e-05,
+      "loss": 3.0327,
+      "step": 2568
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5291703312811496e-05,
+      "loss": 2.9718,
+      "step": 2569
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.523306948109059e-05,
+      "loss": 2.9151,
+      "step": 2570
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.517443564936969e-05,
+      "loss": 3.0394,
+      "step": 2571
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.511580181764878e-05,
+      "loss": 2.9542,
+      "step": 2572
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.505716798592788e-05,
+      "loss": 2.9608,
+      "step": 2573
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.499853415420698e-05,
+      "loss": 2.901,
+      "step": 2574
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.493990032248608e-05,
+      "loss": 2.9036,
+      "step": 2575
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.488126649076517e-05,
+      "loss": 2.9425,
+      "step": 2576
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.482263265904427e-05,
+      "loss": 2.9666,
+      "step": 2577
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.476399882732337e-05,
+      "loss": 2.9727,
+      "step": 2578
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.470536499560247e-05,
+      "loss": 2.9093,
+      "step": 2579
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.464673116388156e-05,
+      "loss": 2.9456,
+      "step": 2580
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.458809733216066e-05,
+      "loss": 2.9856,
+      "step": 2581
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4529463500439756e-05,
+      "loss": 3.0115,
+      "step": 2582
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.447082966871885e-05,
+      "loss": 2.9719,
+      "step": 2583
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4412195836997945e-05,
+      "loss": 2.996,
+      "step": 2584
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4353562005277046e-05,
+      "loss": 3.067,
+      "step": 2585
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.429492817355615e-05,
+      "loss": 3.0018,
+      "step": 2586
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4236294341835235e-05,
+      "loss": 2.9288,
+      "step": 2587
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4177660510114336e-05,
+      "loss": 2.9967,
+      "step": 2588
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.411902667839344e-05,
+      "loss": 2.9905,
+      "step": 2589
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.406039284667254e-05,
+      "loss": 3.0039,
+      "step": 2590
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.4001759014951625e-05,
+      "loss": 2.9316,
+      "step": 2591
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.3943125183230727e-05,
+      "loss": 2.9999,
+      "step": 2592
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.388449135150983e-05,
+      "loss": 2.9992,
+      "step": 2593
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.382585751978892e-05,
+      "loss": 2.9757,
+      "step": 2594
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.3767223688068016e-05,
+      "loss": 2.8192,
+      "step": 2595
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.370858985634711e-05,
+      "loss": 2.8452,
+      "step": 2596
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.364995602462621e-05,
+      "loss": 2.8474,
+      "step": 2597
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.359132219290531e-05,
+      "loss": 2.8424,
+      "step": 2598
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.35326883611844e-05,
+      "loss": 3.046,
+      "step": 2599
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.34740545294635e-05,
+      "loss": 2.914,
+      "step": 2600
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.34154206977426e-05,
+      "loss": 2.9644,
+      "step": 2601
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.3356786866021704e-05,
+      "loss": 2.9215,
+      "step": 2602
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.329815303430079e-05,
+      "loss": 2.872,
+      "step": 2603
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.323951920257989e-05,
+      "loss": 2.9656,
+      "step": 2604
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.318088537085899e-05,
+      "loss": 2.9664,
+      "step": 2605
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.312225153913809e-05,
+      "loss": 2.8479,
+      "step": 2606
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.306361770741718e-05,
+      "loss": 2.9984,
+      "step": 2607
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.3004983875696276e-05,
+      "loss": 3.0044,
+      "step": 2608
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.294635004397538e-05,
+      "loss": 3.0522,
+      "step": 2609
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.2887716212254465e-05,
+      "loss": 2.9028,
+      "step": 2610
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.2829082380533566e-05,
+      "loss": 2.9666,
+      "step": 2611
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.277044854881267e-05,
+      "loss": 2.948,
+      "step": 2612
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.271181471709177e-05,
+      "loss": 2.9854,
+      "step": 2613
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.2653180885370856e-05,
+      "loss": 2.9549,
+      "step": 2614
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.259454705364996e-05,
+      "loss": 2.8702,
+      "step": 2615
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.253591322192906e-05,
+      "loss": 2.9028,
+      "step": 2616
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.247727939020816e-05,
+      "loss": 2.9593,
+      "step": 2617
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.2418645558487247e-05,
+      "loss": 3.0233,
+      "step": 2618
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.236001172676635e-05,
+      "loss": 2.9256,
+      "step": 2619
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.230137789504544e-05,
+      "loss": 2.9698,
+      "step": 2620
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.224274406332454e-05,
+      "loss": 2.8858,
+      "step": 2621
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.218411023160363e-05,
+      "loss": 2.9088,
+      "step": 2622
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.212547639988273e-05,
+      "loss": 2.9003,
+      "step": 2623
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.206684256816183e-05,
+      "loss": 2.9809,
+      "step": 2624
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.2008208736440934e-05,
+      "loss": 2.9083,
+      "step": 2625
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.194957490472002e-05,
+      "loss": 2.9803,
+      "step": 2626
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.189094107299912e-05,
+      "loss": 2.9718,
+      "step": 2627
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.1832307241278224e-05,
+      "loss": 2.9033,
+      "step": 2628
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.1773673409557325e-05,
+      "loss": 3.0781,
+      "step": 2629
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.171503957783641e-05,
+      "loss": 2.9367,
+      "step": 2630
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.165640574611551e-05,
+      "loss": 2.9469,
+      "step": 2631
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.159777191439461e-05,
+      "loss": 2.8456,
+      "step": 2632
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.153913808267371e-05,
+      "loss": 2.9042,
+      "step": 2633
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.1480504250952796e-05,
+      "loss": 2.9669,
+      "step": 2634
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.14218704192319e-05,
+      "loss": 2.9404,
+      "step": 2635
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.1363236587511e-05,
+      "loss": 2.9794,
+      "step": 2636
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.13046027557901e-05,
+      "loss": 3.0569,
+      "step": 2637
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.124596892406919e-05,
+      "loss": 2.9156,
+      "step": 2638
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.118733509234829e-05,
+      "loss": 3.0051,
+      "step": 2639
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.112870126062739e-05,
+      "loss": 2.991,
+      "step": 2640
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 3.2172892093658447,
+      "eval_runtime": 2940.2613,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 2640
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.107006742890648e-05,
+      "loss": 2.8543,
+      "step": 2641
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.101143359718558e-05,
+      "loss": 3.0516,
+      "step": 2642
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.095279976546468e-05,
+      "loss": 2.9691,
+      "step": 2643
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.089416593374377e-05,
+      "loss": 3.0091,
+      "step": 2644
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.083553210202287e-05,
+      "loss": 2.9718,
+      "step": 2645
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.077689827030196e-05,
+      "loss": 2.933,
+      "step": 2646
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.071826443858106e-05,
+      "loss": 2.9114,
+      "step": 2647
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.0659630606860164e-05,
+      "loss": 2.9301,
+      "step": 2648
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.060099677513925e-05,
+      "loss": 2.8686,
+      "step": 2649
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.054236294341835e-05,
+      "loss": 2.9059,
+      "step": 2650
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.0483729111697454e-05,
+      "loss": 3.0058,
+      "step": 2651
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.0425095279976555e-05,
+      "loss": 2.922,
+      "step": 2652
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.036646144825564e-05,
+      "loss": 2.9205,
+      "step": 2653
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.0307827616534744e-05,
+      "loss": 2.9853,
+      "step": 2654
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.0249193784813845e-05,
+      "loss": 2.9316,
+      "step": 2655
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.019055995309294e-05,
+      "loss": 3.0522,
+      "step": 2656
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.0131926121372033e-05,
+      "loss": 2.9679,
+      "step": 2657
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.007329228965113e-05,
+      "loss": 2.9912,
+      "step": 2658
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.001465845793023e-05,
+      "loss": 2.9458,
+      "step": 2659
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.995602462620932e-05,
+      "loss": 2.9799,
+      "step": 2660
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9897390794488424e-05,
+      "loss": 2.9888,
+      "step": 2661
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.983875696276752e-05,
+      "loss": 2.8084,
+      "step": 2662
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.978012313104662e-05,
+      "loss": 2.8867,
+      "step": 2663
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9721489299325714e-05,
+      "loss": 2.9993,
+      "step": 2664
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9662855467604815e-05,
+      "loss": 2.9978,
+      "step": 2665
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.960422163588391e-05,
+      "loss": 2.9658,
+      "step": 2666
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9545587804163004e-05,
+      "loss": 2.9173,
+      "step": 2667
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9486953972442105e-05,
+      "loss": 2.925,
+      "step": 2668
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.94283201407212e-05,
+      "loss": 2.919,
+      "step": 2669
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9369686309000293e-05,
+      "loss": 2.8328,
+      "step": 2670
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.931105247727939e-05,
+      "loss": 3.0428,
+      "step": 2671
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.925241864555849e-05,
+      "loss": 3.0001,
+      "step": 2672
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.919378481383758e-05,
+      "loss": 2.9359,
+      "step": 2673
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.9135150982116684e-05,
+      "loss": 2.9077,
+      "step": 2674
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.907651715039578e-05,
+      "loss": 2.992,
+      "step": 2675
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.901788331867488e-05,
+      "loss": 2.9674,
+      "step": 2676
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8959249486953974e-05,
+      "loss": 2.8465,
+      "step": 2677
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8900615655233075e-05,
+      "loss": 2.8251,
+      "step": 2678
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.884198182351217e-05,
+      "loss": 2.9652,
+      "step": 2679
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.878334799179127e-05,
+      "loss": 2.9478,
+      "step": 2680
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8724714160070365e-05,
+      "loss": 3.0002,
+      "step": 2681
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.866608032834946e-05,
+      "loss": 3.0055,
+      "step": 2682
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8607446496628553e-05,
+      "loss": 2.9612,
+      "step": 2683
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8548812664907655e-05,
