diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.999124215443001,
+  "eval_steps": 500,
+  "global_step": 1712,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011677127426653043,
+      "grad_norm": 2.8304451130782486,
+      "learning_rate": 1.1627906976744187e-07,
+      "loss": 0.731,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023354254853306085,
+      "grad_norm": 2.739602109681187,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 0.7212,
+      "step": 2
+    },
+    {
+      "epoch": 0.003503138227995913,
+      "grad_norm": 2.7308900172867925,
+      "learning_rate": 3.488372093023256e-07,
+      "loss": 0.7317,
+      "step": 3
+    },
+    {
+      "epoch": 0.004670850970661217,
+      "grad_norm": 2.8097974437148174,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.7416,
+      "step": 4
+    },
+    {
+      "epoch": 0.005838563713326522,
+      "grad_norm": 2.7488918765022086,
+      "learning_rate": 5.813953488372094e-07,
+      "loss": 0.7368,
+      "step": 5
+    },
+    {
+      "epoch": 0.007006276455991826,
+      "grad_norm": 2.9086962066727393,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 0.7212,
+      "step": 6
+    },
+    {
+      "epoch": 0.00817398919865713,
+      "grad_norm": 2.674978233489208,
+      "learning_rate": 8.139534883720931e-07,
+      "loss": 0.7247,
+      "step": 7
+    },
+    {
+      "epoch": 0.009341701941322434,
+      "grad_norm": 2.5038568211093626,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.6934,
+      "step": 8
+    },
+    {
+      "epoch": 0.01050941468398774,
+      "grad_norm": 2.6532636891290315,
+      "learning_rate": 1.0465116279069768e-06,
+      "loss": 0.713,
+      "step": 9
+    },
+    {
+      "epoch": 0.011677127426653043,
+      "grad_norm": 2.2367726624232613,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.6842,
+      "step": 10
+    },
+    {
+      "epoch": 0.012844840169318347,
+      "grad_norm": 2.227274773522393,
+      "learning_rate": 1.2790697674418605e-06,
+      "loss": 0.7113,
+      "step": 11
+    },
+    {
+      "epoch": 0.014012552911983653,
+      "grad_norm": 2.080459141847197,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.6965,
+      "step": 12
+    },
+    {
+      "epoch": 0.015180265654648957,
+      "grad_norm": 1.8889695119165482,
+      "learning_rate": 1.5116279069767443e-06,
+      "loss": 0.6747,
+      "step": 13
+    },
+    {
+      "epoch": 0.01634797839731426,
+      "grad_norm": 1.448624048201056,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.6968,
+      "step": 14
+    },
+    {
+      "epoch": 0.017515691139979566,
+      "grad_norm": 1.320650159585573,
+      "learning_rate": 1.7441860465116282e-06,
+      "loss": 0.6451,
+      "step": 15
+    },
+    {
+      "epoch": 0.018683403882644868,
+      "grad_norm": 1.273572139161287,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.6851,
+      "step": 16
+    },
+    {
+      "epoch": 0.019851116625310174,
+      "grad_norm": 1.2443492382755228,
+      "learning_rate": 1.976744186046512e-06,
+      "loss": 0.6596,
+      "step": 17
+    },
+    {
+      "epoch": 0.02101882936797548,
+      "grad_norm": 1.1916282425261147,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.6043,
+      "step": 18
+    },
+    {
+      "epoch": 0.02218654211064078,
+      "grad_norm": 1.814855419015494,
+      "learning_rate": 2.2093023255813954e-06,
+      "loss": 0.6156,
+      "step": 19
+    },
+    {
+      "epoch": 0.023354254853306087,
+      "grad_norm": 2.217087104862634,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.644,
+      "step": 20
+    },
+    {
+      "epoch": 0.024521967595971392,
+      "grad_norm": 2.2736967005413455,
+      "learning_rate": 2.4418604651162793e-06,
+      "loss": 0.6034,
+      "step": 21
+    },
+    {
+      "epoch": 0.025689680338636695,
+      "grad_norm": 1.8576152069772087,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.6324,
+      "step": 22
+    },
+    {
+      "epoch": 0.026857393081302,
+      "grad_norm": 1.451256748229803,
+      "learning_rate": 2.674418604651163e-06,
+      "loss": 0.5839,
+      "step": 23
+    },
+    {
+      "epoch": 0.028025105823967306,
+      "grad_norm": 1.2515195024751948,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.6107,
+      "step": 24
+    },
+    {
+      "epoch": 0.029192818566632608,
+      "grad_norm": 1.0548353228570377,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 0.58,
+      "step": 25
+    },
+    {
+      "epoch": 0.030360531309297913,
+      "grad_norm": 1.0125847218538686,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.5955,
+      "step": 26
+    },
+    {
+      "epoch": 0.031528244051963215,
+      "grad_norm": 1.1331592793824488,
+      "learning_rate": 3.1395348837209307e-06,
+      "loss": 0.5741,
+      "step": 27
+    },
+    {
+      "epoch": 0.03269595679462852,
+      "grad_norm": 1.0653832205442737,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.5789,
+      "step": 28
+    },
+    {
+      "epoch": 0.033863669537293827,
+      "grad_norm": 1.072347534511007,
+      "learning_rate": 3.372093023255814e-06,
+      "loss": 0.5725,
+      "step": 29
+    },
+    {
+      "epoch": 0.03503138227995913,
+      "grad_norm": 1.0804489322853839,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.6065,
+      "step": 30
+    },
+    {
+      "epoch": 0.03619909502262444,
+      "grad_norm": 1.055685844262672,
+      "learning_rate": 3.6046511627906977e-06,
+      "loss": 0.606,
+      "step": 31
+    },
+    {
+      "epoch": 0.037366807765289736,
+      "grad_norm": 0.9033926431968798,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.6183,
+      "step": 32
+    },
+    {
+      "epoch": 0.03853452050795504,
+      "grad_norm": 0.9236696128600146,
+      "learning_rate": 3.837209302325582e-06,
+      "loss": 0.5492,
+      "step": 33
+    },
+    {
+      "epoch": 0.03970223325062035,
+      "grad_norm": 0.9794654987023562,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.554,
+      "step": 34
+    },
+    {
+      "epoch": 0.04086994599328565,
+      "grad_norm": 0.9403999785737593,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 0.5407,
+      "step": 35
+    },
+    {
+      "epoch": 0.04203765873595096,
+      "grad_norm": 0.8596511269136627,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.5213,
+      "step": 36
+    },
+    {
+      "epoch": 0.04320537147861626,
+      "grad_norm": 0.8578716650811907,
+      "learning_rate": 4.302325581395349e-06,
+      "loss": 0.5525,
+      "step": 37
+    },
+    {
+      "epoch": 0.04437308422128156,
+      "grad_norm": 0.8073311903648347,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.558,
+      "step": 38
+    },
+    {
+      "epoch": 0.04554079696394687,
+      "grad_norm": 0.7777749596916469,
+      "learning_rate": 4.5348837209302326e-06,
+      "loss": 0.5546,
+      "step": 39
+    },
+    {
+      "epoch": 0.046708509706612174,
+      "grad_norm": 0.8588392375171646,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.5523,
+      "step": 40
+    },
+    {
+      "epoch": 0.04787622244927748,
+      "grad_norm": 0.8526787335737195,
+      "learning_rate": 4.767441860465117e-06,
+      "loss": 0.5928,
+      "step": 41
+    },
+    {
+      "epoch": 0.049043935191942785,
+      "grad_norm": 0.8284268095403576,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.561,
+      "step": 42
+    },
+    {
+      "epoch": 0.050211647934608084,
+      "grad_norm": 0.7906090287722494,
+      "learning_rate": 5e-06,
+      "loss": 0.5935,
+      "step": 43
+    },
+    {
+      "epoch": 0.05137936067727339,
+      "grad_norm": 0.7275232825323641,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.5157,
+      "step": 44
+    },
+    {
+      "epoch": 0.052547073419938695,
+      "grad_norm": 0.7956003664894934,
+      "learning_rate": 5.232558139534885e-06,
+      "loss": 0.5716,
+      "step": 45
+    },
+    {
+      "epoch": 0.053714786162604,
+      "grad_norm": 0.8247869529376962,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.5234,
+      "step": 46
+    },
+    {
+      "epoch": 0.054882498905269306,
+      "grad_norm": 0.7767405840347296,
+      "learning_rate": 5.465116279069767e-06,
+      "loss": 0.5306,
+      "step": 47
+    },
+    {
+      "epoch": 0.05605021164793461,
+      "grad_norm": 0.7297390390653535,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.5389,
+      "step": 48
+    },
+    {
+      "epoch": 0.05721792439059991,
+      "grad_norm": 0.7280729493753395,
+      "learning_rate": 5.697674418604652e-06,
+      "loss": 0.5452,
+      "step": 49
+    },
+    {
+      "epoch": 0.058385637133265215,
+      "grad_norm": 0.7733732384356662,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.5104,
+      "step": 50
+    },
+    {
+      "epoch": 0.05955334987593052,
+      "grad_norm": 0.7553817406913257,
+      "learning_rate": 5.930232558139536e-06,
+      "loss": 0.525,
+      "step": 51
+    },
+    {
+      "epoch": 0.06072106261859583,
+      "grad_norm": 0.7238237752690643,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.4814,
+      "step": 52
+    },
+    {
+      "epoch": 0.06188877536126113,
+      "grad_norm": 0.7164198548589287,
+      "learning_rate": 6.162790697674419e-06,
+      "loss": 0.5159,
+      "step": 53
+    },
+    {
+      "epoch": 0.06305648810392643,
+      "grad_norm": 0.7689212669798277,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.5279,
+      "step": 54
+    },
+    {
+      "epoch": 0.06422420084659174,
+      "grad_norm": 0.7318695597112662,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 0.4867,
+      "step": 55
+    },
+    {
+      "epoch": 0.06539191358925704,
+      "grad_norm": 0.8033496996816557,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.4942,
+      "step": 56
+    },
+    {
+      "epoch": 0.06655962633192235,
+      "grad_norm": 0.7605821974677298,
+      "learning_rate": 6.627906976744186e-06,
+      "loss": 0.5356,
+      "step": 57
+    },
+    {
+      "epoch": 0.06772733907458765,
+      "grad_norm": 0.7601928975636058,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.5046,
+      "step": 58
+    },
+    {
+      "epoch": 0.06889505181725296,
+      "grad_norm": 0.7599560548238734,
+      "learning_rate": 6.86046511627907e-06,
+      "loss": 0.532,
+      "step": 59
+    },
+    {
+      "epoch": 0.07006276455991826,
+      "grad_norm": 0.7594831339130473,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.4876,
+      "step": 60
+    },
+    {
+      "epoch": 0.07123047730258357,
+      "grad_norm": 0.7660285501295816,
+      "learning_rate": 7.0930232558139545e-06,
+      "loss": 0.5171,
+      "step": 61
+    },
+    {
+      "epoch": 0.07239819004524888,
+      "grad_norm": 0.7608094446347471,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.5348,