+      "loss": 3.0569,
+      "step": 2684
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.849017883318675e-05,
+      "loss": 2.9817,
+      "step": 2685
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.843154500146585e-05,
+      "loss": 2.8774,
+      "step": 2686
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.8372911169744944e-05,
+      "loss": 2.9759,
+      "step": 2687
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.8314277338024045e-05,
+      "loss": 3.0761,
+      "step": 2688
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.825564350630314e-05,
+      "loss": 2.9908,
+      "step": 2689
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.819700967458224e-05,
+      "loss": 2.9525,
+      "step": 2690
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.8138375842861335e-05,
+      "loss": 2.8244,
+      "step": 2691
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.8079742011140436e-05,
+      "loss": 2.8924,
+      "step": 2692
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.802110817941953e-05,
+      "loss": 2.9047,
+      "step": 2693
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7962474347698625e-05,
+      "loss": 3.007,
+      "step": 2694
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.790384051597772e-05,
+      "loss": 3.0086,
+      "step": 2695
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7845206684256813e-05,
+      "loss": 2.9496,
+      "step": 2696
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7786572852535915e-05,
+      "loss": 2.9391,
+      "step": 2697
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.772793902081501e-05,
+      "loss": 2.9309,
+      "step": 2698
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.766930518909411e-05,
+      "loss": 3.0011,
+      "step": 2699
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7610671357373204e-05,
+      "loss": 2.9296,
+      "step": 2700
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7552037525652305e-05,
+      "loss": 2.913,
+      "step": 2701
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.74934036939314e-05,
+      "loss": 2.827,
+      "step": 2702
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.74347698622105e-05,
+      "loss": 2.978,
+      "step": 2703
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7376136030489595e-05,
+      "loss": 3.0295,
+      "step": 2704
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7317502198768696e-05,
+      "loss": 3.0656,
+      "step": 2705
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.725886836704779e-05,
+      "loss": 2.8391,
+      "step": 2706
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7200234535326885e-05,
+      "loss": 2.9854,
+      "step": 2707
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.714160070360598e-05,
+      "loss": 2.8733,
+      "step": 2708
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.708296687188508e-05,
+      "loss": 2.8921,
+      "step": 2709
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.7024333040164175e-05,
+      "loss": 2.8333,
+      "step": 2710
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6965699208443276e-05,
+      "loss": 2.9524,
+      "step": 2711
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.690706537672237e-05,
+      "loss": 2.9885,
+      "step": 2712
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.684843154500147e-05,
+      "loss": 3.07,
+      "step": 2713
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6789797713280565e-05,
+      "loss": 2.9899,
+      "step": 2714
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6731163881559667e-05,
+      "loss": 2.9843,
+      "step": 2715
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.667253004983876e-05,
+      "loss": 3.0218,
+      "step": 2716
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6613896218117855e-05,
+      "loss": 2.8883,
+      "step": 2717
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.655526238639695e-05,
+      "loss": 2.9983,
+      "step": 2718
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.649662855467605e-05,
+      "loss": 2.91,
+      "step": 2719
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6437994722955145e-05,
+      "loss": 2.9078,
+      "step": 2720
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.6379360891234246e-05,
+      "loss": 2.8622,
+      "step": 2721
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.632072705951334e-05,
+      "loss": 3.057,
+      "step": 2722
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.6262093227792435e-05,
+      "loss": 2.8854,
+      "step": 2723
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.6203459396071536e-05,
+      "loss": 2.9078,
+      "step": 2724
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.614482556435063e-05,
+      "loss": 2.9365,
+      "step": 2725
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.608619173262973e-05,
+      "loss": 2.9439,
+      "step": 2726
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.6027557900908825e-05,
+      "loss": 2.9007,
+      "step": 2727
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5968924069187927e-05,
+      "loss": 2.8987,
+      "step": 2728
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.591029023746702e-05,
+      "loss": 2.8122,
+      "step": 2729
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5851656405746115e-05,
+      "loss": 2.9194,
+      "step": 2730
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.579302257402521e-05,
+      "loss": 2.9381,
+      "step": 2731
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.573438874230431e-05,
+      "loss": 2.9534,
+      "step": 2732
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5675754910583405e-05,
+      "loss": 2.7799,
+      "step": 2733
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5617121078862506e-05,
+      "loss": 2.9023,
+      "step": 2734
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.55584872471416e-05,
+      "loss": 2.9027,
+      "step": 2735
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.54998534154207e-05,
+      "loss": 2.9513,
+      "step": 2736
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5441219583699796e-05,
+      "loss": 2.8047,
+      "step": 2737
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.53825857519789e-05,
+      "loss": 3.0478,
+      "step": 2738
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.532395192025799e-05,
+      "loss": 2.9872,
+      "step": 2739
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.526531808853709e-05,
+      "loss": 2.9807,
+      "step": 2740
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5206684256816187e-05,
+      "loss": 2.9503,
+      "step": 2741
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.514805042509528e-05,
+      "loss": 2.9607,
+      "step": 2742
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5089416593374375e-05,
+      "loss": 3.0141,
+      "step": 2743
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.5030782761653476e-05,
+      "loss": 2.8885,
+      "step": 2744
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.497214892993257e-05,
+      "loss": 3.0052,
+      "step": 2745
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.491351509821167e-05,
+      "loss": 2.9363,
+      "step": 2746
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4854881266490766e-05,
+      "loss": 3.0528,
+      "step": 2747
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.479624743476987e-05,
+      "loss": 2.9836,
+      "step": 2748
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.473761360304896e-05,
+      "loss": 2.9265,
+      "step": 2749
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.467897977132806e-05,
+      "loss": 2.9269,
+      "step": 2750
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.462034593960716e-05,
+      "loss": 2.933,
+      "step": 2751
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.456171210788625e-05,
+      "loss": 2.9605,
+      "step": 2752
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.450307827616535e-05,
+      "loss": 2.9411,
+      "step": 2753
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 2.9598,
+      "step": 2754
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.438581061272354e-05,
+      "loss": 2.9167,
+      "step": 2755
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4327176781002635e-05,
+      "loss": 2.9279,
+      "step": 2756
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4268542949281736e-05,
+      "loss": 2.9237,
+      "step": 2757
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.420990911756083e-05,
+      "loss": 2.903,
+      "step": 2758
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.415127528583993e-05,
+      "loss": 2.841,
+      "step": 2759
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4092641454119026e-05,
+      "loss": 2.9219,
+      "step": 2760
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.403400762239813e-05,
+      "loss": 2.9157,
+      "step": 2761
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.397537379067722e-05,
+      "loss": 2.8828,
+      "step": 2762
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.391673995895632e-05,
+      "loss": 2.9777,
+      "step": 2763
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.385810612723542e-05,
+      "loss": 2.9269,
+      "step": 2764
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.379947229551452e-05,
+      "loss": 2.9936,
+      "step": 2765
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.374083846379361e-05,
+      "loss": 2.901,
+      "step": 2766
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3682204632072707e-05,
+      "loss": 2.8971,
+      "step": 2767
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.36235708003518e-05,
+      "loss": 2.9404,
+      "step": 2768
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.35649369686309e-05,
+      "loss": 2.9661,
+      "step": 2769
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3506303136909996e-05,
+      "loss": 2.8264,
+      "step": 2770
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.34476693051891e-05,
+      "loss": 2.977,
+      "step": 2771
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.338903547346819e-05,
+      "loss": 2.8653,
+      "step": 2772
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.333040164174729e-05,
+      "loss": 2.9227,
+      "step": 2773
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.327176781002639e-05,
+      "loss": 2.9449,
+      "step": 2774
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.321313397830549e-05,
+      "loss": 2.9402,
+      "step": 2775
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.315450014658458e-05,
+      "loss": 2.8669,
+      "step": 2776
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3095866314863684e-05,
+      "loss": 2.971,
+      "step": 2777
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.303723248314278e-05,
+      "loss": 2.8939,
+      "step": 2778
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.297859865142187e-05,
+      "loss": 2.9303,
+      "step": 2779
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.291996481970097e-05,
+      "loss": 2.9258,
+      "step": 2780
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.286133098798006e-05,
+      "loss": 2.8606,
+      "step": 2781
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.280269715625916e-05,
+      "loss": 2.7927,
+      "step": 2782
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2744063324538256e-05,
+      "loss": 2.9105,
+      "step": 2783
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.268542949281736e-05,
+      "loss": 2.9119,
+      "step": 2784
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.262679566109645e-05,
+      "loss": 3.0501,
+      "step": 2785
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.256816182937555e-05,
+      "loss": 2.9486,
+      "step": 2786
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.250952799765465e-05,
+      "loss": 2.9263,
+      "step": 2787
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.245089416593375e-05,
+      "loss": 2.8871,
+      "step": 2788
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.239226033421284e-05,
+      "loss": 2.9066,
+      "step": 2789
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2333626502491944e-05,
+      "loss": 2.956,
+      "step": 2790
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.227499267077104e-05,
+      "loss": 3.0425,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.221635883905013e-05,
+      "loss": 2.9842,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.215772500732923e-05,
+      "loss": 2.9224,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.209909117560833e-05,
+      "loss": 2.8652,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.204045734388742e-05,
+      "loss": 2.9478,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.198182351216652e-05,
+      "loss": 2.9776,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.192318968044562e-05,
+      "loss": 2.9747,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.186455584872472e-05,
+      "loss": 2.9551,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.180592201700381e-05,
+      "loss": 2.8735,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1747288185282914e-05,
+      "loss": 2.8173,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.168865435356201e-05,
+      "loss": 2.9245,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.163002052184111e-05,
+      "loss": 2.9679,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1571386690120204e-05,
+      "loss": 2.9496,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.15127528583993e-05,
+      "loss": 2.9174,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.145411902667839e-05,
+      "loss": 2.9739,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1395485194957493e-05,
+      "loss": 2.9363,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.133685136323659e-05,
+      "loss": 2.9018,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.127821753151568e-05,
+      "loss": 2.9554,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.121958369979478e-05,
+      "loss": 2.89,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.116094986807388e-05,
+      "loss": 3.0036,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.110231603635298e-05,
+      "loss": 2.9772,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.104368220463207e-05,
+      "loss": 2.9327,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.0985048372911174e-05,
+      "loss": 2.9426,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.092641454119027e-05,
+      "loss": 2.9169,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.086778070946937e-05,
+      "loss": 2.9826,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.0809146877748464e-05,
+      "loss": 2.9663,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 3.1972124576568604,
+      "eval_runtime": 2940.1891,
+      "eval_samples_per_second": 6.968,
+      "eval_steps_per_second": 2.323,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.075051304602756e-05,
+      "loss": 2.9791,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.069187921430665e-05,
+      "loss": 2.9606,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.0633245382585753e-05,
+      "loss": 2.953,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.057461155086485e-05,
+      "loss": 2.9999,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.051597771914395e-05,
+      "loss": 2.7495,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.045734388742304e-05,
+      "loss": 2.9623,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.0398710055702144e-05,
+      "loss": 2.9181,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.034007622398124e-05,
+      "loss": 2.9388,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.028144239226034e-05,
+      "loss": 2.8328,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.0222808560539434e-05,
+      "loss": 2.9606,
+      "step": 2826
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.0164174728818535e-05,
+      "loss": 2.984,
+      "step": 2827
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.010554089709763e-05,
+      "loss": 2.9417,
+      "step": 2828
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.0046907065376724e-05,
+      "loss": 2.8587,
+      "step": 2829
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.998827323365582e-05,
+      "loss": 2.9365,
+      "step": 2830
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.992963940193492e-05,
+      "loss": 2.9864,
+      "step": 2831
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9871005570214013e-05,
+      "loss": 2.9776,
+      "step": 2832
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9812371738493115e-05,
+      "loss": 2.9485,
+      "step": 2833
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.975373790677221e-05,
+      "loss": 2.9256,
+      "step": 2834
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.969510407505131e-05,
+      "loss": 2.9352,
+      "step": 2835
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9636470243330404e-05,
+      "loss": 2.9328,
+      "step": 2836
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.95778364116095e-05,
+      "loss": 2.9756,
+      "step": 2837
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.95192025798886e-05,
+      "loss": 2.875,
+      "step": 2838
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9460568748167694e-05,
+      "loss": 2.9342,
+      "step": 2839
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9401934916446795e-05,
+      "loss": 2.908,
+      "step": 2840
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.934330108472589e-05,
+      "loss": 3.0439,
+      "step": 2841
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9284667253004984e-05,
+      "loss": 3.0013,
+      "step": 2842
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.922603342128408e-05,
+      "loss": 2.9613,
+      "step": 2843
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.916739958956318e-05,
+      "loss": 2.9669,
+      "step": 2844
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9108765757842274e-05,
+      "loss": 2.9951,
+      "step": 2845
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9050131926121375e-05,
+      "loss": 2.9353,
+      "step": 2846
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.899149809440047e-05,
+      "loss": 2.9397,
+      "step": 2847
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.893286426267957e-05,
+      "loss": 2.927,
+      "step": 2848
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8874230430958664e-05,
+      "loss": 2.8031,
+      "step": 2849
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8815596599237765e-05,
+      "loss": 2.9521,
+      "step": 2850
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.875696276751686e-05,
+      "loss": 2.8733,
+      "step": 2851
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.869832893579596e-05,
+      "loss": 2.9939,
+      "step": 2852
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8639695104075055e-05,
+      "loss": 2.866,
+      "step": 2853
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.858106127235415e-05,
+      "loss": 2.896,
+      "step": 2854
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8522427440633244e-05,
+      "loss": 3.0004,
+      "step": 2855
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8463793608912345e-05,
+      "loss": 2.895,
+      "step": 2856
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.840515977719144e-05,
+      "loss": 2.9782,
+      "step": 2857
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.834652594547054e-05,
+      "loss": 2.896,
+      "step": 2858
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8287892113749635e-05,
+      "loss": 2.9681,
+      "step": 2859
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8229258282028736e-05,
+      "loss": 2.9011,
+      "step": 2860
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.817062445030783e-05,
+      "loss": 3.0072,
+      "step": 2861
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.811199061858693e-05,
+      "loss": 2.8798,
+      "step": 2862
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8053356786866025e-05,
+      "loss": 2.8416,
+      "step": 2863
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.799472295514512e-05,
+      "loss": 2.8416,
+      "step": 2864
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.793608912342422e-05,
+      "loss": 2.9927,
+      "step": 2865
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7877455291703315e-05,
+      "loss": 2.8251,
+      "step": 2866
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.781882145998241e-05,
+      "loss": 2.8113,
+      "step": 2867
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7760187628261504e-05,
+      "loss": 2.8884,
+      "step": 2868
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7701553796540605e-05,
+      "loss": 2.9539,
+      "step": 2869
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.76429199648197e-05,
+      "loss": 2.9121,
+      "step": 2870
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.75842861330988e-05,
+      "loss": 2.9654,
+      "step": 2871
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7525652301377895e-05,
+      "loss": 2.9765,
+      "step": 2872
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7467018469656996e-05,
+      "loss": 2.7985,
+      "step": 2873
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.740838463793609e-05,
+      "loss": 2.9105,
+      "step": 2874
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.734975080621519e-05,
+      "loss": 2.9365,
+      "step": 2875
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7291116974494285e-05,
+      "loss": 2.9142,
+      "step": 2876
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7232483142773387e-05,
+      "loss": 3.0456,
+      "step": 2877
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.717384931105248e-05,
+      "loss": 2.9197,
+      "step": 2878
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7115215479331575e-05,
+      "loss": 2.8813,
+      "step": 2879
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.705658164761067e-05,
+      "loss": 2.909,
+      "step": 2880
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.699794781588977e-05,
+      "loss": 2.7869,
+      "step": 2881
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6939313984168865e-05,
+      "loss": 2.9119,
+      "step": 2882
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6880680152447966e-05,
+      "loss": 2.9985,
+      "step": 2883
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.682204632072706e-05,
+      "loss": 2.949,
+      "step": 2884
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.676341248900616e-05,
+      "loss": 2.9885,
+      "step": 2885
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6704778657285256e-05,
+      "loss": 2.9018,
+      "step": 2886
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.664614482556436e-05,
+      "loss": 3.0304,
+      "step": 2887
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.658751099384345e-05,
+      "loss": 2.8653,
+      "step": 2888
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.652887716212255e-05,
+      "loss": 3.0575,
+      "step": 2889
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6470243330401647e-05,
+      "loss": 2.9043,
+      "step": 2890
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.641160949868074e-05,
+      "loss": 3.0285,
+      "step": 2891
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6352975666959835e-05,
+      "loss": 3.032,
+      "step": 2892
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.629434183523893e-05,
+      "loss": 3.0223,
+      "step": 2893
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.623570800351803e-05,
+      "loss": 2.9317,
+      "step": 2894
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6177074171797125e-05,
+      "loss": 2.9454,
+      "step": 2895
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6118440340076226e-05,
+      "loss": 2.9658,
+      "step": 2896
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.605980650835532e-05,
+      "loss": 2.9197,
+      "step": 2897
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.600117267663442e-05,
+      "loss": 3.0083,
+      "step": 2898
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5942538844913516e-05,
+      "loss": 2.9619,
+      "step": 2899
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.588390501319262e-05,
+      "loss": 2.9261,
+      "step": 2900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.582527118147171e-05,
+      "loss": 2.9368,
+      "step": 2901
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.576663734975081e-05,
+      "loss": 2.9306,
+      "step": 2902
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5708003518029907e-05,
+      "loss": 2.9412,
+      "step": 2903
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5649369686309e-05,
+      "loss": 2.902,
+      "step": 2904
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5590735854588095e-05,
+      "loss": 2.8198,
+      "step": 2905
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5532102022867196e-05,
+      "loss": 2.8804,
+      "step": 2906
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.547346819114629e-05,
+      "loss": 2.8733,
+      "step": 2907
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.541483435942539e-05,
+      "loss": 2.8826,
+      "step": 2908
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5356200527704486e-05,
+      "loss": 2.8929,
+      "step": 2909
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.529756669598359e-05,
+      "loss": 2.8247,
+      "step": 2910
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.523893286426268e-05,
+      "loss": 2.802,
+      "step": 2911
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.518029903254178e-05,
+      "loss": 3.0505,
+      "step": 2912
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.512166520082088e-05,
+      "loss": 2.9077,
+      "step": 2913
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.506303136909997e-05,
+      "loss": 2.9342,
+      "step": 2914
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5004397537379066e-05,
+      "loss": 2.9597,
+      "step": 2915
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.494576370565817e-05,
+      "loss": 2.9522,
+      "step": 2916
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.488712987393726e-05,
+      "loss": 2.8228,
+      "step": 2917
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.482849604221636e-05,
+      "loss": 2.9209,
+      "step": 2918
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4769862210495456e-05,
+      "loss": 2.9564,
+      "step": 2919
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.471122837877456e-05,
+      "loss": 2.9294,
+      "step": 2920
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.465259454705365e-05,
+      "loss": 2.9424,
+      "step": 2921
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4593960715332746e-05,
+      "loss": 2.9069,
+      "step": 2922
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.453532688361185e-05,
+      "loss": 2.9826,
+      "step": 2923
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.447669305189094e-05,