+      "step": 62
+    },
+    {
+      "epoch": 0.07356590278791417,
+      "grad_norm": 0.7487560526364506,
+      "learning_rate": 7.325581395348837e-06,
+      "loss": 0.5279,
+      "step": 63
+    },
+    {
+      "epoch": 0.07473361553057947,
+      "grad_norm": 0.7296290974034273,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.499,
+      "step": 64
+    },
+    {
+      "epoch": 0.07590132827324478,
+      "grad_norm": 0.8023779040180614,
+      "learning_rate": 7.5581395348837215e-06,
+      "loss": 0.5309,
+      "step": 65
+    },
+    {
+      "epoch": 0.07706904101591008,
+      "grad_norm": 0.7831288722572679,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.5229,
+      "step": 66
+    },
+    {
+      "epoch": 0.07823675375857539,
+      "grad_norm": 0.740525316516137,
+      "learning_rate": 7.790697674418605e-06,
+      "loss": 0.5148,
+      "step": 67
+    },
+    {
+      "epoch": 0.0794044665012407,
+      "grad_norm": 0.7447190045112109,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.4979,
+      "step": 68
+    },
+    {
+      "epoch": 0.080572179243906,
+      "grad_norm": 0.8320182341303636,
+      "learning_rate": 8.023255813953488e-06,
+      "loss": 0.5412,
+      "step": 69
+    },
+    {
+      "epoch": 0.0817398919865713,
+      "grad_norm": 0.7770533483386607,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.5225,
+      "step": 70
+    },
+    {
+      "epoch": 0.08290760472923661,
+      "grad_norm": 0.7938488710236006,
+      "learning_rate": 8.255813953488374e-06,
+      "loss": 0.4587,
+      "step": 71
+    },
+    {
+      "epoch": 0.08407531747190192,
+      "grad_norm": 0.8017107798734227,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.5183,
+      "step": 72
+    },
+    {
+      "epoch": 0.08524303021456722,
+      "grad_norm": 0.7469207989044206,
+      "learning_rate": 8.488372093023256e-06,
+      "loss": 0.4728,
+      "step": 73
+    },
+    {
+      "epoch": 0.08641074295723251,
+      "grad_norm": 0.7315388202220444,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.4603,
+      "step": 74
+    },
+    {
+      "epoch": 0.08757845569989782,
+      "grad_norm": 0.7598325225327908,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 0.4797,
+      "step": 75
+    },
+    {
+      "epoch": 0.08874616844256313,
+      "grad_norm": 0.7197690318425074,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.4858,
+      "step": 76
+    },
+    {
+      "epoch": 0.08991388118522843,
+      "grad_norm": 0.7471705500776821,
+      "learning_rate": 8.953488372093024e-06,
+      "loss": 0.4909,
+      "step": 77
+    },
+    {
+      "epoch": 0.09108159392789374,
+      "grad_norm": 0.7302685220708665,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.4851,
+      "step": 78
+    },
+    {
+      "epoch": 0.09224930667055904,
+      "grad_norm": 0.7802546087964382,
+      "learning_rate": 9.186046511627908e-06,
+      "loss": 0.5138,
+      "step": 79
+    },
+    {
+      "epoch": 0.09341701941322435,
+      "grad_norm": 0.7243276010947527,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.5223,
+      "step": 80
+    },
+    {
+      "epoch": 0.09458473215588965,
+      "grad_norm": 0.7823042026924807,
+      "learning_rate": 9.418604651162791e-06,
+      "loss": 0.4935,
+      "step": 81
+    },
+    {
+      "epoch": 0.09575244489855496,
+      "grad_norm": 0.7467470861203891,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.4933,
+      "step": 82
+    },
+    {
+      "epoch": 0.09692015764122026,
+      "grad_norm": 0.7740867537297575,
+      "learning_rate": 9.651162790697676e-06,
+      "loss": 0.468,
+      "step": 83
+    },
+    {
+      "epoch": 0.09808787038388557,
+      "grad_norm": 0.8103069550982679,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.512,
+      "step": 84
+    },
+    {
+      "epoch": 0.09925558312655088,
+      "grad_norm": 0.765124197439174,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 0.4821,
+      "step": 85
+    },
+    {
+      "epoch": 0.10042329586921617,
+      "grad_norm": 0.9061261053610111,
+      "learning_rate": 1e-05,
+      "loss": 0.4795,
+      "step": 86
+    },
+    {
+      "epoch": 0.10159100861188147,
+      "grad_norm": 0.778167992234207,
+      "learning_rate": 1.0116279069767443e-05,
+      "loss": 0.5228,
+      "step": 87
+    },
+    {
+      "epoch": 0.10275872135454678,
+      "grad_norm": 0.7538145547009709,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 0.4794,
+      "step": 88
+    },
+    {
+      "epoch": 0.10392643409721208,
+      "grad_norm": 0.8474409226159929,
+      "learning_rate": 1.0348837209302327e-05,
+      "loss": 0.488,
+      "step": 89
+    },
+    {
+      "epoch": 0.10509414683987739,
+      "grad_norm": 0.7262012188790375,
+      "learning_rate": 1.046511627906977e-05,
+      "loss": 0.4771,
+      "step": 90
+    },
+    {
+      "epoch": 0.1062618595825427,
+      "grad_norm": 0.7931901554348757,
+      "learning_rate": 1.058139534883721e-05,
+      "loss": 0.5256,
+      "step": 91
+    },
+    {
+      "epoch": 0.107429572325208,
+      "grad_norm": 0.7667972840804032,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 0.4786,
+      "step": 92
+    },
+    {
+      "epoch": 0.1085972850678733,
+      "grad_norm": 0.7167299849823746,
+      "learning_rate": 1.0813953488372094e-05,
+      "loss": 0.4924,
+      "step": 93
+    },
+    {
+      "epoch": 0.10976499781053861,
+      "grad_norm": 0.8152897734453837,
+      "learning_rate": 1.0930232558139535e-05,
+      "loss": 0.478,
+      "step": 94
+    },
+    {
+      "epoch": 0.11093271055320392,
+      "grad_norm": 0.7628887742594963,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 0.4668,
+      "step": 95
+    },
+    {
+      "epoch": 0.11210042329586922,
+      "grad_norm": 0.7718427854495195,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 0.5057,
+      "step": 96
+    },
+    {
+      "epoch": 0.11326813603853451,
+      "grad_norm": 0.7993734486714776,
+      "learning_rate": 1.1279069767441861e-05,
+      "loss": 0.4951,
+      "step": 97
+    },
+    {
+      "epoch": 0.11443584878119982,
+      "grad_norm": 0.7273043245636739,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 0.4741,
+      "step": 98
+    },
+    {
+      "epoch": 0.11560356152386513,
+      "grad_norm": 0.7171974787133631,
+      "learning_rate": 1.1511627906976746e-05,
+      "loss": 0.486,
+      "step": 99
+    },
+    {
+      "epoch": 0.11677127426653043,
+      "grad_norm": 0.7523830129457043,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.4883,
+      "step": 100
+    },
+    {
+      "epoch": 0.11793898700919574,
+      "grad_norm": 0.754771318399461,
+      "learning_rate": 1.174418604651163e-05,
+      "loss": 0.4637,
+      "step": 101
+    },
+    {
+      "epoch": 0.11910669975186104,
+      "grad_norm": 0.8277750877947945,
+      "learning_rate": 1.1860465116279072e-05,
+      "loss": 0.4888,
+      "step": 102
+    },
+    {
+      "epoch": 0.12027441249452635,
+      "grad_norm": 0.7742341608761458,
+      "learning_rate": 1.1976744186046511e-05,
+      "loss": 0.4587,
+      "step": 103
+    },
+    {
+      "epoch": 0.12144212523719165,
+      "grad_norm": 0.726627838676463,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 0.4459,
+      "step": 104
+    },
+    {
+      "epoch": 0.12260983797985696,
+      "grad_norm": 0.83069761432513,
+      "learning_rate": 1.2209302325581397e-05,
+      "loss": 0.4654,
+      "step": 105
+    },
+    {
+      "epoch": 0.12377755072252226,
+      "grad_norm": 0.7018891835437143,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 0.4666,
+      "step": 106
+    },
+    {
+      "epoch": 0.12494526346518757,
+      "grad_norm": 0.7949493245103832,
+      "learning_rate": 1.244186046511628e-05,
+      "loss": 0.4751,
+      "step": 107
+    },
+    {
+      "epoch": 0.12611297620785286,
+      "grad_norm": 0.8329409810432516,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 0.522,
+      "step": 108
+    },
+    {
+      "epoch": 0.12728068895051817,
+      "grad_norm": 0.7148966383270693,
+      "learning_rate": 1.2674418604651164e-05,
+      "loss": 0.5122,
+      "step": 109
+    },
+    {
+      "epoch": 0.12844840169318347,
+      "grad_norm": 0.8208978608978629,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 0.5025,
+      "step": 110
+    },
+    {
+      "epoch": 0.12961611443584878,
+      "grad_norm": 0.7262122866653611,
+      "learning_rate": 1.2906976744186049e-05,
+      "loss": 0.5105,
+      "step": 111
+    },
+    {
+      "epoch": 0.13078382717851408,
+      "grad_norm": 0.7407655234445234,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 0.4751,
+      "step": 112
+    },
+    {
+      "epoch": 0.1319515399211794,
+      "grad_norm": 0.7795544686630245,
+      "learning_rate": 1.313953488372093e-05,
+      "loss": 0.5055,
+      "step": 113
+    },
+    {
+      "epoch": 0.1331192526638447,
+      "grad_norm": 0.7811531241162247,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 0.485,
+      "step": 114
+    },
+    {
+      "epoch": 0.13428696540651,
+      "grad_norm": 0.8707907790715469,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 0.517,
+      "step": 115
+    },
+    {
+      "epoch": 0.1354546781491753,
+      "grad_norm": 0.7268430498575351,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 0.4799,
+      "step": 116
+    },
+    {
+      "epoch": 0.1366223908918406,
+      "grad_norm": 0.8229147405053716,
+      "learning_rate": 1.36046511627907e-05,
+      "loss": 0.4972,
+      "step": 117
+    },
+    {
+      "epoch": 0.13779010363450592,
+      "grad_norm": 0.7583738990369542,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 0.4734,
+      "step": 118
+    },
+    {
+      "epoch": 0.13895781637717122,
+      "grad_norm": 0.760203613427415,
+      "learning_rate": 1.3837209302325583e-05,
+      "loss": 0.4965,
+      "step": 119
+    },
+    {
+      "epoch": 0.14012552911983653,
+      "grad_norm": 0.8064036575163311,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 0.4751,
+      "step": 120
+    },
+    {
+      "epoch": 0.14129324186250183,
+      "grad_norm": 0.7648966579497968,
+      "learning_rate": 1.4069767441860466e-05,
+      "loss": 0.4985,
+      "step": 121
+    },
+    {
+      "epoch": 0.14246095460516714,
+      "grad_norm": 0.7894308656877415,
+      "learning_rate": 1.4186046511627909e-05,
+      "loss": 0.4917,
+      "step": 122
+    },
+    {
+      "epoch": 0.14362866734783244,
+      "grad_norm": 0.8534620161520874,
+      "learning_rate": 1.4302325581395352e-05,
+      "loss": 0.4806,
+      "step": 123
+    },
+    {
+      "epoch": 0.14479638009049775,
+      "grad_norm": 0.774706649909284,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 0.4941,