+      "loss": 2.8924,
+      "step": 2924
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.441805922017004e-05,
+      "loss": 2.8812,
+      "step": 2925
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.435942538844914e-05,
+      "loss": 2.8904,
+      "step": 2926
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.430079155672823e-05,
+      "loss": 2.9383,
+      "step": 2927
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4242157725007326e-05,
+      "loss": 2.8946,
+      "step": 2928
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.418352389328643e-05,
+      "loss": 2.9094,
+      "step": 2929
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.412489006156552e-05,
+      "loss": 2.8616,
+      "step": 2930
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.406625622984462e-05,
+      "loss": 2.9005,
+      "step": 2931
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4007622398123716e-05,
+      "loss": 3.0549,
+      "step": 2932
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.394898856640282e-05,
+      "loss": 2.8597,
+      "step": 2933
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.389035473468191e-05,
+      "loss": 2.8892,
+      "step": 2934
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.383172090296101e-05,
+      "loss": 2.9433,
+      "step": 2935
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.377308707124011e-05,
+      "loss": 2.7032,
+      "step": 2936
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.371445323951921e-05,
+      "loss": 2.9428,
+      "step": 2937
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.36558194077983e-05,
+      "loss": 2.9722,
+      "step": 2938
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.35971855760774e-05,
+      "loss": 2.8409,
+      "step": 2939
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.353855174435649e-05,
+      "loss": 2.9067,
+      "step": 2940
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.347991791263559e-05,
+      "loss": 2.6999,
+      "step": 2941
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.342128408091469e-05,
+      "loss": 2.8469,
+      "step": 2942
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.336265024919379e-05,
+      "loss": 2.9083,
+      "step": 2943
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.330401641747288e-05,
+      "loss": 2.9588,
+      "step": 2944
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.324538258575198e-05,
+      "loss": 2.8951,
+      "step": 2945
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.318674875403108e-05,
+      "loss": 2.898,
+      "step": 2946
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.312811492231018e-05,
+      "loss": 3.0032,
+      "step": 2947
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.306948109058927e-05,
+      "loss": 2.8667,
+      "step": 2948
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.301084725886837e-05,
+      "loss": 2.911,
+      "step": 2949
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.295221342714747e-05,
+      "loss": 2.9187,
+      "step": 2950
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.289357959542656e-05,
+      "loss": 2.9395,
+      "step": 2951
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.283494576370566e-05,
+      "loss": 2.8935,
+      "step": 2952
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.277631193198475e-05,
+      "loss": 2.9818,
+      "step": 2953
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.271767810026385e-05,
+      "loss": 2.9324,
+      "step": 2954
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.265904426854295e-05,
+      "loss": 2.9162,
+      "step": 2955
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.260041043682205e-05,
+      "loss": 2.9753,
+      "step": 2956
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.254177660510114e-05,
+      "loss": 2.9244,
+      "step": 2957
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.248314277338024e-05,
+      "loss": 2.9027,
+      "step": 2958
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.242450894165934e-05,
+      "loss": 2.8489,
+      "step": 2959
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.236587510993844e-05,
+      "loss": 2.9391,
+      "step": 2960
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.230724127821753e-05,
+      "loss": 2.8787,
+      "step": 2961
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2248607446496634e-05,
+      "loss": 2.8082,
+      "step": 2962
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.218997361477573e-05,
+      "loss": 2.8509,
+      "step": 2963
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.213133978305482e-05,
+      "loss": 2.8784,
+      "step": 2964
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.207270595133392e-05,
+      "loss": 2.9297,
+      "step": 2965
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.201407211961302e-05,
+      "loss": 2.8596,
+      "step": 2966
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.195543828789211e-05,
+      "loss": 2.8075,
+      "step": 2967
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1896804456171213e-05,
+      "loss": 2.9482,
+      "step": 2968
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.183817062445031e-05,
+      "loss": 2.8439,
+      "step": 2969
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.177953679272941e-05,
+      "loss": 2.9943,
+      "step": 2970
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.17209029610085e-05,
+      "loss": 2.8396,
+      "step": 2971
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1662269129287604e-05,
+      "loss": 2.7967,
+      "step": 2972
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.16036352975667e-05,
+      "loss": 2.9674,
+      "step": 2973
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.15450014658458e-05,
+      "loss": 2.8858,
+      "step": 2974
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1486367634124894e-05,
+      "loss": 2.8675,
+      "step": 2975
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.142773380240399e-05,
+      "loss": 2.9529,
+      "step": 2976
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.136909997068308e-05,
+      "loss": 2.8245,
+      "step": 2977
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.131046613896218e-05,
+      "loss": 2.9489,
+      "step": 2978
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.125183230724128e-05,
+      "loss": 2.8507,
+      "step": 2979
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.119319847552037e-05,
+      "loss": 2.8228,
+      "step": 2980
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1134564643799474e-05,
+      "loss": 2.9688,
+      "step": 2981
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.107593081207857e-05,
+      "loss": 2.8488,
+      "step": 2982
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.101729698035767e-05,
+      "loss": 2.8388,
+      "step": 2983
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.095866314863676e-05,
+      "loss": 2.8861,
+      "step": 2984
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0900029316915864e-05,
+      "loss": 2.9253,
+      "step": 2985
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.084139548519496e-05,
+      "loss": 2.8586,
+      "step": 2986
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.078276165347406e-05,
+      "loss": 2.8695,
+      "step": 2987
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0724127821753154e-05,
+      "loss": 2.9596,
+      "step": 2988
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.066549399003225e-05,
+      "loss": 2.8596,
+      "step": 2989
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.060686015831134e-05,
+      "loss": 2.9009,
+      "step": 2990
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0548226326590444e-05,
+      "loss": 2.8724,
+      "step": 2991
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.048959249486954e-05,
+      "loss": 2.8179,
+      "step": 2992
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 3.1793813705444336,
+      "eval_runtime": 2939.5659,
+      "eval_samples_per_second": 6.969,
+      "eval_steps_per_second": 2.323,
+      "step": 2992
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.043095866314864e-05,
+      "loss": 2.8459,
+      "step": 2993
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0372324831427734e-05,
+      "loss": 2.8515,
+      "step": 2994
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0313690999706835e-05,
+      "loss": 2.79,
+      "step": 2995
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.025505716798593e-05,
+      "loss": 2.8396,
+      "step": 2996
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.019642333626503e-05,
+      "loss": 2.7678,
+      "step": 2997
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0137789504544124e-05,
+      "loss": 2.895,
+      "step": 2998
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0079155672823222e-05,
+      "loss": 2.8841,
+      "step": 2999
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0020521841102316e-05,
+      "loss": 2.8433,
+      "step": 3000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9961888009381417e-05,
+      "loss": 2.9279,
+      "step": 3001
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9903254177660512e-05,
+      "loss": 2.8772,
+      "step": 3002
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.984462034593961e-05,
+      "loss": 2.8162,
+      "step": 3003
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9785986514218707e-05,
+      "loss": 2.8856,
+      "step": 3004
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.97273526824978e-05,
+      "loss": 2.8785,
+      "step": 3005
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.96687188507769e-05,
+      "loss": 2.8341,
+      "step": 3006
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9610085019055994e-05,
+      "loss": 2.894,
+      "step": 3007
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9551451187335095e-05,
+      "loss": 2.9335,
+      "step": 3008
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.949281735561419e-05,
+      "loss": 2.8878,
+      "step": 3009
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9434183523893287e-05,
+      "loss": 2.8899,
+      "step": 3010
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9375549692172384e-05,
+      "loss": 2.8749,
+      "step": 3011
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9316915860451482e-05,
+      "loss": 2.9289,
+      "step": 3012
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9258282028730576e-05,
+      "loss": 2.9509,
+      "step": 3013
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9199648197009678e-05,
+      "loss": 2.9502,
+      "step": 3014
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9141014365288772e-05,
+      "loss": 2.9716,
+      "step": 3015
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.908238053356787e-05,
+      "loss": 2.8487,
+      "step": 3016
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.9023746701846964e-05,
+      "loss": 2.9988,
+      "step": 3017
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8965112870126065e-05,
+      "loss": 2.8262,
+      "step": 3018
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.890647903840516e-05,
+      "loss": 2.8904,
+      "step": 3019
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.884784520668426e-05,
+      "loss": 2.7983,
+      "step": 3020
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8789211374963355e-05,
+      "loss": 2.8936,
+      "step": 3021
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8730577543242452e-05,
+      "loss": 2.9329,
+      "step": 3022
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8671943711521547e-05,
+      "loss": 2.9266,
+      "step": 3023
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8613309879800648e-05,
+      "loss": 2.8731,
+      "step": 3024
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8554676048079742e-05,
+      "loss": 2.9438,
+      "step": 3025
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8496042216358843e-05,
+      "loss": 2.8051,
+      "step": 3026
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8437408384637938e-05,
+      "loss": 2.8465,
+      "step": 3027
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8378774552917035e-05,
+      "loss": 2.89,
+      "step": 3028
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.832014072119613e-05,
+      "loss": 2.9555,
+      "step": 3029
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.826150688947523e-05,
+      "loss": 2.9402,
+      "step": 3030
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8202873057754325e-05,
+      "loss": 2.8841,
+      "step": 3031
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8144239226033426e-05,
+      "loss": 2.9948,
+      "step": 3032
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.808560539431252e-05,
+      "loss": 2.8503,
+      "step": 3033
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8026971562591615e-05,
+      "loss": 2.9322,
+      "step": 3034
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7968337730870712e-05,
+      "loss": 2.812,
+      "step": 3035
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7909703899149807e-05,
+      "loss": 2.8251,
+      "step": 3036
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7851070067428908e-05,
+      "loss": 2.9553,
+      "step": 3037
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7792436235708002e-05,
+      "loss": 2.942,
+      "step": 3038
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7733802403987103e-05,
+      "loss": 2.9353,
+      "step": 3039
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7675168572266198e-05,
+      "loss": 2.8833,
+      "step": 3040
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7616534740545295e-05,
+      "loss": 2.8876,
+      "step": 3041
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.755790090882439e-05,
+      "loss": 2.9271,
+      "step": 3042
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.749926707710349e-05,
+      "loss": 2.9971,
+      "step": 3043
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7440633245382585e-05,
+      "loss": 2.8194,
+      "step": 3044
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7381999413661686e-05,
+      "loss": 2.9262,
+      "step": 3045
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.732336558194078e-05,
+      "loss": 2.9073,
+      "step": 3046
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7264731750219878e-05,
+      "loss": 2.9261,
+      "step": 3047
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7206097918498972e-05,
+      "loss": 2.8828,
+      "step": 3048
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7147464086778074e-05,
+      "loss": 2.9443,
+      "step": 3049
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7088830255057168e-05,
+      "loss": 2.9206,
+      "step": 3050
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.703019642333627e-05,
+      "loss": 2.8375,
+      "step": 3051
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6971562591615363e-05,
+      "loss": 2.9127,
+      "step": 3052
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.691292875989446e-05,
+      "loss": 2.8895,
+      "step": 3053
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6854294928173555e-05,
+      "loss": 2.9373,
+      "step": 3054
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6795661096452656e-05,
+      "loss": 3.0038,
+      "step": 3055
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.673702726473175e-05,
+      "loss": 2.9043,
+      "step": 3056
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6678393433010852e-05,
+      "loss": 2.8431,
+      "step": 3057
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6619759601289946e-05,
+      "loss": 2.94,
+      "step": 3058
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6561125769569044e-05,
+      "loss": 2.9705,
+      "step": 3059
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6502491937848138e-05,
+      "loss": 2.8743,
+      "step": 3060
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6443858106127232e-05,
+      "loss": 2.9339,
+      "step": 3061
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6385224274406334e-05,
+      "loss": 2.8471,
+      "step": 3062
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6326590442685428e-05,
+      "loss": 2.9231,
+      "step": 3063
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.626795661096453e-05,
+      "loss": 2.895,
+      "step": 3064
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6209322779243623e-05,
+      "loss": 2.8238,
+      "step": 3065
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.615068894752272e-05,
+      "loss": 2.8911,
+      "step": 3066
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6092055115801815e-05,
+      "loss": 2.9757,
+      "step": 3067
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6033421284080916e-05,
+      "loss": 2.8924,
+      "step": 3068
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.597478745236001e-05,
+      "loss": 2.9346,
+      "step": 3069
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.5916153620639112e-05,
+      "loss": 2.909,
+      "step": 3070
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.5857519788918206e-05,
+      "loss": 2.9504,
+      "step": 3071
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.5798885957197304e-05,
+      "loss": 2.929,
+      "step": 3072
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5740252125476398e-05,
+      "loss": 3.0145,
+      "step": 3073
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.56816182937555e-05,
+      "loss": 2.8655,
+      "step": 3074
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5622984462034594e-05,
+      "loss": 2.9567,
+      "step": 3075
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5564350630313695e-05,
+      "loss": 2.8592,
+      "step": 3076
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.550571679859279e-05,
+      "loss": 2.8847,
+      "step": 3077
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5447082966871887e-05,
+      "loss": 2.8842,
+      "step": 3078
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.538844913515098e-05,
+      "loss": 2.9071,
+      "step": 3079
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5329815303430082e-05,
+      "loss": 2.8896,
+      "step": 3080
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5271181471709176e-05,
+      "loss": 2.9599,
+      "step": 3081
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5212547639988278e-05,
+      "loss": 2.9521,
+      "step": 3082
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5153913808267372e-05,
+      "loss": 2.8697,
+      "step": 3083
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.509527997654647e-05,
+      "loss": 2.8008,
+      "step": 3084
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.5036646144825564e-05,
+      "loss": 2.9064,
+      "step": 3085
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.497801231310466e-05,
+      "loss": 2.8085,
+      "step": 3086
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.491937848138376e-05,
+      "loss": 2.8273,
+      "step": 3087
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4860744649662857e-05,
+      "loss": 2.8658,
+      "step": 3088
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4802110817941955e-05,
+      "loss": 2.9475,
+      "step": 3089
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4743476986221052e-05,
+      "loss": 2.9246,
+      "step": 3090
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4684843154500147e-05,
+      "loss": 2.8969,
+      "step": 3091
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4626209322779244e-05,
+      "loss": 2.8539,
+      "step": 3092
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4567575491058342e-05,
+      "loss": 2.8667,
+      "step": 3093
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.450894165933744e-05,
+      "loss": 3.0136,
+      "step": 3094
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4450307827616538e-05,
+      "loss": 2.885,
+      "step": 3095
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4391673995895635e-05,
+      "loss": 2.8927,
+      "step": 3096
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.433304016417473e-05,
+      "loss": 2.8121,
+      "step": 3097
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4274406332453827e-05,
+      "loss": 2.9706,
+      "step": 3098
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4215772500732925e-05,
+      "loss": 2.7851,
+      "step": 3099
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4157138669012023e-05,
+      "loss": 2.9091,
+      "step": 3100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.409850483729112e-05,
+      "loss": 2.9729,
+      "step": 3101
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4039871005570218e-05,
+      "loss": 2.8444,
+      "step": 3102
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.3981237173849312e-05,
+      "loss": 2.8912,
+      "step": 3103
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.3922603342128407e-05,
+      "loss": 2.9623,
+      "step": 3104
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.3863969510407504e-05,
+      "loss": 2.8596,
+      "step": 3105
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.3805335678686602e-05,
+      "loss": 2.9232,
+      "step": 3106
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.37467018469657e-05,
+      "loss": 2.8836,
+      "step": 3107
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3688068015244798e-05,
+      "loss": 2.9154,
+      "step": 3108
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3629434183523895e-05,
+      "loss": 2.8813,
+      "step": 3109
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.357080035180299e-05,
+      "loss": 2.8312,
+      "step": 3110
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3512166520082087e-05,
+      "loss": 2.9847,
+      "step": 3111
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3453532688361185e-05,
+      "loss": 2.8094,
+      "step": 3112
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3394898856640283e-05,
+      "loss": 2.9362,
+      "step": 3113
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.333626502491938e-05,
+      "loss": 2.6475,
+      "step": 3114
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3277631193198475e-05,
+      "loss": 2.9112,
+      "step": 3115
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3218997361477572e-05,
+      "loss": 2.9251,
+      "step": 3116
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.316036352975667e-05,
+      "loss": 2.9166,
+      "step": 3117
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3101729698035768e-05,
+      "loss": 3.003,
+      "step": 3118
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3043095866314866e-05,
+      "loss": 2.8892,
+      "step": 3119
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2984462034593963e-05,
+      "loss": 2.8297,
+      "step": 3120
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2925828202873058e-05,
+      "loss": 2.9204,
+      "step": 3121
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2867194371152155e-05,
+      "loss": 2.9136,
+      "step": 3122
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2808560539431253e-05,
+      "loss": 2.8663,
+      "step": 3123
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.274992670771035e-05,
+      "loss": 2.8577,
+      "step": 3124
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.269129287598945e-05,
+      "loss": 2.8965,
+      "step": 3125
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2632659044268546e-05,
+      "loss": 2.8784,
+      "step": 3126
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.257402521254764e-05,
+      "loss": 2.8439,
+      "step": 3127
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2515391380826738e-05,
+      "loss": 2.95,
+      "step": 3128
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2456757549105836e-05,
+      "loss": 2.9214,
+      "step": 3129
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2398123717384934e-05,
+      "loss": 2.8409,
+      "step": 3130
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.233948988566403e-05,
+      "loss": 2.8276,
+      "step": 3131
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2280856053943126e-05,
+      "loss": 2.9065,
+      "step": 3132
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 2.8985,
+      "step": 3133
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2163588390501318e-05,
+      "loss": 2.917,
+      "step": 3134
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2104954558780415e-05,
+      "loss": 2.8759,
+      "step": 3135
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2046320727059513e-05,
+      "loss": 2.9307,
+      "step": 3136
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.198768689533861e-05,
+      "loss": 2.9063,
+      "step": 3137
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.192905306361771e-05,
+      "loss": 2.8303,
+      "step": 3138
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1870419231896806e-05,
+      "loss": 2.9144,
+      "step": 3139
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.18117854001759e-05,
+      "loss": 2.9602,
+      "step": 3140
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1753151568454998e-05,
+      "loss": 2.9569,
+      "step": 3141
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1694517736734096e-05,
+      "loss": 2.9886,
+      "step": 3142
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1635883905013194e-05,
+      "loss": 2.8674,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.157725007329229e-05,
+      "loss": 2.7881,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.151861624157139e-05,
+      "loss": 2.8678,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1459982409850483e-05,
+      "loss": 2.83,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.140134857812958e-05,
+      "loss": 2.8217,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.134271474640868e-05,
+      "loss": 2.8975,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1284080914687776e-05,
+      "loss": 2.8039,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1225447082966874e-05,
+      "loss": 2.8653,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1166813251245972e-05,
+      "loss": 2.8633,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1108179419525066e-05,
+      "loss": 2.9128,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.1049545587804164e-05,
+      "loss": 2.9029,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.099091175608326e-05,
+      "loss": 2.7887,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.093227792436236e-05,
+      "loss": 2.822,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0873644092641457e-05,
+      "loss": 2.9808,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0815010260920555e-05,
+      "loss": 2.9651,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.075637642919965e-05,
+      "loss": 2.8369,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0697742597478747e-05,
+      "loss": 2.8609,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.063910876575784e-05,