+      "step": 124
+    },
+    {
+      "epoch": 0.14596409283316303,
+      "grad_norm": 0.8359880696743944,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 0.4829,
+      "step": 125
+    },
+    {
+      "epoch": 0.14713180557582833,
+      "grad_norm": 0.802288022042228,
+      "learning_rate": 1.4651162790697674e-05,
+      "loss": 0.4963,
+      "step": 126
+    },
+    {
+      "epoch": 0.14829951831849364,
+      "grad_norm": 0.7645377136880702,
+      "learning_rate": 1.4767441860465117e-05,
+      "loss": 0.4901,
+      "step": 127
+    },
+    {
+      "epoch": 0.14946723106115894,
+      "grad_norm": 0.8081433719407812,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 0.4745,
+      "step": 128
+    },
+    {
+      "epoch": 0.15063494380382425,
+      "grad_norm": 0.8395750979545705,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4837,
+      "step": 129
+    },
+    {
+      "epoch": 0.15180265654648956,
+      "grad_norm": 0.8785533818560177,
+      "learning_rate": 1.5116279069767443e-05,
+      "loss": 0.4709,
+      "step": 130
+    },
+    {
+      "epoch": 0.15297036928915486,
+      "grad_norm": 0.775901157067453,
+      "learning_rate": 1.5232558139534886e-05,
+      "loss": 0.4676,
+      "step": 131
+    },
+    {
+      "epoch": 0.15413808203182017,
+      "grad_norm": 0.8527855459854962,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 0.4742,
+      "step": 132
+    },
+    {
+      "epoch": 0.15530579477448547,
+      "grad_norm": 0.8216619380870416,
+      "learning_rate": 1.546511627906977e-05,
+      "loss": 0.4695,
+      "step": 133
+    },
+    {
+      "epoch": 0.15647350751715078,
+      "grad_norm": 0.8447156806357727,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 0.493,
+      "step": 134
+    },
+    {
+      "epoch": 0.15764122025981608,
+      "grad_norm": 0.7554358305947614,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 0.4827,
+      "step": 135
+    },
+    {
+      "epoch": 0.1588089330024814,
+      "grad_norm": 0.8202294856372566,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 0.4786,
+      "step": 136
+    },
+    {
+      "epoch": 0.1599766457451467,
+      "grad_norm": 0.7315916473519791,
+      "learning_rate": 1.5930232558139536e-05,
+      "loss": 0.4509,
+      "step": 137
+    },
+    {
+      "epoch": 0.161144358487812,
+      "grad_norm": 2.119604897197923,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 0.5086,
+      "step": 138
+    },
+    {
+      "epoch": 0.1623120712304773,
+      "grad_norm": 9.527500988494516,
+      "learning_rate": 1.616279069767442e-05,
+      "loss": 0.5722,
+      "step": 139
+    },
+    {
+      "epoch": 0.1634797839731426,
+      "grad_norm": 4.355186823389993,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.4924,
+      "step": 140
+    },
+    {
+      "epoch": 0.16464749671580792,
+      "grad_norm": 1.1160401681400667,
+      "learning_rate": 1.6395348837209303e-05,
+      "loss": 0.4719,
+      "step": 141
+    },
+    {
+      "epoch": 0.16581520945847322,
+      "grad_norm": 3.186485365068782,
+      "learning_rate": 1.6511627906976747e-05,
+      "loss": 0.5031,
+      "step": 142
+    },
+    {
+      "epoch": 0.16698292220113853,
+      "grad_norm": 1.4688820326372403,
+      "learning_rate": 1.6627906976744188e-05,
+      "loss": 0.4803,
+      "step": 143
+    },
+    {
+      "epoch": 0.16815063494380383,
+      "grad_norm": 3.1471517885543094,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 0.5256,
+      "step": 144
+    },
+    {
+      "epoch": 0.16931834768646914,
+      "grad_norm": 1.0565094290414523,
+      "learning_rate": 1.6860465116279073e-05,
+      "loss": 0.4768,
+      "step": 145
+    },
+    {
+      "epoch": 0.17048606042913445,
+      "grad_norm": 2.302534455729968,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 0.5294,
+      "step": 146
+    },
+    {
+      "epoch": 0.17165377317179975,
+      "grad_norm": 1.4224434846517153,
+      "learning_rate": 1.7093023255813955e-05,
+      "loss": 0.4762,
+      "step": 147
+    },
+    {
+      "epoch": 0.17282148591446503,
+      "grad_norm": 7.069981347837623,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.5071,
+      "step": 148
+    },
+    {
+      "epoch": 0.17398919865713033,
+      "grad_norm": 4.232679030435236,
+      "learning_rate": 1.7325581395348837e-05,
+      "loss": 0.432,
+      "step": 149
+    },
+    {
+      "epoch": 0.17515691139979564,
+      "grad_norm": 3.709254552243265,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 0.4968,
+      "step": 150
+    },
+    {
+      "epoch": 0.17632462414246095,
+      "grad_norm": 1.4062366539178452,
+      "learning_rate": 1.7558139534883722e-05,
+      "loss": 0.4998,
+      "step": 151
+    },
+    {
+      "epoch": 0.17749233688512625,
+      "grad_norm": 0.9092021343335508,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 0.4801,
+      "step": 152
+    },
+    {
+      "epoch": 0.17866004962779156,
+      "grad_norm": 1.1960283844587885,
+      "learning_rate": 1.7790697674418608e-05,
+      "loss": 0.5147,
+      "step": 153
+    },
+    {
+      "epoch": 0.17982776237045686,
+      "grad_norm": 0.9703655262620353,
+      "learning_rate": 1.790697674418605e-05,
+      "loss": 0.4616,
+      "step": 154
+    },
+    {
+      "epoch": 0.18099547511312217,
+      "grad_norm": 0.9720569450445637,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 0.4312,
+      "step": 155
+    },
+    {
+      "epoch": 0.18216318785578747,
+      "grad_norm": 1.139983318608259,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 0.4589,
+      "step": 156
+    },
+    {
+      "epoch": 0.18333090059845278,
+      "grad_norm": 0.8525713127642296,
+      "learning_rate": 1.825581395348837e-05,
+      "loss": 0.5054,
+      "step": 157
+    },
+    {
+      "epoch": 0.18449861334111808,
+      "grad_norm": 0.9281643656726715,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 0.4689,
+      "step": 158
+    },
+    {
+      "epoch": 0.1856663260837834,
+      "grad_norm": 0.973550005511157,
+      "learning_rate": 1.8488372093023256e-05,
+      "loss": 0.467,
+      "step": 159
+    },
+    {
+      "epoch": 0.1868340388264487,
+      "grad_norm": 0.7724996723890557,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 0.4942,
+      "step": 160
+    },
+    {
+      "epoch": 0.188001751569114,
+      "grad_norm": 0.7803969291300282,
+      "learning_rate": 1.872093023255814e-05,
+      "loss": 0.438,
+      "step": 161
+    },
+    {
+      "epoch": 0.1891694643117793,
+      "grad_norm": 0.7728568751561969,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 0.4403,
+      "step": 162
+    },
+    {
+      "epoch": 0.1903371770544446,
+      "grad_norm": 0.8421944523506198,
+      "learning_rate": 1.8953488372093027e-05,
+      "loss": 0.4977,
+      "step": 163
+    },
+    {
+      "epoch": 0.19150488979710992,
+      "grad_norm": 0.7902049729149769,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.478,
+      "step": 164
+    },
+    {
+      "epoch": 0.19267260253977522,
+      "grad_norm": 0.8036181261476156,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 0.4947,
+      "step": 165
+    },
+    {
+      "epoch": 0.19384031528244053,
+      "grad_norm": 0.7930904069630152,
+      "learning_rate": 1.9302325581395353e-05,
+      "loss": 0.4933,
+      "step": 166
+    },
+    {
+      "epoch": 0.19500802802510583,
+      "grad_norm": 0.7746721911011534,
+      "learning_rate": 1.941860465116279e-05,
+      "loss": 0.5104,
+      "step": 167
+    },
+    {
+      "epoch": 0.19617574076777114,
+      "grad_norm": 0.711761296019644,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 0.4696,
+      "step": 168
+    },
+    {
+      "epoch": 0.19734345351043645,
+      "grad_norm": 0.8291210091629917,
+      "learning_rate": 1.9651162790697676e-05,
+      "loss": 0.4889,
+      "step": 169
+    },
+    {
+      "epoch": 0.19851116625310175,
+      "grad_norm": 0.7399145847969631,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.4684,
+      "step": 170
+    },
+    {
+      "epoch": 0.19967887899576703,
+      "grad_norm": 0.7469696263122528,
+      "learning_rate": 1.988372093023256e-05,
+      "loss": 0.5038,
+      "step": 171
+    },
+    {
+      "epoch": 0.20084659173843233,
+      "grad_norm": 0.8689785689891113,
+      "learning_rate": 2e-05,
+      "loss": 0.5108,
+      "step": 172
+    },
+    {
+      "epoch": 0.20201430448109764,
+      "grad_norm": 0.7530499758226269,
+      "learning_rate": 1.9999979192104532e-05,
+      "loss": 0.5,
+      "step": 173
+    },
+    {
+      "epoch": 0.20318201722376295,
+      "grad_norm": 0.7227121116177873,
+      "learning_rate": 1.9999916768504724e-05,
+      "loss": 0.4687,
+      "step": 174
+    },
+    {
+      "epoch": 0.20434972996642825,
+      "grad_norm": 0.8436104386429739,
+      "learning_rate": 1.9999812729460353e-05,
+      "loss": 0.4789,
+      "step": 175
+    },
+    {
+      "epoch": 0.20551744270909356,
+      "grad_norm": 0.754514325551959,
+      "learning_rate": 1.9999667075404385e-05,
+      "loss": 0.4691,
+      "step": 176
+    },
+    {
+      "epoch": 0.20668515545175886,
+      "grad_norm": 0.8047448260707747,
+      "learning_rate": 1.9999479806942976e-05,
+      "loss": 0.4822,
+      "step": 177
+    },
+    {
+      "epoch": 0.20785286819442417,
+      "grad_norm": 0.7990775161801204,
+      "learning_rate": 1.9999250924855458e-05,
+      "loss": 0.4667,
+      "step": 178
+    },
+    {
+      "epoch": 0.20902058093708947,
+      "grad_norm": 0.7144387127913208,
+      "learning_rate": 1.9998980430094333e-05,
+      "loss": 0.4672,
+      "step": 179
+    },
+    {
+      "epoch": 0.21018829367975478,
+      "grad_norm": 0.7145625716323254,
+      "learning_rate": 1.9998668323785298e-05,
+      "loss": 0.4356,
+      "step": 180
+    },
+    {
+      "epoch": 0.21135600642242008,
+      "grad_norm": 0.7409030752947819,
+      "learning_rate": 1.9998314607227203e-05,
+      "loss": 0.4825,
+      "step": 181
+    },
+    {
+      "epoch": 0.2125237191650854,
+      "grad_norm": 0.7772810994534644,
+      "learning_rate": 1.9997919281892066e-05,
+      "loss": 0.4518,
+      "step": 182
+    },
+    {
+      "epoch": 0.2136914319077507,
+      "grad_norm": 0.76827264611303,
+      "learning_rate": 1.999748234942507e-05,
+      "loss": 0.4795,
+      "step": 183
+    },
+    {
+      "epoch": 0.214859144650416,
+      "grad_norm": 0.7968726730980179,
+      "learning_rate": 1.9997003811644534e-05,
+      "loss": 0.4851,
+      "step": 184
+    },
+    {
+      "epoch": 0.2160268573930813,
+      "grad_norm": 0.7784677678487992,
+      "learning_rate": 1.999648367054194e-05,
+      "loss": 0.4989,
+      "step": 185
+    },
+    {