+      "loss": 2.8827,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.058047493403694e-05,
+      "loss": 2.8084,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0521841102316036e-05,
+      "loss": 2.7938,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0463207270595134e-05,
+      "loss": 2.9767,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0404573438874232e-05,
+      "loss": 2.9485,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0345939607153326e-05,
+      "loss": 2.9543,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0287305775432424e-05,
+      "loss": 2.9186,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.022867194371152e-05,
+      "loss": 2.9111,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.017003811199062e-05,
+      "loss": 2.836,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 3.160200357437134,
+      "eval_runtime": 2939.8509,
+      "eval_samples_per_second": 6.969,
+      "eval_steps_per_second": 2.323,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0111404280269717e-05,
+      "loss": 2.8292,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0052770448548815e-05,
+      "loss": 2.8157,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.999413661682791e-05,
+      "loss": 2.8245,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9935502785107007e-05,
+      "loss": 2.9398,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9876868953386104e-05,
+      "loss": 2.796,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9818235121665202e-05,
+      "loss": 2.8846,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.97596012899443e-05,
+      "loss": 2.7891,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9700967458223398e-05,
+      "loss": 2.9822,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9642333626502492e-05,
+      "loss": 2.9676,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.958369979478159e-05,
+      "loss": 2.8681,
+      "step": 3178
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9525065963060687e-05,
+      "loss": 2.8187,
+      "step": 3179
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9466432131339785e-05,
+      "loss": 2.8531,
+      "step": 3180
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9407798299618883e-05,
+      "loss": 2.8477,
+      "step": 3181
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.934916446789798e-05,
+      "loss": 2.9817,
+      "step": 3182
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9290530636177075e-05,
+      "loss": 2.8434,
+      "step": 3183
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9231896804456172e-05,
+      "loss": 2.9335,
+      "step": 3184
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.917326297273527e-05,
+      "loss": 2.7555,
+      "step": 3185
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9114629141014368e-05,
+      "loss": 2.83,
+      "step": 3186
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9055995309293466e-05,
+      "loss": 2.7177,
+      "step": 3187
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.899736147757256e-05,
+      "loss": 2.8393,
+      "step": 3188
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8938727645851658e-05,
+      "loss": 2.9234,
+      "step": 3189
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8880093814130752e-05,
+      "loss": 2.9678,
+      "step": 3190
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.882145998240985e-05,
+      "loss": 2.9529,
+      "step": 3191
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8762826150688947e-05,
+      "loss": 2.9177,
+      "step": 3192
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8704192318968045e-05,
+      "loss": 2.8308,
+      "step": 3193
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8645558487247143e-05,
+      "loss": 2.8068,
+      "step": 3194
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.858692465552624e-05,
+      "loss": 2.8895,
+      "step": 3195
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8528290823805335e-05,
+      "loss": 2.8562,
+      "step": 3196
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8469656992084432e-05,
+      "loss": 2.9185,
+      "step": 3197
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.841102316036353e-05,
+      "loss": 2.8922,
+      "step": 3198
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8352389328642628e-05,
+      "loss": 2.7684,
+      "step": 3199
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8293755496921726e-05,
+      "loss": 2.9996,
+      "step": 3200
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8235121665200823e-05,
+      "loss": 2.8695,
+      "step": 3201
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8176487833479918e-05,
+      "loss": 2.8365,
+      "step": 3202
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8117854001759015e-05,
+      "loss": 2.8094,
+      "step": 3203
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8059220170038113e-05,
+      "loss": 2.9036,
+      "step": 3204
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.800058633831721e-05,
+      "loss": 2.7422,
+      "step": 3205
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.794195250659631e-05,
+      "loss": 2.84,
+      "step": 3206
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7883318674875406e-05,
+      "loss": 2.8855,
+      "step": 3207
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.78246848431545e-05,
+      "loss": 2.8716,
+      "step": 3208
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7766051011433598e-05,
+      "loss": 2.9577,
+      "step": 3209
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7707417179712696e-05,
+      "loss": 2.9218,
+      "step": 3210
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7648783347991794e-05,
+      "loss": 2.8567,
+      "step": 3211
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.759014951627089e-05,
+      "loss": 2.941,
+      "step": 3212
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7531515684549986e-05,
+      "loss": 2.8864,
+      "step": 3213
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7472881852829083e-05,
+      "loss": 2.8474,
+      "step": 3214
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.741424802110818e-05,
+      "loss": 2.8396,
+      "step": 3215
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.735561418938728e-05,
+      "loss": 2.813,
+      "step": 3216
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7296980357666373e-05,
+      "loss": 2.8872,
+      "step": 3217
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.723834652594547e-05,
+      "loss": 2.9066,
+      "step": 3218
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.717971269422457e-05,
+      "loss": 2.9,
+      "step": 3219
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7121078862503663e-05,
+      "loss": 2.9115,
+      "step": 3220
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.706244503078276e-05,
+      "loss": 2.7752,
+      "step": 3221
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7003811199061858e-05,
+      "loss": 2.9374,
+      "step": 3222
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6945177367340956e-05,
+      "loss": 2.8273,
+      "step": 3223
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6886543535620054e-05,
+      "loss": 2.7782,
+      "step": 3224
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.682790970389915e-05,
+      "loss": 2.9504,
+      "step": 3225
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6769275872178246e-05,
+      "loss": 2.8504,
+      "step": 3226
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6710642040457343e-05,
+      "loss": 2.8606,
+      "step": 3227
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.665200820873644e-05,
+      "loss": 2.7783,
+      "step": 3228
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.659337437701554e-05,
+      "loss": 2.8033,
+      "step": 3229
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6534740545294636e-05,
+      "loss": 2.8692,
+      "step": 3230
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6476106713573734e-05,
+      "loss": 2.8594,
+      "step": 3231
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.641747288185283e-05,
+      "loss": 2.8659,
+      "step": 3232
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6358839050131926e-05,
+      "loss": 2.8606,
+      "step": 3233
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6300205218411024e-05,
+      "loss": 2.7963,
+      "step": 3234
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.624157138669012e-05,
+      "loss": 2.8598,
+      "step": 3235
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.618293755496922e-05,
+      "loss": 2.8262,
+      "step": 3236
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6124303723248317e-05,
+      "loss": 2.8283,
+      "step": 3237
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.606566989152741e-05,
+      "loss": 3.0016,
+      "step": 3238
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.600703605980651e-05,
+      "loss": 2.8427,
+      "step": 3239
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5948402228085607e-05,
+      "loss": 3.0162,
+      "step": 3240
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5889768396364704e-05,
+      "loss": 2.7804,
+      "step": 3241
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5831134564643802e-05,
+      "loss": 2.8232,
+      "step": 3242
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.57725007329229e-05,
+      "loss": 2.9136,
+      "step": 3243
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5713866901201994e-05,
+      "loss": 2.8262,
+      "step": 3244
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.565523306948109e-05,
+      "loss": 2.8576,
+      "step": 3245
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5596599237760186e-05,
+      "loss": 2.85,
+      "step": 3246
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5537965406039284e-05,
+      "loss": 2.9155,
+      "step": 3247
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.547933157431838e-05,
+      "loss": 2.9287,
+      "step": 3248
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.542069774259748e-05,
+      "loss": 2.9685,
+      "step": 3249
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5362063910876577e-05,
+      "loss": 2.9452,
+      "step": 3250
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.530343007915567e-05,
+      "loss": 2.839,
+      "step": 3251
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.524479624743477e-05,
+      "loss": 2.8205,
+      "step": 3252
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5186162415713867e-05,
+      "loss": 2.965,
+      "step": 3253
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5127528583992964e-05,
+      "loss": 2.9346,
+      "step": 3254
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5068894752272062e-05,
+      "loss": 2.8178,
+      "step": 3255
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5010260920551158e-05,
+      "loss": 2.7931,
+      "step": 3256
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4951627088830256e-05,
+      "loss": 2.8281,
+      "step": 3257
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4892993257109354e-05,
+      "loss": 2.8357,
+      "step": 3258
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.483435942538845e-05,
+      "loss": 2.8288,
+      "step": 3259
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4775725593667547e-05,
+      "loss": 2.8825,
+      "step": 3260
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4717091761946643e-05,
+      "loss": 2.7498,
+      "step": 3261
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4658457930225741e-05,
+      "loss": 2.806,
+      "step": 3262
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4599824098504839e-05,
+      "loss": 2.9031,
+      "step": 3263
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4541190266783935e-05,
+      "loss": 2.8699,
+      "step": 3264
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4482556435063032e-05,
+      "loss": 2.9579,
+      "step": 3265
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.442392260334213e-05,
+      "loss": 2.7564,
+      "step": 3266
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4365288771621226e-05,
+      "loss": 2.9202,
+      "step": 3267
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4306654939900324e-05,
+      "loss": 2.8848,
+      "step": 3268
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4248021108179422e-05,
+      "loss": 2.8895,
+      "step": 3269
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4189387276458518e-05,
+      "loss": 2.8821,
+      "step": 3270
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4130753444737615e-05,
+      "loss": 2.9572,
+      "step": 3271
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4072119613016713e-05,
+      "loss": 2.9551,
+      "step": 3272
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4013485781295807e-05,
+      "loss": 2.7267,
+      "step": 3273
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3954851949574903e-05,
+      "loss": 2.7621,
+      "step": 3274
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3896218117854001e-05,
+      "loss": 2.8636,
+      "step": 3275
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3837584286133099e-05,
+      "loss": 2.8203,
+      "step": 3276
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3778950454412195e-05,
+      "loss": 2.7836,
+      "step": 3277
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3720316622691292e-05,
+      "loss": 2.8999,
+      "step": 3278
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.366168279097039e-05,
+      "loss": 2.851,
+      "step": 3279