+      "epoch": 0.2171945701357466,
+      "grad_norm": 0.7618128942058454,
+      "learning_rate": 1.9995921928281893e-05,
+      "loss": 0.4893,
+      "step": 186
+    },
+    {
+      "epoch": 0.21836228287841192,
+      "grad_norm": 0.7699566034393192,
+      "learning_rate": 1.9995318587202132e-05,
+      "loss": 0.4654,
+      "step": 187
+    },
+    {
+      "epoch": 0.21952999562107722,
+      "grad_norm": 0.7689051541323366,
+      "learning_rate": 1.99946736498135e-05,
+      "loss": 0.4741,
+      "step": 188
+    },
+    {
+      "epoch": 0.22069770836374253,
+      "grad_norm": 0.7642886618828265,
+      "learning_rate": 1.999398711879996e-05,
+      "loss": 0.492,
+      "step": 189
+    },
+    {
+      "epoch": 0.22186542110640783,
+      "grad_norm": 0.8620831536968766,
+      "learning_rate": 1.9993258997018564e-05,
+      "loss": 0.4672,
+      "step": 190
+    },
+    {
+      "epoch": 0.22303313384907314,
+      "grad_norm": 0.7507174789270951,
+      "learning_rate": 1.9992489287499454e-05,
+      "loss": 0.4576,
+      "step": 191
+    },
+    {
+      "epoch": 0.22420084659173845,
+      "grad_norm": 0.8359644912383936,
+      "learning_rate": 1.9991677993445832e-05,
+      "loss": 0.4769,
+      "step": 192
+    },
+    {
+      "epoch": 0.22536855933440372,
+      "grad_norm": 0.7845794269054861,
+      "learning_rate": 1.9990825118233958e-05,
+      "loss": 0.4435,
+      "step": 193
+    },
+    {
+      "epoch": 0.22653627207706903,
+      "grad_norm": 0.7364043570445138,
+      "learning_rate": 1.9989930665413148e-05,
+      "loss": 0.4385,
+      "step": 194
+    },
+    {
+      "epoch": 0.22770398481973433,
+      "grad_norm": 0.7879757131086841,
+      "learning_rate": 1.9988994638705734e-05,
+      "loss": 0.485,
+      "step": 195
+    },
+    {
+      "epoch": 0.22887169756239964,
+      "grad_norm": 0.8035636652589458,
+      "learning_rate": 1.9988017042007066e-05,
+      "loss": 0.4647,
+      "step": 196
+    },
+    {
+      "epoch": 0.23003941030506495,
+      "grad_norm": 0.7766298498418136,
+      "learning_rate": 1.9986997879385488e-05,
+      "loss": 0.466,
+      "step": 197
+    },
+    {
+      "epoch": 0.23120712304773025,
+      "grad_norm": 0.7796285597120038,
+      "learning_rate": 1.998593715508233e-05,
+      "loss": 0.4706,
+      "step": 198
+    },
+    {
+      "epoch": 0.23237483579039556,
+      "grad_norm": 0.7363727814372553,
+      "learning_rate": 1.9984834873511873e-05,
+      "loss": 0.4782,
+      "step": 199
+    },
+    {
+      "epoch": 0.23354254853306086,
+      "grad_norm": 0.8206908089342946,
+      "learning_rate": 1.9983691039261358e-05,
+      "loss": 0.4739,
+      "step": 200
+    },
+    {
+      "epoch": 0.23471026127572617,
+      "grad_norm": 0.7687109812474855,
+      "learning_rate": 1.9982505657090936e-05,
+      "loss": 0.4937,
+      "step": 201
+    },
+    {
+      "epoch": 0.23587797401839147,
+      "grad_norm": 0.7778605200657507,
+      "learning_rate": 1.998127873193367e-05,
+      "loss": 0.4718,
+      "step": 202
+    },
+    {
+      "epoch": 0.23704568676105678,
+      "grad_norm": 0.7752985541048024,
+      "learning_rate": 1.9980010268895507e-05,
+      "loss": 0.4759,
+      "step": 203
+    },
+    {
+      "epoch": 0.23821339950372208,
+      "grad_norm": 0.7214523421124862,
+      "learning_rate": 1.9978700273255254e-05,
+      "loss": 0.5243,
+      "step": 204
+    },
+    {
+      "epoch": 0.2393811122463874,
+      "grad_norm": 0.8267853577501512,
+      "learning_rate": 1.997734875046456e-05,
+      "loss": 0.4668,
+      "step": 205
+    },
+    {
+      "epoch": 0.2405488249890527,
+      "grad_norm": 0.7254140905087999,
+      "learning_rate": 1.9975955706147902e-05,
+      "loss": 0.4851,
+      "step": 206
+    },
+    {
+      "epoch": 0.241716537731718,
+      "grad_norm": 0.7709805019965509,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.4736,
+      "step": 207
+    },
+    {
+      "epoch": 0.2428842504743833,
+      "grad_norm": 0.7680838826374472,
+      "learning_rate": 1.99730450762985e-05,
+      "loss": 0.4426,
+      "step": 208
+    },
+    {
+      "epoch": 0.2440519632170486,
+      "grad_norm": 0.8048628478355198,
+      "learning_rate": 1.9971527502878576e-05,
+      "loss": 0.4827,
+      "step": 209
+    },
+    {
+      "epoch": 0.24521967595971392,
+      "grad_norm": 0.7773043090178662,
+      "learning_rate": 1.9969968432158266e-05,
+      "loss": 0.446,
+      "step": 210
+    },
+    {
+      "epoch": 0.24638738870237922,
+      "grad_norm": 0.7670232679862734,
+      "learning_rate": 1.9968367870625762e-05,
+      "loss": 0.468,
+      "step": 211
+    },
+    {
+      "epoch": 0.24755510144504453,
+      "grad_norm": 0.7964341369422844,
+      "learning_rate": 1.9966725824941933e-05,
+      "loss": 0.4847,
+      "step": 212
+    },
+    {
+      "epoch": 0.24872281418770983,
+      "grad_norm": 0.7265060930031542,
+      "learning_rate": 1.996504230194028e-05,
+      "loss": 0.4899,
+      "step": 213
+    },
+    {
+      "epoch": 0.24989052693037514,
+      "grad_norm": 0.8005182587622003,
+      "learning_rate": 1.9963317308626916e-05,
+      "loss": 0.5106,
+      "step": 214
+    },
+    {
+      "epoch": 0.2510582396730404,
+      "grad_norm": 0.7227742624841775,
+      "learning_rate": 1.9961550852180535e-05,
+      "loss": 0.4812,
+      "step": 215
+    },
+    {
+      "epoch": 0.2522259524157057,
+      "grad_norm": 0.7500551803007318,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.4645,
+      "step": 216
+    },
+    {
+      "epoch": 0.25339366515837103,
+      "grad_norm": 0.7006902581646431,
+      "learning_rate": 1.9957893579466252e-05,
+      "loss": 0.455,
+      "step": 217
+    },
+    {
+      "epoch": 0.25456137790103633,
+      "grad_norm": 0.7477340605178123,
+      "learning_rate": 1.995600277841837e-05,
+      "loss": 0.4639,
+      "step": 218
+    },
+    {
+      "epoch": 0.25572909064370164,
+      "grad_norm": 0.8317071937088114,
+      "learning_rate": 1.9954070544677476e-05,
+      "loss": 0.4941,
+      "step": 219
+    },
+    {
+      "epoch": 0.25689680338636695,
+      "grad_norm": 0.7883424767737056,
+      "learning_rate": 1.995209688628471e-05,
+      "loss": 0.4669,
+      "step": 220
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.8089967193764374,
+      "learning_rate": 1.9950081811453598e-05,
+      "loss": 0.4668,
+      "step": 221
+    },
+    {
+      "epoch": 0.25923222887169756,
+      "grad_norm": 0.728768928323264,
+      "learning_rate": 1.994802532857004e-05,
+      "loss": 0.49,
+      "step": 222
+    },
+    {
+      "epoch": 0.26039994161436286,
+      "grad_norm": 0.7805174987688834,
+      "learning_rate": 1.9945927446192256e-05,
+      "loss": 0.4706,
+      "step": 223
+    },
+    {
+      "epoch": 0.26156765435702817,
+      "grad_norm": 0.8082384190203897,
+      "learning_rate": 1.9943788173050743e-05,
+      "loss": 0.5143,
+      "step": 224
+    },
+    {
+      "epoch": 0.2627353670996935,
+      "grad_norm": 0.7642405210098879,
+      "learning_rate": 1.994160751804826e-05,
+      "loss": 0.4643,
+      "step": 225
+    },
+    {
+      "epoch": 0.2639030798423588,
+      "grad_norm": 0.8189664591372509,
+      "learning_rate": 1.9939385490259773e-05,
+      "loss": 0.5034,
+      "step": 226
+    },
+    {
+      "epoch": 0.2650707925850241,
+      "grad_norm": 0.7810569875869899,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 0.4661,
+      "step": 227
+    },
+    {
+      "epoch": 0.2662385053276894,
+      "grad_norm": 0.7474483143603757,
+      "learning_rate": 1.99348173534855e-05,
+      "loss": 0.4471,
+      "step": 228
+    },
+    {
+      "epoch": 0.2674062180703547,
+      "grad_norm": 0.729392927896885,
+      "learning_rate": 1.9932471263510385e-05,
+      "loss": 0.4946,
+      "step": 229
+    },
+    {
+      "epoch": 0.26857393081302,
+      "grad_norm": 0.7475077207479482,
+      "learning_rate": 1.9930083838770504e-05,
+      "loss": 0.4732,
+      "step": 230
+    },
+    {
+      "epoch": 0.2697416435556853,
+      "grad_norm": 0.7215732939312259,
+      "learning_rate": 1.9927655089201327e-05,
+      "loss": 0.4357,
+      "step": 231
+    },
+    {
+      "epoch": 0.2709093562983506,
+      "grad_norm": 0.7280417972188489,
+      "learning_rate": 1.992518502491028e-05,
+      "loss": 0.4684,
+      "step": 232
+    },
+    {
+      "epoch": 0.2720770690410159,
+      "grad_norm": 0.7577411472224214,
+      "learning_rate": 1.9922673656176733e-05,
+      "loss": 0.4655,
+      "step": 233
+    },
+    {
+      "epoch": 0.2732447817836812,
+      "grad_norm": 0.7398046350817632,
+      "learning_rate": 1.9920120993451947e-05,
+      "loss": 0.4758,
+      "step": 234
+    },
+    {
+      "epoch": 0.27441249452634653,
+      "grad_norm": 0.7491622038317146,
+      "learning_rate": 1.991752704735903e-05,
+      "loss": 0.4842,
+      "step": 235
+    },
+    {
+      "epoch": 0.27558020726901183,
+      "grad_norm": 0.7101091468645754,
+      "learning_rate": 1.991489182869289e-05,
+      "loss": 0.4729,
+      "step": 236
+    },
+    {
+      "epoch": 0.27674792001167714,
+      "grad_norm": 0.6886932654250374,
+      "learning_rate": 1.99122153484202e-05,
+      "loss": 0.4358,
+      "step": 237
+    },
+    {
+      "epoch": 0.27791563275434245,
+      "grad_norm": 0.7640547747447162,
+      "learning_rate": 1.990949761767935e-05,
+      "loss": 0.4495,
+      "step": 238
+    },
+    {
+      "epoch": 0.27908334549700775,
+      "grad_norm": 0.7748899162697885,
+      "learning_rate": 1.9906738647780384e-05,
+      "loss": 0.477,
+      "step": 239
+    },
+    {
+      "epoch": 0.28025105823967306,
+      "grad_norm": 0.6550151306429578,
+      "learning_rate": 1.9903938450204972e-05,
+      "loss": 0.4442,
+      "step": 240
+    },
+    {
+      "epoch": 0.28141877098233836,
+      "grad_norm": 0.7700348262436073,
+      "learning_rate": 1.990109703660637e-05,
+      "loss": 0.4803,
+      "step": 241
+    },
+    {
+      "epoch": 0.28258648372500367,
+      "grad_norm": 0.6561126962316909,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4264,
+      "step": 242
+    },
+    {
+      "epoch": 0.283754196467669,
+      "grad_norm": 0.7497613735583678,
+      "learning_rate": 1.9895290608810103e-05,
+      "loss": 0.478,
+      "step": 243
+    },
+    {
+      "epoch": 0.2849219092103343,
+      "grad_norm": 0.659211471388419,
+      "learning_rate": 1.9892325618776353e-05,
+      "loss": 0.4372,
+      "step": 244
+    },
+    {
+      "epoch": 0.2860896219529996,
+      "grad_norm": 0.672911987820084,
+      "learning_rate": 1.988931946104712e-05,
+      "loss": 0.4797,
+      "step": 245
+    },
+    {
+      "epoch": 0.2872573346956649,
+      "grad_norm": 0.7005351644080285,