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3603048959249486e-05,
+      "loss": 2.862,
+      "step": 3280
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3544415127528584e-05,
+      "loss": 2.8626,
+      "step": 3281
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3485781295807682e-05,
+      "loss": 2.9474,
+      "step": 3282
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3427147464086778e-05,
+      "loss": 2.8565,
+      "step": 3283
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3368513632365875e-05,
+      "loss": 2.9163,
+      "step": 3284
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3309879800644973e-05,
+      "loss": 2.8821,
+      "step": 3285
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3251245968924069e-05,
+      "loss": 2.8121,
+      "step": 3286
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3192612137203167e-05,
+      "loss": 2.9233,
+      "step": 3287
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3133978305482264e-05,
+      "loss": 2.9479,
+      "step": 3288
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.307534447376136e-05,
+      "loss": 2.8446,
+      "step": 3289
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3016710642040458e-05,
+      "loss": 2.8088,
+      "step": 3290
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2958076810319556e-05,
+      "loss": 2.8534,
+      "step": 3291
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2899442978598652e-05,
+      "loss": 2.7208,
+      "step": 3292
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.284080914687775e-05,
+      "loss": 2.8408,
+      "step": 3293
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2782175315156847e-05,
+      "loss": 2.8883,
+      "step": 3294
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2723541483435943e-05,
+      "loss": 2.8794,
+      "step": 3295
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2664907651715041e-05,
+      "loss": 2.8654,
+      "step": 3296
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2606273819994139e-05,
+      "loss": 2.8625,
+      "step": 3297
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2547639988273235e-05,
+      "loss": 2.9146,
+      "step": 3298
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.248900615655233e-05,
+      "loss": 2.839,
+      "step": 3299
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2430372324831428e-05,
+      "loss": 2.9716,
+      "step": 3300
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2371738493110526e-05,
+      "loss": 2.8428,
+      "step": 3301
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2313104661389622e-05,
+      "loss": 2.9266,
+      "step": 3302
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.225447082966872e-05,
+      "loss": 2.8197,
+      "step": 3303
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2195836997947818e-05,
+      "loss": 2.9407,
+      "step": 3304
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2137203166226914e-05,
+      "loss": 2.8513,
+      "step": 3305
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2078569334506011e-05,
+      "loss": 2.8183,
+      "step": 3306
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2019935502785109e-05,
+      "loss": 2.8864,
+      "step": 3307
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1961301671064203e-05,
+      "loss": 2.9143,
+      "step": 3308
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1902667839343301e-05,
+      "loss": 2.8703,
+      "step": 3309
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1844034007622399e-05,
+      "loss": 2.9207,
+      "step": 3310
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1785400175901495e-05,
+      "loss": 2.9206,
+      "step": 3311
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1726766344180592e-05,
+      "loss": 2.8417,
+      "step": 3312
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.166813251245969e-05,
+      "loss": 2.7817,
+      "step": 3313
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1609498680738786e-05,
+      "loss": 2.8918,
+      "step": 3314
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1550864849017884e-05,
+      "loss": 2.8252,
+      "step": 3315
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1492231017296982e-05,
+      "loss": 2.8636,
+      "step": 3316
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1433597185576078e-05,
+      "loss": 2.9615,
+      "step": 3317
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1374963353855175e-05,
+      "loss": 2.9234,
+      "step": 3318
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1316329522134273e-05,
+      "loss": 2.8497,
+      "step": 3319
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1257695690413369e-05,
+      "loss": 2.9749,
+      "step": 3320
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1199061858692467e-05,
+      "loss": 2.9241,
+      "step": 3321
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1140428026971563e-05,
+      "loss": 2.873,
+      "step": 3322
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1081794195250659e-05,
+      "loss": 2.851,
+      "step": 3323
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1023160363529757e-05,
+      "loss": 2.8547,
+      "step": 3324
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0964526531808854e-05,
+      "loss": 2.8402,
+      "step": 3325
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.090589270008795e-05,
+      "loss": 2.9167,
+      "step": 3326
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0847258868367048e-05,
+      "loss": 2.8132,
+      "step": 3327
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0788625036646146e-05,
+      "loss": 2.9386,
+      "step": 3328
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0729991204925242e-05,
+      "loss": 2.8653,
+      "step": 3329
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.067135737320434e-05,
+      "loss": 2.862,
+      "step": 3330
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0612723541483437e-05,
+      "loss": 2.7229,
+      "step": 3331
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0554089709762533e-05,
+      "loss": 2.8003,
+      "step": 3332
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.049545587804163e-05,
+      "loss": 2.8708,
+      "step": 3333
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0436822046320728e-05,
+      "loss": 2.8852,
+      "step": 3334
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0378188214599825e-05,
+      "loss": 2.8857,
+      "step": 3335
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.031955438287892e-05,
+      "loss": 2.7855,
+      "step": 3336
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0260920551158018e-05,
+      "loss": 2.8554,
+      "step": 3337
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0202286719437116e-05,
+      "loss": 2.8743,
+      "step": 3338
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0143652887716212e-05,
+      "loss": 2.8712,
+      "step": 3339
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.008501905599531e-05,
+      "loss": 3.0037,
+      "step": 3340
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0026385224274407e-05,
+      "loss": 2.8358,
+      "step": 3341
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.967751392553503e-06,
+      "loss": 2.869,
+      "step": 3342
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.909117560832601e-06,
+      "loss": 2.8204,
+      "step": 3343
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.850483729111699e-06,
+      "loss": 2.9861,
+      "step": 3344
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 3.1451783180236816,
+      "eval_runtime": 2939.0886,
+      "eval_samples_per_second": 6.971,
+      "eval_steps_per_second": 2.324,
+      "step": 3344
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.791849897390795e-06,
+      "loss": 2.8953,
+      "step": 3345
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.733216065669893e-06,
+      "loss": 2.8766,
+      "step": 3346
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.67458223394899e-06,
+      "loss": 2.9005,
+      "step": 3347
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.615948402228086e-06,
+      "loss": 2.8427,
+      "step": 3348
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.557314570507184e-06,
+      "loss": 2.8354,
+      "step": 3349
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.49868073878628e-06,
+      "loss": 2.8795,
+      "step": 3350
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.440046907065376e-06,
+      "loss": 2.8449,
+      "step": 3351
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.381413075344474e-06,
+      "loss": 2.9221,
+      "step": 3352
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.322779243623571e-06,
+      "loss": 2.8642,
+      "step": 3353
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.264145411902667e-06,
+      "loss": 2.8635,
+      "step": 3354
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.205511580181765e-06,
+      "loss": 2.9202,
+      "step": 3355
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.146877748460863e-06,
+      "loss": 2.8686,
+      "step": 3356
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.088243916739959e-06,
+      "loss": 2.9704,
+      "step": 3357
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.029610085019057e-06,
+      "loss": 2.8669,
+      "step": 3358
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.970976253298154e-06,
+      "loss": 2.8702,
+      "step": 3359
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.91234242157725e-06,
+      "loss": 2.8498,
+      "step": 3360
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.853708589856348e-06,
+      "loss": 2.8873,
+      "step": 3361
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.795074758135446e-06,
+      "loss": 2.8563,
+      "step": 3362
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.736440926414542e-06,
+      "loss": 2.9415,
+      "step": 3363
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.67780709469364e-06,
+      "loss": 2.8061,
+      "step": 3364
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.619173262972735e-06,
+      "loss": 2.7686,
+      "step": 3365
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.560539431251831e-06,
+      "loss": 2.9245,
+      "step": 3366
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.501905599530929e-06,
+      "loss": 2.882,
+      "step": 3367
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.443271767810027e-06,
+      "loss": 2.7816,
+      "step": 3368
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.384637936089123e-06,
+      "loss": 2.9509,
+      "step": 3369
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.32600410436822e-06,
+      "loss": 2.8633,
+      "step": 3370
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.267370272647318e-06,
+      "loss": 2.9092,
+      "step": 3371
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.208736440926414e-06,
+      "loss": 2.8923,
+      "step": 3372
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.150102609205512e-06,
+      "loss": 2.8727,
+      "step": 3373
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.09146877748461e-06,
+      "loss": 2.8452,
+      "step": 3374
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.032834945763706e-06,
+      "loss": 2.8718,
+      "step": 3375
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.974201114042803e-06,
+      "loss": 2.8629,
+      "step": 3376
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.915567282321901e-06,
+      "loss": 2.9427,
+      "step": 3377
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.856933450600997e-06,
+      "loss": 2.8703,
+      "step": 3378
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.798299618880093e-06,
+      "loss": 2.851,
+      "step": 3379
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.73966578715919e-06,
+      "loss": 2.922,
+      "step": 3380
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.681031955438289e-06,
+      "loss": 2.9027,
+      "step": 3381
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.622398123717385e-06,
+      "loss": 2.8629,
+      "step": 3382
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.563764291996482e-06,
+      "loss": 2.8226,
+      "step": 3383
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.505130460275579e-06,
+      "loss": 2.9194,
+      "step": 3384
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.446496628554677e-06,
+      "loss": 2.8865,
+      "step": 3385
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.387862796833774e-06,
+      "loss": 2.7308,
+      "step": 3386
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.3292289651128705e-06,
+      "loss": 2.809,
+      "step": 3387
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.270595133391967e-06,
+      "loss": 2.8759,
+      "step": 3388
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.211961301671065e-06,
+      "loss": 2.9575,
+      "step": 3389
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.153327469950162e-06,
+      "loss": 2.8045,
+      "step": 3390
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.094693638229259e-06,
+      "loss": 2.8413,
+      "step": 3391
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.0360598065083565e-06,
+      "loss": 2.8926,
+      "step": 3392
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.977425974787452e-06,
+      "loss": 2.8948,
+      "step": 3393
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.918792143066549e-06,
+      "loss": 2.855,
+      "step": 3394
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.860158311345646e-06,
+      "loss": 2.8372,
+      "step": 3395