+      "learning_rate": 1.988627214813277e-05,
+      "loss": 0.4628,
+      "step": 246
+    },
+    {
+      "epoch": 0.2884250474383302,
+      "grad_norm": 0.6828551948058207,
+      "learning_rate": 1.9883183692714935e-05,
+      "loss": 0.4699,
+      "step": 247
+    },
+    {
+      "epoch": 0.2895927601809955,
+      "grad_norm": 0.7594278229765398,
+      "learning_rate": 1.9880054107646467e-05,
+      "loss": 0.4615,
+      "step": 248
+    },
+    {
+      "epoch": 0.2907604729236608,
+      "grad_norm": 0.7321225856834818,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.4812,
+      "step": 249
+    },
+    {
+      "epoch": 0.29192818566632606,
+      "grad_norm": 0.8002957386264957,
+      "learning_rate": 1.98736716008248e-05,
+      "loss": 0.4936,
+      "step": 250
+    },
+    {
+      "epoch": 0.29309589840899136,
+      "grad_norm": 0.7245652712729834,
+      "learning_rate": 1.987041870563291e-05,
+      "loss": 0.4705,
+      "step": 251
+    },
+    {
+      "epoch": 0.29426361115165667,
+      "grad_norm": 0.649023253250857,
+      "learning_rate": 1.986712473391289e-05,
+      "loss": 0.4684,
+      "step": 252
+    },
+    {
+      "epoch": 0.295431323894322,
+      "grad_norm": 0.7443947002272074,
+      "learning_rate": 1.9863789699372863e-05,
+      "loss": 0.4608,
+      "step": 253
+    },
+    {
+      "epoch": 0.2965990366369873,
+      "grad_norm": 0.6595812565629166,
+      "learning_rate": 1.986041361589184e-05,
+      "loss": 0.4482,
+      "step": 254
+    },
+    {
+      "epoch": 0.2977667493796526,
+      "grad_norm": 0.6516525598937988,
+      "learning_rate": 1.985699649751966e-05,
+      "loss": 0.4841,
+      "step": 255
+    },
+    {
+      "epoch": 0.2989344621223179,
+      "grad_norm": 0.8294717897137015,
+      "learning_rate": 1.9853538358476933e-05,
+      "loss": 0.5009,
+      "step": 256
+    },
+    {
+      "epoch": 0.3001021748649832,
+      "grad_norm": 0.6715152889109127,
+      "learning_rate": 1.9850039213154972e-05,
+      "loss": 0.4353,
+      "step": 257
+    },
+    {
+      "epoch": 0.3012698876076485,
+      "grad_norm": 0.7851156714469566,
+      "learning_rate": 1.984649907611575e-05,
+      "loss": 0.4706,
+      "step": 258
+    },
+    {
+      "epoch": 0.3024376003503138,
+      "grad_norm": 0.7778442893515388,
+      "learning_rate": 1.984291796209183e-05,
+      "loss": 0.4729,
+      "step": 259
+    },
+    {
+      "epoch": 0.3036053130929791,
+      "grad_norm": 0.6809936000508532,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.4774,
+      "step": 260
+    },
+    {
+      "epoch": 0.3047730258356444,
+      "grad_norm": 0.8644329377524221,
+      "learning_rate": 1.9835632862872712e-05,
+      "loss": 0.4758,
+      "step": 261
+    },
+    {
+      "epoch": 0.3059407385783097,
+      "grad_norm": 0.686903660333163,
+      "learning_rate": 1.9831928907995032e-05,
+      "loss": 0.4509,
+      "step": 262
+    },
+    {
+      "epoch": 0.30710845132097503,
+      "grad_norm": 0.8673334454426852,
+      "learning_rate": 1.9828184036767556e-05,
+      "loss": 0.4727,
+      "step": 263
+    },
+    {
+      "epoch": 0.30827616406364033,
+      "grad_norm": 0.7799693445551544,
+      "learning_rate": 1.982439826477487e-05,
+      "loss": 0.4817,
+      "step": 264
+    },
+    {
+      "epoch": 0.30944387680630564,
+      "grad_norm": 0.7334982597879803,
+      "learning_rate": 1.9820571607771756e-05,
+      "loss": 0.4822,
+      "step": 265
+    },
+    {
+      "epoch": 0.31061158954897095,
+      "grad_norm": 0.8339485991447102,
+      "learning_rate": 1.9816704081683152e-05,
+      "loss": 0.4275,
+      "step": 266
+    },
+    {
+      "epoch": 0.31177930229163625,
+      "grad_norm": 0.656768459068204,
+      "learning_rate": 1.9812795702604073e-05,
+      "loss": 0.448,
+      "step": 267
+    },
+    {
+      "epoch": 0.31294701503430156,
+      "grad_norm": 0.7770159884963295,
+      "learning_rate": 1.980884648679955e-05,
+      "loss": 0.451,
+      "step": 268
+    },
+    {
+      "epoch": 0.31411472777696686,
+      "grad_norm": 0.6512230730023202,
+      "learning_rate": 1.980485645070455e-05,
+      "loss": 0.4534,
+      "step": 269
+    },
+    {
+      "epoch": 0.31528244051963217,
+      "grad_norm": 0.7134932640932927,
+      "learning_rate": 1.9800825610923937e-05,
+      "loss": 0.4905,
+      "step": 270
+    },
+    {
+      "epoch": 0.3164501532622975,
+      "grad_norm": 0.6713538794477137,
+      "learning_rate": 1.9796753984232357e-05,
+      "loss": 0.4865,
+      "step": 271
+    },
+    {
+      "epoch": 0.3176178660049628,
+      "grad_norm": 0.66347182641543,
+      "learning_rate": 1.9792641587574212e-05,
+      "loss": 0.4442,
+      "step": 272
+    },
+    {
+      "epoch": 0.3187855787476281,
+      "grad_norm": 0.7299404611180915,
+      "learning_rate": 1.9788488438063566e-05,
+      "loss": 0.4703,
+      "step": 273
+    },
+    {
+      "epoch": 0.3199532914902934,
+      "grad_norm": 0.6981767308734392,
+      "learning_rate": 1.9784294552984077e-05,
+      "loss": 0.464,
+      "step": 274
+    },
+    {
+      "epoch": 0.3211210042329587,
+      "grad_norm": 0.6441845967007052,
+      "learning_rate": 1.9780059949788935e-05,
+      "loss": 0.4444,
+      "step": 275
+    },
+    {
+      "epoch": 0.322288716975624,
+      "grad_norm": 0.68355692334444,
+      "learning_rate": 1.9775784646100768e-05,
+      "loss": 0.4774,
+      "step": 276
+    },
+    {
+      "epoch": 0.3234564297182893,
+      "grad_norm": 0.7053274347741544,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.4867,
+      "step": 277
+    },
+    {
+      "epoch": 0.3246241424609546,
+      "grad_norm": 0.7082193751200652,
+      "learning_rate": 1.9767112008582736e-05,
+      "loss": 0.4421,
+      "step": 278
+    },
+    {
+      "epoch": 0.3257918552036199,
+      "grad_norm": 0.6855509292197659,
+      "learning_rate": 1.9762714710844738e-05,
+      "loss": 0.4706,
+      "step": 279
+    },
+    {
+      "epoch": 0.3269595679462852,
+      "grad_norm": 0.7471378695952058,
+      "learning_rate": 1.97582767847973e-05,
+      "loss": 0.468,
+      "step": 280
+    },
+    {
+      "epoch": 0.32812728068895053,
+      "grad_norm": 0.7634805656045042,
+      "learning_rate": 1.975379824890921e-05,
+      "loss": 0.4561,
+      "step": 281
+    },
+    {
+      "epoch": 0.32929499343161583,
+      "grad_norm": 0.6258738459348818,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.4352,
+      "step": 282
+    },
+    {
+      "epoch": 0.33046270617428114,
+      "grad_norm": 0.7083858123251475,
+      "learning_rate": 1.9744719422331098e-05,
+      "loss": 0.4752,
+      "step": 283
+    },
+    {
+      "epoch": 0.33163041891694645,
+      "grad_norm": 0.7272393763946423,
+      "learning_rate": 1.9740119169423337e-05,
+      "loss": 0.4689,
+      "step": 284
+    },
+    {
+      "epoch": 0.33279813165961175,
+      "grad_norm": 0.7283496757917828,
+      "learning_rate": 1.973547838223927e-05,
+      "loss": 0.4704,
+      "step": 285
+    },
+    {
+      "epoch": 0.33396584440227706,
+      "grad_norm": 0.651124394757982,
+      "learning_rate": 1.9730797080091904e-05,
+      "loss": 0.4264,
+      "step": 286
+    },
+    {
+      "epoch": 0.33513355714494236,
+      "grad_norm": 0.7001818628152243,
+      "learning_rate": 1.9726075282462847e-05,
+      "loss": 0.4729,
+      "step": 287
+    },
+    {
+      "epoch": 0.33630126988760767,
+      "grad_norm": 0.739638547428972,
+      "learning_rate": 1.9721313009002228e-05,
+      "loss": 0.4863,
+      "step": 288
+    },
+    {
+      "epoch": 0.337468982630273,
+      "grad_norm": 0.6698272560708307,
+      "learning_rate": 1.9716510279528632e-05,
+      "loss": 0.4541,
+      "step": 289
+    },
+    {
+      "epoch": 0.3386366953729383,
+      "grad_norm": 0.6578736674255748,
+      "learning_rate": 1.971166711402899e-05,
+      "loss": 0.4695,
+      "step": 290
+    },
+    {
+      "epoch": 0.3398044081156036,
+      "grad_norm": 0.7262860532300098,
+      "learning_rate": 1.9706783532658528e-05,
+      "loss": 0.4357,
+      "step": 291
+    },
+    {
+      "epoch": 0.3409721208582689,
+      "grad_norm": 0.6502724929546747,
+      "learning_rate": 1.9701859555740647e-05,
+      "loss": 0.4476,
+      "step": 292
+    },
+    {
+      "epoch": 0.3421398336009342,
+      "grad_norm": 0.6814606732550476,
+      "learning_rate": 1.969689520376687e-05,
+      "loss": 0.4577,
+      "step": 293
+    },
+    {
+      "epoch": 0.3433075463435995,
+      "grad_norm": 0.6455730618926196,
+      "learning_rate": 1.969189049739674e-05,
+      "loss": 0.465,
+      "step": 294
+    },
+    {
+      "epoch": 0.34447525908626475,
+      "grad_norm": 0.6562879177765564,
+      "learning_rate": 1.968684545745774e-05,
+      "loss": 0.4497,
+      "step": 295
+    },
+    {
+      "epoch": 0.34564297182893006,
+      "grad_norm": 0.6533462617901878,
+      "learning_rate": 1.9681760104945203e-05,
+      "loss": 0.4581,
+      "step": 296
+    },
+    {
+      "epoch": 0.34681068457159536,
+      "grad_norm": 0.6239510584045519,
+      "learning_rate": 1.9676634461022222e-05,
+      "loss": 0.4494,
+      "step": 297
+    },
+    {
+      "epoch": 0.34797839731426067,
+      "grad_norm": 0.6651014991016091,
+      "learning_rate": 1.9671468547019575e-05,
+      "loss": 0.4376,
+      "step": 298
+    },
+    {
+      "epoch": 0.349146110056926,
+      "grad_norm": 0.6959089993530782,
+      "learning_rate": 1.966626238443561e-05,
+      "loss": 0.4532,
+      "step": 299
+    },
+    {
+      "epoch": 0.3503138227995913,
+      "grad_norm": 0.641883707470629,
+      "learning_rate": 1.9661015994936204e-05,
+      "loss": 0.4295,
+      "step": 300
+    },
+    {
+      "epoch": 0.3514815355422566,
+      "grad_norm": 0.716373274030357,
+      "learning_rate": 1.9655729400354607e-05,
+      "loss": 0.4446,
+      "step": 301
+    },
+    {
+      "epoch": 0.3526492482849219,
+      "grad_norm": 0.6376254754294192,
+      "learning_rate": 1.96504026226914e-05,
+      "loss": 0.4668,
+      "step": 302
+    },
+    {
+      "epoch": 0.3538169610275872,
+      "grad_norm": 0.6933079821055188,
+      "learning_rate": 1.9645035684114398e-05,
+      "loss": 0.454,
+      "step": 303
+    },
+    {
+      "epoch": 0.3549846737702525,
+      "grad_norm": 0.7351751738485525,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 0.4845,
+      "step": 304
+    },
+    {
+      "epoch": 0.3561523865129178,
+      "grad_norm": 0.6839105459647862,
+      "learning_rate": 1.963418141372579e-05,
+      "loss": 0.4578,
+      "step": 305
+    },
+    {
+      "epoch": 0.3573200992555831,
+      "grad_norm": 0.670077124805405,
+      "learning_rate": 1.9628694127085094e-05,
+      "loss": 0.4559,
+      "step": 306
+    },
+    {
+      "epoch": 0.3584878119982484,
+      "grad_norm": 0.7700209300473069,
+      "learning_rate": 1.9623166769872216e-05,
+      "loss": 0.4552,
+      "step": 307
+    },
+    {
+      "epoch": 0.3596555247409137,
+      "grad_norm": 0.6684732645227414,
+      "learning_rate": 1.9617599365089693e-05,
+      "loss": 0.4452,
+      "step": 308
+    },
+    {
+      "epoch": 0.36082323748357903,
+      "grad_norm": 0.7364035843514483,
+      "learning_rate": 1.9611991935906727e-05,
+      "loss": 0.4537,
+      "step": 309
+    },
+    {
+      "epoch": 0.36199095022624433,
+      "grad_norm": 0.6688564510066136,
+      "learning_rate": 1.9606344505659072e-05,
+      "loss": 0.4468,
+      "step": 310
+    },
+    {
+      "epoch": 0.36315866296890964,
+      "grad_norm": 0.7135754705907367,
+      "learning_rate": 1.9600657097848957e-05,
+      "loss": 0.4446,
+      "step": 311
+    },
+    {
+      "epoch": 0.36432637571157495,
+      "grad_norm": 0.6801468051828141,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4405,
+      "step": 312
+    },
+    {
+      "epoch": 0.36549408845424025,
+      "grad_norm": 0.6765796518543027,
+      "learning_rate": 1.9589162444382e-05,
+      "loss": 0.4651,
+      "step": 313
+    },
+    {
+      "epoch": 0.36666180119690556,
+      "grad_norm": 0.6671406721706945,
+      "learning_rate": 1.9583355246561076e-05,
+      "loss": 0.4846,
+      "step": 314
+    },
+    {
+      "epoch": 0.36782951393957086,
+      "grad_norm": 0.5927001650084065,
+      "learning_rate": 1.9577508166849308e-05,
+      "loss": 0.4596,
+      "step": 315
+    },
+    {
+      "epoch": 0.36899722668223617,
+      "grad_norm": 0.7599228743151991,
+      "learning_rate": 1.957162122957978e-05,
+      "loss": 0.4544,
+      "step": 316
+    },
+    {
+      "epoch": 0.3701649394249015,
+      "grad_norm": 0.6025081777379674,
+      "learning_rate": 1.9565694459251457e-05,
+      "loss": 0.4644,
+      "step": 317
+    },
+    {
+      "epoch": 0.3713326521675668,
+      "grad_norm": 0.673030812858708,
+      "learning_rate": 1.9559727880529057e-05,
+      "loss": 0.4842,
+      "step": 318
+    },
+    {
+      "epoch": 0.3725003649102321,
+      "grad_norm": 0.7134002336006805,
+      "learning_rate": 1.955372151824297e-05,
+      "loss": 0.4319,
+      "step": 319
+    },
+    {
+      "epoch": 0.3736680776528974,
+      "grad_norm": 0.6593542646301118,
+      "learning_rate": 1.9547675397389144e-05,
+      "loss": 0.461,
+      "step": 320
+    },
+    {
+      "epoch": 0.3748357903955627,
+      "grad_norm": 0.7396053147970731,
+      "learning_rate": 1.954158954312899e-05,
+      "loss": 0.4289,
+      "step": 321
+    },
+    {
+      "epoch": 0.376003503138228,
+      "grad_norm": 0.6274124132087732,
+      "learning_rate": 1.9535463980789277e-05,
+      "loss": 0.458,
+      "step": 322
+    },
+    {
+      "epoch": 0.3771712158808933,
+      "grad_norm": 0.7235863435945558,
+      "learning_rate": 1.9529298735862015e-05,
+      "loss": 0.4082,
+      "step": 323
+    },
+    {
+      "epoch": 0.3783389286235586,
+      "grad_norm": 0.6965260860425936,
+      "learning_rate": 1.9523093834004358e-05,
+      "loss": 0.4568,
+      "step": 324
+    },
+    {
+      "epoch": 0.3795066413662239,
+      "grad_norm": 0.7201003137248683,
+      "learning_rate": 1.951684930103849e-05,
+      "loss": 0.4648,
+      "step": 325
+    },
+    {
+      "epoch": 0.3806743541088892,
+      "grad_norm": 0.7264251245303639,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.4548,
+      "step": 326
+    },
+    {
+      "epoch": 0.38184206685155453,
+      "grad_norm": 0.6443645434115727,
+      "learning_rate": 1.9504241445895434e-05,
+      "loss": 0.4619,
+      "step": 327
+    },
+    {
+      "epoch": 0.38300977959421983,
+      "grad_norm": 0.705346819128517,
+      "learning_rate": 1.949787817618683e-05,
+      "loss": 0.4326,
+      "step": 328
+    },
+    {
+      "epoch": 0.38417749233688514,
+      "grad_norm": 0.6891314871058418,
+      "learning_rate": 1.9491475380306975e-05,
+      "loss": 0.459,
+      "step": 329
+    },
+    {
+      "epoch": 0.38534520507955045,
+      "grad_norm": 0.6608128952857641,
+      "learning_rate": 1.9485033084901606e-05,
+      "loss": 0.4605,
+      "step": 330
+    },
+    {
+      "epoch": 0.38651291782221575,
+      "grad_norm": 0.7665353014556447,
+      "learning_rate": 1.9478551316780855e-05,
+      "loss": 0.4571,
+      "step": 331
+    },
+    {
+      "epoch": 0.38768063056488106,
+      "grad_norm": 0.6849772297099399,
+      "learning_rate": 1.9472030102919102e-05,
+      "loss": 0.4446,
+      "step": 332
+    },
+    {
+      "epoch": 0.38884834330754636,
+      "grad_norm": 0.691674182772583,
+      "learning_rate": 1.94654694704549e-05,
+      "loss": 0.4637,
+      "step": 333
+    },
+    {
+      "epoch": 0.39001605605021167,
+      "grad_norm": 0.7112229129153583,
+      "learning_rate": 1.945886944669084e-05,
+      "loss": 0.4584,
+      "step": 334
+    },
+    {
+      "epoch": 0.391183768792877,
+      "grad_norm": 0.6670118457015922,
+      "learning_rate": 1.9452230059093442e-05,
+      "loss": 0.4769,
+      "step": 335
+    },
+    {
+      "epoch": 0.3923514815355423,
+      "grad_norm": 0.8029443673006625,
+      "learning_rate": 1.944555133529304e-05,
+      "loss": 0.4261,
+      "step": 336
+    },
+    {
+      "epoch": 0.3935191942782076,
+      "grad_norm": 0.6532012294587536,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 0.4509,
+      "step": 337
+    },
+    {
+      "epoch": 0.3946869070208729,
+      "grad_norm": 0.711893571566433,
+      "learning_rate": 1.943207599042297e-05,
+      "loss": 0.4628,
+      "step": 338
+    },
+    {
+      "epoch": 0.3958546197635382,
+      "grad_norm": 0.7557332085841042,
+      "learning_rate": 1.9425279425432016e-05,
+      "loss": 0.4272,
+      "step": 339
+    },
+    {
+      "epoch": 0.3970223325062035,
+      "grad_norm": 0.6288948399349257,
+      "learning_rate": 1.941844363639525e-05,
+      "loss": 0.4441,
+      "step": 340
+    },
+    {
+      "epoch": 0.39819004524886875,
+      "grad_norm": 0.7535663650732167,
+      "learning_rate": 1.941156865176035e-05,
+      "loss": 0.4924,
+      "step": 341
+    },
+    {
+      "epoch": 0.39935775799153406,
+      "grad_norm": 0.6478635213602633,
+      "learning_rate": 1.9404654500138117e-05,
+      "loss": 0.4441,
+      "step": 342
+    },
+    {
+      "epoch": 0.40052547073419936,
+      "grad_norm": 0.6216678850528717,
+      "learning_rate": 1.939770121030233e-05,
+      "loss": 0.4547,
+      "step": 343
+    },
+    {
+      "epoch": 0.40169318347686467,
+      "grad_norm": 0.730100544828049,
+      "learning_rate": 1.939070881118966e-05,
+      "loss": 0.4545,
+      "step": 344
+    },
+    {
+      "epoch": 0.40286089621953,
+      "grad_norm": 0.6373962855683577,
+      "learning_rate": 1.938367733189953e-05,
+      "loss": 0.458,
+      "step": 345
+    },
+    {
+      "epoch": 0.4040286089621953,
+      "grad_norm": 0.7133144024159934,
+      "learning_rate": 1.937660680169399e-05,
+      "loss": 0.4659,
+      "step": 346
+    },
+    {
+      "epoch": 0.4051963217048606,
+      "grad_norm": 0.6872855918108276,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.4689,
+      "step": 347
+    },
+    {
+      "epoch": 0.4063640344475259,
+      "grad_norm": 0.6735126088469916,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.4481,
+      "step": 348
+    },
+    {
+      "epoch": 0.4075317471901912,
+      "grad_norm": 0.6620173671660297,
+      "learning_rate": 1.9355161200642485e-05,
+      "loss": 0.4524,
+      "step": 349
+    },
+    {
+      "epoch": 0.4086994599328565,
+      "grad_norm": 0.7237116997279236,
+      "learning_rate": 1.9347934762644326e-05,
+      "loss": 0.4544,
+      "step": 350
+    },
+    {
+      "epoch": 0.4098671726755218,
+      "grad_norm": 0.6785866980354694,
+      "learning_rate": 1.934066942247629e-05,
+      "loss": 0.4432,
+      "step": 351
+    },
+    {
+      "epoch": 0.4110348854181871,
+      "grad_norm": 0.6464034095951653,
+      "learning_rate": 1.9333365210373668e-05,
+      "loss": 0.4684,
+      "step": 352
+    },
+    {
+      "epoch": 0.4122025981608524,
+      "grad_norm": 0.7347569046710745,
+      "learning_rate": 1.9326022156733514e-05,
+      "loss": 0.442,
+      "step": 353
+    },
+    {
+      "epoch": 0.4133703109035177,
+      "grad_norm": 0.6872990529386536,
+      "learning_rate": 1.9318640292114526e-05,
+      "loss": 0.433,
+      "step": 354
+    },
+    {
+      "epoch": 0.41453802364618303,
+      "grad_norm": 0.669658452483006,
+      "learning_rate": 1.9311219647236914e-05,
+      "loss": 0.4575,
+      "step": 355
+    },
+    {
+      "epoch": 0.41570573638884833,
+      "grad_norm": 0.6892582919449695,
+      "learning_rate": 1.9303760252982287e-05,
+      "loss": 0.4427,
+      "step": 356
+    },
+    {
+      "epoch": 0.41687344913151364,
+      "grad_norm": 0.6742886739987458,
+      "learning_rate": 1.9296262140393498e-05,
+      "loss": 0.4632,
+      "step": 357
+    },
+    {
+      "epoch": 0.41804116187417895,
+      "grad_norm": 0.6268115094153417,
+      "learning_rate": 1.9288725340674536e-05,
+      "loss": 0.4359,
+      "step": 358
+    },
+    {
+      "epoch": 0.41920887461684425,
+      "grad_norm": 0.6344438912418683,
+      "learning_rate": 1.928114988519039e-05,
+      "loss": 0.4652,
+      "step": 359
+    },
+    {
+      "epoch": 0.42037658735950956,
+      "grad_norm": 0.6208055579819224,
+      "learning_rate": 1.927353580546692e-05,
+      "loss": 0.4188,
+      "step": 360
+    },
+    {
+      "epoch": 0.42154430010217486,
+      "grad_norm": 0.6907496290526792,
+      "learning_rate": 1.9265883133190715e-05,
+      "loss": 0.4624,
+      "step": 361
+    },
+    {
+      "epoch": 0.42271201284484017,
+      "grad_norm": 0.6600688039615262,
+      "learning_rate": 1.925819190020898e-05,
+      "loss": 0.4447,
+      "step": 362
+    },
+    {
+      "epoch": 0.4238797255875055,
+      "grad_norm": 0.6625642292473344,
+      "learning_rate": 1.925046213852939e-05,
+      "loss": 0.4813,
+      "step": 363
+    },
+    {
+      "epoch": 0.4250474383301708,
+      "grad_norm": 0.6727312546125456,
+      "learning_rate": 1.9242693880319962e-05,
+      "loss": 0.48,
+      "step": 364
+    },
+    {
+      "epoch": 0.4262151510728361,
+      "grad_norm": 0.6672974213964593,
+      "learning_rate": 1.923488715790891e-05,
+      "loss": 0.4423,
+      "step": 365
+    },
+    {
+      "epoch": 0.4273828638155014,
+      "grad_norm": 0.6939274329395789,
+      "learning_rate": 1.922704200378453e-05,
+      "loss": 0.4473,
+      "step": 366
+    },
+    {
+      "epoch": 0.4285505765581667,
+      "grad_norm": 0.6957144232797671,