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.801524479624743e-06,
+      "loss": 2.8268,
+      "step": 3396
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.742890647903841e-06,
+      "loss": 2.8013,
+      "step": 3397
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.684256816182938e-06,
+      "loss": 2.8179,
+      "step": 3398
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.6256229844620345e-06,
+      "loss": 2.875,
+      "step": 3399
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.566989152741132e-06,
+      "loss": 2.8707,
+      "step": 3400
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.508355321020229e-06,
+      "loss": 2.8618,
+      "step": 3401
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.449721489299326e-06,
+      "loss": 2.8761,
+      "step": 3402
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.391087657578424e-06,
+      "loss": 2.8959,
+      "step": 3403
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.3324538258575205e-06,
+      "loss": 2.8393,
+      "step": 3404
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.273819994136617e-06,
+      "loss": 2.9073,
+      "step": 3405
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.215186162415714e-06,
+      "loss": 2.9331,
+      "step": 3406
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.156552330694811e-06,
+      "loss": 2.8051,
+      "step": 3407
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.097918498973909e-06,
+      "loss": 2.8643,
+      "step": 3408
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.039284667253006e-06,
+      "loss": 2.9364,
+      "step": 3409
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.980650835532102e-06,
+      "loss": 2.8516,
+      "step": 3410
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.922017003811199e-06,
+      "loss": 2.9023,
+      "step": 3411
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.863383172090296e-06,
+      "loss": 2.863,
+      "step": 3412
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.804749340369393e-06,
+      "loss": 2.8365,
+      "step": 3413
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.746115508648491e-06,
+      "loss": 2.8407,
+      "step": 3414
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.687481676927588e-06,
+      "loss": 2.7719,
+      "step": 3415
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.6288478452066845e-06,
+      "loss": 2.8312,
+      "step": 3416
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.570214013485781e-06,
+      "loss": 2.8941,
+      "step": 3417
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.511580181764878e-06,
+      "loss": 2.8098,
+      "step": 3418
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.452946350043975e-06,
+      "loss": 2.8687,
+      "step": 3419
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.394312518323073e-06,
+      "loss": 2.8045,
+      "step": 3420
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.33567868660217e-06,
+      "loss": 2.6793,
+      "step": 3421
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.2770448548812665e-06,
+      "loss": 2.8539,
+      "step": 3422
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.218411023160364e-06,
+      "loss": 2.8605,
+      "step": 3423
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.15977719143946e-06,
+      "loss": 2.834,
+      "step": 3424
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.101143359718558e-06,
+      "loss": 2.8213,
+      "step": 3425
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.042509527997655e-06,
+      "loss": 2.8129,
+      "step": 3426
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.983875696276752e-06,
+      "loss": 2.9176,
+      "step": 3427
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.925241864555849e-06,
+      "loss": 2.895,
+      "step": 3428
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.866608032834946e-06,
+      "loss": 2.7044,
+      "step": 3429
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.807974201114043e-06,
+      "loss": 2.891,
+      "step": 3430
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.74934036939314e-06,
+      "loss": 2.7879,
+      "step": 3431
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.690706537672237e-06,
+      "loss": 2.8971,
+      "step": 3432
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.632072705951334e-06,
+      "loss": 2.8768,
+      "step": 3433
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.573438874230431e-06,
+      "loss": 2.7405,
+      "step": 3434
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.514805042509528e-06,
+      "loss": 2.8209,
+      "step": 3435
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.456171210788625e-06,
+      "loss": 2.8784,
+      "step": 3436
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.397537379067723e-06,
+      "loss": 2.7199,
+      "step": 3437
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.33890354734682e-06,
+      "loss": 2.912,
+      "step": 3438
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.280269715625916e-06,
+      "loss": 2.8641,
+      "step": 3439
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.221635883905013e-06,
+      "loss": 2.7801,
+      "step": 3440
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.16300205218411e-06,
+      "loss": 2.8512,
+      "step": 3441
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.104368220463207e-06,
+      "loss": 2.7977,
+      "step": 3442
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.045734388742305e-06,
+      "loss": 2.8873,
+      "step": 3443
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.987100557021402e-06,
+      "loss": 2.8778,
+      "step": 3444
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.9284667253004985e-06,
+      "loss": 2.731,
+      "step": 3445
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.869832893579595e-06,
+      "loss": 2.9343,
+      "step": 3446
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.8111990618586927e-06,
+      "loss": 2.7655,
+      "step": 3447
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.7525652301377895e-06,
+      "loss": 2.9145,
+      "step": 3448
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.693931398416887e-06,
+      "loss": 2.7719,
+      "step": 3449
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.6352975666959837e-06,
+      "loss": 2.8715,
+      "step": 3450
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.576663734975081e-06,
+      "loss": 2.7961,
+      "step": 3451
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.5180299032541783e-06,
+      "loss": 2.8308,
+      "step": 3452
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4593960715332747e-06,
+      "loss": 2.9132,
+      "step": 3453
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4007622398123716e-06,
+      "loss": 2.87,
+      "step": 3454
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.342128408091469e-06,
+      "loss": 2.8685,
+      "step": 3455
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.283494576370566e-06,
+      "loss": 2.9182,
+      "step": 3456
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.224860744649663e-06,
+      "loss": 2.8645,
+      "step": 3457
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1662269129287603e-06,
+      "loss": 2.8448,
+      "step": 3458
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.107593081207857e-06,
+      "loss": 2.9324,
+      "step": 3459
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0489592494869544e-06,
+      "loss": 2.8694,
+      "step": 3460
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.990325417766051e-06,
+      "loss": 2.6942,
+      "step": 3461
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.931691586045148e-06,
+      "loss": 2.8406,
+      "step": 3462
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.8730577543242454e-06,
+      "loss": 2.8226,
+      "step": 3463
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.8144239226033423e-06,
+      "loss": 2.8704,
+      "step": 3464
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.755790090882439e-06,
+      "loss": 2.8336,
+      "step": 3465
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6971562591615364e-06,
+      "loss": 2.8579,
+      "step": 3466
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6385224274406333e-06,
+      "loss": 2.8385,
+      "step": 3467
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.57988859571973e-06,
+      "loss": 2.93,
+      "step": 3468
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5212547639988274e-06,
+      "loss": 3.0071,
+      "step": 3469
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4626209322779247e-06,
+      "loss": 2.8082,
+      "step": 3470
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4039871005570216e-06,
+      "loss": 2.8131,
+      "step": 3471
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3453532688361184e-06,
+      "loss": 2.9279,
+      "step": 3472
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2867194371152157e-06,
+      "loss": 2.8262,
+      "step": 3473
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2280856053943126e-06,
+      "loss": 2.8332,
+      "step": 3474
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.16945177367341e-06,
+      "loss": 2.8972,
+      "step": 3475
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.1108179419525067e-06,
+      "loss": 2.8598,
+      "step": 3476
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.0521841102316036e-06,
+      "loss": 2.8697,
+      "step": 3477
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.993550278510701e-06,
+      "loss": 2.8586,
+      "step": 3478
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9349164467897977e-06,
+      "loss": 2.8463,
+      "step": 3479
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8762826150688948e-06,
+      "loss": 2.7904,
+      "step": 3480
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8176487833479918e-06,
+      "loss": 2.8856,
+      "step": 3481
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.7590149516270891e-06,
+      "loss": 2.8751,
+      "step": 3482
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.7003811199061858e-06,
+      "loss": 2.8387,
+      "step": 3483
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.641747288185283e-06,
+      "loss": 2.8042,
+      "step": 3484
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5831134564643801e-06,
+      "loss": 2.7155,
+      "step": 3485
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5244796247434772e-06,
+      "loss": 2.7341,
+      "step": 3486
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.465845793022574e-06,
+      "loss": 2.9217,
+      "step": 3487
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4072119613016711e-06,
+      "loss": 2.8898,
+      "step": 3488
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.3485781295807682e-06,
+      "loss": 2.8511,
+      "step": 3489
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.289944297859865e-06,
+      "loss": 2.7788,
+      "step": 3490
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2313104661389623e-06,
+      "loss": 2.7344,
+      "step": 3491
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1726766344180592e-06,
+      "loss": 2.8135,
+      "step": 3492
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1140428026971563e-06,
+      "loss": 2.8553,
+      "step": 3493
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0554089709762534e-06,
+      "loss": 2.7931,
+      "step": 3494
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.967751392553504e-07,
+      "loss": 2.8761,
+      "step": 3495
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.381413075344474e-07,
+      "loss": 2.8658,
+      "step": 3496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.795074758135446e-07,
+      "loss": 2.8199,
+      "step": 3497
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.208736440926415e-07,
+      "loss": 2.8918,
+      "step": 3498
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.622398123717386e-07,
+      "loss": 2.8117,
+      "step": 3499
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.036059806508356e-07,
+      "loss": 2.8405,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.449721489299325e-07,
+      "loss": 2.9002,
+      "step": 3501
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.863383172090296e-07,
+      "loss": 2.8177,
+      "step": 3502
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.277044854881267e-07,
+      "loss": 2.9016,
+      "step": 3503
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.690706537672237e-07,
+      "loss": 2.87,
+      "step": 3504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1043682204632077e-07,
+      "loss": 2.8673,
+      "step": 3505
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.518029903254178e-07,
+      "loss": 2.8511,
+      "step": 3506
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.931691586045148e-07,
+      "loss": 2.8761,
+      "step": 3507
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3453532688361185e-07,
+      "loss": 2.8007,
+      "step": 3508
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.759014951627089e-07,
+      "loss": 2.8836,
+      "step": 3509
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1726766344180592e-07,
+      "loss": 2.8452,
+      "step": 3510
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.863383172090296e-08,
+      "loss": 2.8898,
+      "step": 3511
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3511,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 275654859816960.0,
+  "trial_name": null,
+  "trial_params": null
+}