+      "learning_rate": 1.921915845059505e-05,
+      "loss": 0.4467,
+      "step": 367
+    },
+    {
+      "epoch": 0.429718289300832,
+      "grad_norm": 0.7107072623431014,
+      "learning_rate": 1.92112365311485e-05,
+      "loss": 0.4601,
+      "step": 368
+    },
+    {
+      "epoch": 0.4308860020434973,
+      "grad_norm": 0.6897560143620715,
+      "learning_rate": 1.920327627841258e-05,
+      "loss": 0.4516,
+      "step": 369
+    },
+    {
+      "epoch": 0.4320537147861626,
+      "grad_norm": 0.6487658920502853,
+      "learning_rate": 1.919527772551451e-05,
+      "loss": 0.4391,
+      "step": 370
+    },
+    {
+      "epoch": 0.4332214275288279,
+      "grad_norm": 0.7365144422522776,
+      "learning_rate": 1.9187240905740893e-05,
+      "loss": 0.4837,
+      "step": 371
+    },
+    {
+      "epoch": 0.4343891402714932,
+      "grad_norm": 0.6695479491700603,
+      "learning_rate": 1.9179165852537596e-05,
+      "loss": 0.4281,
+      "step": 372
+    },
+    {
+      "epoch": 0.43555685301415853,
+      "grad_norm": 0.66512261906684,
+      "learning_rate": 1.917105259950959e-05,
+      "loss": 0.4193,
+      "step": 373
+    },
+    {
+      "epoch": 0.43672456575682383,
+      "grad_norm": 0.6636173415941347,
+      "learning_rate": 1.916290118042082e-05,
+      "loss": 0.4266,
+      "step": 374
+    },
+    {
+      "epoch": 0.43789227849948914,
+      "grad_norm": 0.6337000897613849,
+      "learning_rate": 1.9154711629194062e-05,
+      "loss": 0.4227,
+      "step": 375
+    },
+    {
+      "epoch": 0.43905999124215445,
+      "grad_norm": 0.6235815758527582,
+      "learning_rate": 1.914648397991078e-05,
+      "loss": 0.4493,
+      "step": 376
+    },
+    {
+      "epoch": 0.44022770398481975,
+      "grad_norm": 0.659012390718152,
+      "learning_rate": 1.913821826681099e-05,
+      "loss": 0.4313,
+      "step": 377
+    },
+    {
+      "epoch": 0.44139541672748506,
+      "grad_norm": 0.6958265298516533,
+      "learning_rate": 1.9129914524293102e-05,
+      "loss": 0.4593,
+      "step": 378
+    },
+    {
+      "epoch": 0.44256312947015036,
+      "grad_norm": 0.6630174571365495,
+      "learning_rate": 1.9121572786913808e-05,
+      "loss": 0.4599,
+      "step": 379
+    },
+    {
+      "epoch": 0.44373084221281567,
+      "grad_norm": 0.6162601488226267,
+      "learning_rate": 1.91131930893879e-05,
+      "loss": 0.4629,
+      "step": 380
+    },
+    {
+      "epoch": 0.444898554955481,
+      "grad_norm": 0.6333065496637394,
+      "learning_rate": 1.9104775466588162e-05,
+      "loss": 0.4611,
+      "step": 381
+    },
+    {
+      "epoch": 0.4460662676981463,
+      "grad_norm": 0.6146471410279516,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4412,
+      "step": 382
+    },
+    {
+      "epoch": 0.4472339804408116,
+      "grad_norm": 0.6589653662414898,
+      "learning_rate": 1.9087826585447264e-05,
+      "loss": 0.4569,
+      "step": 383
+    },
+    {
+      "epoch": 0.4484016931834769,
+      "grad_norm": 0.6828477472587944,
+      "learning_rate": 1.9079295397640215e-05,
+      "loss": 0.4758,
+      "step": 384
+    },
+    {
+      "epoch": 0.4495694059261422,
+      "grad_norm": 0.5884428895429085,
+      "learning_rate": 1.9070726425627257e-05,
+      "loss": 0.4414,
+      "step": 385
+    },
+    {
+      "epoch": 0.45073711866880745,
+      "grad_norm": 0.7156583473964074,
+      "learning_rate": 1.9062119705068843e-05,
+      "loss": 0.4647,
+      "step": 386
+    },
+    {
+      "epoch": 0.45190483141147275,
+      "grad_norm": 0.6785741134898847,
+      "learning_rate": 1.9053475271782523e-05,
+      "loss": 0.4605,
+      "step": 387
+    },
+    {
+      "epoch": 0.45307254415413806,
+      "grad_norm": 0.718762580074301,
+      "learning_rate": 1.9044793161742783e-05,
+      "loss": 0.4826,
+      "step": 388
+    },
+    {
+      "epoch": 0.45424025689680336,
+      "grad_norm": 0.6953481697307446,
+      "learning_rate": 1.9036073411080917e-05,
+      "loss": 0.462,
+      "step": 389
+    },
+    {
+      "epoch": 0.45540796963946867,
+      "grad_norm": 0.6206396354799149,
+      "learning_rate": 1.902731605608486e-05,
+      "loss": 0.4307,
+      "step": 390
+    },
+    {
+      "epoch": 0.456575682382134,
+      "grad_norm": 0.6300768540447623,
+      "learning_rate": 1.901852113319903e-05,
+      "loss": 0.437,
+      "step": 391
+    },
+    {
+      "epoch": 0.4577433951247993,
+      "grad_norm": 0.7851633918626141,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.4281,
+      "step": 392
+    },
+    {
+      "epoch": 0.4589111078674646,
+      "grad_norm": 0.6761221346855514,
+      "learning_rate": 1.900081873031731e-05,
+      "loss": 0.4741,
+      "step": 393
+    },
+    {
+      "epoch": 0.4600788206101299,
+      "grad_norm": 0.7176365459129033,
+      "learning_rate": 1.899191132399138e-05,
+      "loss": 0.4957,
+      "step": 394
+    },
+    {
+      "epoch": 0.4612465333527952,
+      "grad_norm": 0.6427082220926086,
+      "learning_rate": 1.898296649711527e-05,
+      "loss": 0.4492,
+      "step": 395
+    },
+    {
+      "epoch": 0.4624142460954605,
+      "grad_norm": 0.6495623703181399,
+      "learning_rate": 1.8973984286913584e-05,
+      "loss": 0.4352,
+      "step": 396
+    },
+    {
+      "epoch": 0.4635819588381258,
+      "grad_norm": 0.6972525336380327,
+      "learning_rate": 1.896496473076651e-05,
+      "loss": 0.4571,
+      "step": 397
+    },
+    {
+      "epoch": 0.4647496715807911,
+      "grad_norm": 0.7014685487488379,
+      "learning_rate": 1.895590786620963e-05,
+      "loss": 0.4666,
+      "step": 398
+    },
+    {
+      "epoch": 0.4659173843234564,
+      "grad_norm": 0.6612128620331346,
+      "learning_rate": 1.8946813730933816e-05,
+      "loss": 0.4971,
+      "step": 399
+    },
+    {
+      "epoch": 0.4670850970661217,
+      "grad_norm": 0.7139324517541192,
+      "learning_rate": 1.8937682362785025e-05,
+      "loss": 0.44,
+      "step": 400
+    },
+    {
+      "epoch": 0.46825280980878703,
+      "grad_norm": 0.6280454910262269,
+      "learning_rate": 1.8928513799764164e-05,
+      "loss": 0.4352,
+      "step": 401
+    },
+    {
+      "epoch": 0.46942052255145233,
+      "grad_norm": 0.7134090744651425,
+      "learning_rate": 1.891930808002694e-05,
+      "loss": 0.4412,
+      "step": 402
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.6111428069753303,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.4286,
+      "step": 403
+    },
+    {
+      "epoch": 0.47175594803678295,
+      "grad_norm": 0.6942896129053439,
+      "learning_rate": 1.890078532379919e-05,
+      "loss": 0.4313,
+      "step": 404
+    },
+    {
+      "epoch": 0.47292366077944825,
+      "grad_norm": 0.6624321678873382,
+      "learning_rate": 1.8891468364392577e-05,
+      "loss": 0.4439,
+      "step": 405
+    },
+    {
+      "epoch": 0.47409137352211356,
+      "grad_norm": 0.6263405217337588,
+      "learning_rate": 1.8882114402437105e-05,
+      "loss": 0.4296,
+      "step": 406
+    },
+    {
+      "epoch": 0.47525908626477886,
+      "grad_norm": 0.6480239240799481,
+      "learning_rate": 1.8872723476860033e-05,
+      "loss": 0.4257,
+      "step": 407
+    },
+    {
+      "epoch": 0.47642679900744417,
+      "grad_norm": 0.5848844018067597,
+      "learning_rate": 1.8863295626742438e-05,
+      "loss": 0.4355,
+      "step": 408
+    },
+    {
+      "epoch": 0.4775945117501095,
+      "grad_norm": 0.6494481345063974,
+      "learning_rate": 1.885383089131906e-05,
+      "loss": 0.4668,
+      "step": 409
+    },
+    {
+      "epoch": 0.4787622244927748,
+      "grad_norm": 0.6557814637114865,
+      "learning_rate": 1.8844329309978146e-05,
+      "loss": 0.4606,
+      "step": 410
+    },
+    {
+      "epoch": 0.4799299372354401,
+      "grad_norm": 0.6543315349611579,
+      "learning_rate": 1.8834790922261278e-05,
+      "loss": 0.4438,
+      "step": 411
+    },
+    {
+      "epoch": 0.4810976499781054,
+      "grad_norm": 0.6598269726460232,
+      "learning_rate": 1.8825215767863215e-05,
+      "loss": 0.4393,
+      "step": 412
+    },
+    {
+      "epoch": 0.4822653627207707,
+      "grad_norm": 0.674243608307105,
+      "learning_rate": 1.8815603886631715e-05,
+      "loss": 0.4272,
+      "step": 413
+    },
+    {
+      "epoch": 0.483433075463436,
+      "grad_norm": 0.6211925813343548,
+      "learning_rate": 1.880595531856738e-05,
+      "loss": 0.459,
+      "step": 414
+    },
+    {
+      "epoch": 0.4846007882061013,
+      "grad_norm": 0.7081294104572783,
+      "learning_rate": 1.8796270103823494e-05,
+      "loss": 0.4716,
+      "step": 415
+    },
+    {
+      "epoch": 0.4857685009487666,
+      "grad_norm": 0.6092525463903073,
+      "learning_rate": 1.8786548282705847e-05,
+      "loss": 0.4801,
+      "step": 416
+    },
+    {
+      "epoch": 0.4869362136914319,
+      "grad_norm": 0.6337419151303799,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.4429,
+      "step": 417
+    },
+    {
+      "epoch": 0.4881039264340972,
+      "grad_norm": 0.6773908181498255,
+      "learning_rate": 1.876699498333393e-05,
+      "loss": 0.4479,
+      "step": 418
+    },
+    {
+      "epoch": 0.48927163917676253,
+      "grad_norm": 0.5967103095868506,
+      "learning_rate": 1.8757163586452266e-05,
+      "loss": 0.4392,
+      "step": 419
+    },
+    {
+      "epoch": 0.49043935191942784,
+      "grad_norm": 0.651022112752135,
+      "learning_rate": 1.8747295745941705e-05,
+      "loss": 0.4685,
+      "step": 420
+    },
+    {
+      "epoch": 0.49160706466209314,
+      "grad_norm": 0.6526288684651269,
+      "learning_rate": 1.8737391502868038e-05,
+      "loss": 0.4596,
+      "step": 421
+    },
+    {
+      "epoch": 0.49277477740475845,
+      "grad_norm": 0.6111990487526071,
+      "learning_rate": 1.8727450898448562e-05,
+      "loss": 0.4288,
+      "step": 422
+    },
+    {
+      "epoch": 0.49394249014742375,
+      "grad_norm": 0.7357988108483421,
+      "learning_rate": 1.8717473974051888e-05,
+      "loss": 0.4175,
+      "step": 423
+    },
+    {
+      "epoch": 0.49511020289008906,
+      "grad_norm": 0.6281695293775557,
+      "learning_rate": 1.8707460771197773e-05,
+      "loss": 0.4419,
+      "step": 424
+    },
+    {
+      "epoch": 0.49627791563275436,
+      "grad_norm": 0.628416030439757,
+      "learning_rate": 1.8697411331556958e-05,
+      "loss": 0.4412,
+      "step": 425
+    },
+    {
+      "epoch": 0.49744562837541967,
+      "grad_norm": 0.6980798660235025,
+      "learning_rate": 1.8687325696950972e-05,
+      "loss": 0.4772,
+      "step": 426
+    },
+    {
+      "epoch": 0.498613341118085,
+      "grad_norm": 0.6492166654096987,
+      "learning_rate": 1.867720390935199e-05,
+      "loss": 0.4516,
+      "step": 427
+    },
+    {
+      "epoch": 0.4997810538607503,
+      "grad_norm": 0.7037397095023976,
+      "learning_rate": 1.8667046010882627e-05,
+      "loss": 0.4489,
+      "step": 428
+    },
+    {
+      "epoch": 0.5009487666034156,
+      "grad_norm": 0.6617284557887854,
+      "learning_rate": 1.865685204381578e-05,
+      "loss": 0.4447,
+      "step": 429
+    },
+    {
+      "epoch": 0.5021164793460808,
+      "grad_norm": 0.6267738258866716,
+      "learning_rate": 1.8646622050574456e-05,
+      "loss": 0.4352,
+      "step": 430
+    },
+    {
+      "epoch": 0.5032841920887462,
+      "grad_norm": 0.6805664497811045,
+      "learning_rate": 1.863635607373157e-05,
+      "loss": 0.4448,
+      "step": 431
+    },
+    {
+      "epoch": 0.5044519048314114,
+      "grad_norm": 0.7311326059529988,
+      "learning_rate": 1.8626054156009807e-05,
+      "loss": 0.4208,
+      "step": 432
+    },
+    {
+      "epoch": 0.5056196175740768,
+      "grad_norm": 0.7198309446394008,
+      "learning_rate": 1.8615716340281408e-05,
+      "loss": 0.4814,
+      "step": 433
+    },
+    {
+      "epoch": 0.5067873303167421,
+      "grad_norm": 0.7202641023712736,
+      "learning_rate": 1.8605342669568013e-05,
+      "loss": 0.4462,
+      "step": 434
+    },
+    {
+      "epoch": 0.5079550430594074,
+      "grad_norm": 0.6366678747293129,
+      "learning_rate": 1.8594933187040468e-05,
+      "loss": 0.4329,
+      "step": 435
+    },
+    {
+      "epoch": 0.5091227558020727,
+      "grad_norm": 0.7150085103686383,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.4237,
+      "step": 436
+    },
+    {
+      "epoch": 0.510290468544738,
+      "grad_norm": 0.686923292506756,
+      "learning_rate": 1.8574006959971335e-05,
+      "loss": 0.4649,
+      "step": 437
+    },
+    {
+      "epoch": 0.5114581812874033,
+      "grad_norm": 0.6791379808676635,
+      "learning_rate": 1.856349030251589e-05,
+      "loss": 0.4405,
+      "step": 438
+    },
+    {
+      "epoch": 0.5126258940300686,
+      "grad_norm": 0.7671276541653522,
+      "learning_rate": 1.8552938007418237e-05,
+      "loss": 0.4416,
+      "step": 439
+    },
+    {
+      "epoch": 0.5137936067727339,
+      "grad_norm": 0.6756080371949191,
+      "learning_rate": 1.8542350118592585e-05,
+      "loss": 0.4317,
+      "step": 440
+    },
+    {
+      "epoch": 0.5149613195153993,
+      "grad_norm": 0.6638277622609118,
+      "learning_rate": 1.8531726680101268e-05,
+      "loss": 0.4424,
+      "step": 441
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.7684793182389605,
+      "learning_rate": 1.8521067736154567e-05,
+      "loss": 0.4811,
+      "step": 442
+    },
+    {
+      "epoch": 0.5172967450007299,
+      "grad_norm": 0.6430459776616491,
+      "learning_rate": 1.851037333111052e-05,
+      "loss": 0.4594,
+      "step": 443
+    },
+    {
+      "epoch": 0.5184644577433951,
+      "grad_norm": 0.7097924894219443,
+      "learning_rate": 1.8499643509474738e-05,
+      "loss": 0.4288,
+      "step": 444
+    },
+    {
+      "epoch": 0.5196321704860605,
+      "grad_norm": 0.6256230397983086,
+      "learning_rate": 1.8488878315900228e-05,
+      "loss": 0.4152,
+      "step": 445
+    },
+    {
+      "epoch": 0.5207998832287257,
+      "grad_norm": 0.6154742493269747,
+      "learning_rate": 1.8478077795187188e-05,
+      "loss": 0.4797,
+      "step": 446
+    },
+    {
+      "epoch": 0.5219675959713911,
+      "grad_norm": 0.7578241327245697,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.4833,
+      "step": 447
+    },
+    {
+      "epoch": 0.5231353087140563,
+      "grad_norm": 0.6920351563097324,
+      "learning_rate": 1.8456370952281243e-05,
+      "loss": 0.4524,
+      "step": 448
+    },
+    {
+      "epoch": 0.5243030214567217,
+      "grad_norm": 0.6608437017546993,
+      "learning_rate": 1.844546472042308e-05,
+      "loss": 0.453,
+      "step": 449
+    },
+    {
+      "epoch": 0.525470734199387,
+      "grad_norm": 0.715515514516088,
+      "learning_rate": 1.84345233420955e-05,
+      "loss": 0.4785,
+      "step": 450
+    },
+    {
+      "epoch": 0.5266384469420523,
+      "grad_norm": 0.6466681653389168,
+      "learning_rate": 1.8423546862831917e-05,
+      "loss": 0.4625,
+      "step": 451
+    },
+    {
+      "epoch": 0.5278061596847176,
+      "grad_norm": 0.6184257021245961,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.434,
+      "step": 452
+    },
+    {
+      "epoch": 0.5289738724273829,
+      "grad_norm": 0.6925592810588838,
+      "learning_rate": 1.8401488784360567e-05,
+      "loss": 0.4288,
+      "step": 453
+    },
+    {
+      "epoch": 0.5301415851700482,
+      "grad_norm": 0.5736548242616697,
+      "learning_rate": 1.8390407276949235e-05,
+      "loss": 0.4401,
+      "step": 454
+    },
+    {
+      "epoch": 0.5313092979127134,
+      "grad_norm": 0.6705204756606166,
+      "learning_rate": 1.8379290852194395e-05,
+      "loss": 0.4583,
+      "step": 455
+    },
+    {
+      "epoch": 0.5324770106553788,
+      "grad_norm": 0.7548052243243242,
+      "learning_rate": 1.836813955635793e-05,
+      "loss": 0.4461,
+      "step": 456
+    },
+    {
+      "epoch": 0.533644723398044,
+      "grad_norm": 0.7014862489781978,
+      "learning_rate": 1.835695343584683e-05,
+      "loss": 0.4317,
+      "step": 457
+    },
+    {
+      "epoch": 0.5348124361407094,
+      "grad_norm": 0.6686481821066799,
+      "learning_rate": 1.834573253721303e-05,
+      "loss": 0.4541,
+      "step": 458
+    },
+    {
+      "epoch": 0.5359801488833746,
+      "grad_norm": 0.6006437112572678,
+      "learning_rate": 1.8334476907153177e-05,
+      "loss": 0.4197,
+      "step": 459
+    },
+    {
+      "epoch": 0.53714786162604,
+      "grad_norm": 0.652069162162329,
+      "learning_rate": 1.8323186592508474e-05,
+      "loss": 0.4525,
+      "step": 460
+    },
+    {
+      "epoch": 0.5383155743687053,
+      "grad_norm": 0.6150191889626871,
+      "learning_rate": 1.8311861640264456e-05,
+      "loss": 0.4683,
+      "step": 461
+    },
+    {
+      "epoch": 0.5394832871113706,
+      "grad_norm": 0.6606696130972726,
+      "learning_rate": 1.8300502097550807e-05,
+      "loss": 0.4258,
+      "step": 462
+    },
+    {
+      "epoch": 0.5406509998540359,
+      "grad_norm": 0.6180307112422921,
+      "learning_rate": 1.8289108011641163e-05,
+      "loss": 0.4583,
+      "step": 463
+    },
+    {
+      "epoch": 0.5418187125967012,
+      "grad_norm": 0.6341599868487502,
+      "learning_rate": 1.8277679429952913e-05,
+      "loss": 0.4049,
+      "step": 464
+    },
+    {
+      "epoch": 0.5429864253393665,
+      "grad_norm": 0.6299134048984188,
+      "learning_rate": 1.8266216400047002e-05,
+      "loss": 0.4307,
+      "step": 465
+    },
+    {
+      "epoch": 0.5441541380820318,
+      "grad_norm": 0.6648838955461025,
+      "learning_rate": 1.825471896962774e-05,
+      "loss": 0.4009,
+      "step": 466
+    },
+    {
+      "epoch": 0.5453218508246971,
+      "grad_norm": 0.6343365121785837,
+      "learning_rate": 1.8243187186542594e-05,
+      "loss": 0.4245,
+      "step": 467
+    },
+    {
+      "epoch": 0.5464895635673624,
+      "grad_norm": 0.6601066045922249,
+      "learning_rate": 1.8231621098781983e-05,
+      "loss": 0.4566,
+      "step": 468
+    },
+    {
+      "epoch": 0.5476572763100277,
+      "grad_norm": 0.6390680610252225,
+      "learning_rate": 1.8220020754479104e-05,
+      "loss": 0.447,
+      "step": 469
+    },
+    {
+      "epoch": 0.5488249890526931,
+      "grad_norm": 0.6537008938312688,
+      "learning_rate": 1.8208386201909702e-05,
+      "loss": 0.4552,
+      "step": 470
+    },
+    {
+      "epoch": 0.5499927017953583,
+      "grad_norm": 0.6478841343689015,
+      "learning_rate": 1.8196717489491885e-05,
+      "loss": 0.4395,
+      "step": 471
+    },
+    {
+      "epoch": 0.5511604145380237,
+      "grad_norm": 0.6477767802073617,
+      "learning_rate": 1.8185014665785936e-05,
+      "loss": 0.4378,
+      "step": 472
+    },
+    {
+      "epoch": 0.5523281272806889,
+      "grad_norm": 0.71259899520148,
+      "learning_rate": 1.817327777949407e-05,
+      "loss": 0.4335,
+      "step": 473
+    },
+    {
+      "epoch": 0.5534958400233543,
+      "grad_norm": 0.6742514322574882,
+      "learning_rate": 1.8161506879460274e-05,
+      "loss": 0.4575,
+      "step": 474
+    },
+    {
+      "epoch": 0.5546635527660195,
+      "grad_norm": 0.7064131965894895,
+      "learning_rate": 1.8149702014670074e-05,
+      "loss": 0.4165,
+      "step": 475
+    },
+    {
+      "epoch": 0.5558312655086849,
+      "grad_norm": 0.7608374131897381,
+      "learning_rate": 1.8137863234250346e-05,
+      "loss": 0.5094,
+      "step": 476
+    },
+    {
+      "epoch": 0.5569989782513501,
+      "grad_norm": 0.6032293766999365,
+      "learning_rate": 1.8125990587469125e-05,
+      "loss": 0.4166,
+      "step": 477
+    },
+    {
+      "epoch": 0.5581666909940155,
+      "grad_norm": 0.6960636073518784,
+      "learning_rate": 1.8114084123735355e-05,
+      "loss": 0.462,
+      "step": 478
+    },
+    {
+      "epoch": 0.5593344037366808,
+      "grad_norm": 0.7399696058765737,
+      "learning_rate": 1.8102143892598734e-05,
+      "loss": 0.4404,
+      "step": 479
+    },
+    {
+      "epoch": 0.5605021164793461,
+      "grad_norm": 0.6714469817830911,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.4454,
+      "step": 480
+    },
+    {
+      "epoch": 0.5616698292220114,
+      "grad_norm": 0.7385049192619568,
+      "learning_rate": 1.8078162327018115e-05,
+      "loss": 0.4583,
+      "step": 481
+    },
+    {
+      "epoch": 0.5628375419646767,
+      "grad_norm": 0.6048728417885677,
+      "learning_rate": 1.8066121092375303e-05,
+      "loss": 0.4206,
+      "step": 482
+    },
+    {
+      "epoch": 0.564005254707342,
+      "grad_norm": 0.7233012295069223,
+      "learning_rate": 1.8054046289931585e-05,
+      "loss": 0.465,
+      "step": 483
+    },
+    {
+      "epoch": 0.5651729674500073,
+      "grad_norm": 0.6660650409719168,
+      "learning_rate": 1.8041937969937206e-05,
+      "loss": 0.4252,
+      "step": 484
+    },
+    {
+      "epoch": 0.5663406801926726,
+      "grad_norm": 0.6360435940734128,
+      "learning_rate": 1.80297961827819e-05,
+      "loss": 0.43,
+      "step": 485
+    },
+    {
+      "epoch": 0.567508392935338,
+      "grad_norm": 0.6286789422925646,
+      "learning_rate": 1.8017620978994676e-05,
+      "loss": 0.4246,
+      "step": 486
+    },
+    {
+      "epoch": 0.5686761056780032,
+      "grad_norm": 0.6471963896147364,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.4434,
+      "step": 487
+    },
+    {
+      "epoch": 0.5698438184206686,
+      "grad_norm": 0.5612509025524243,
+      "learning_rate": 1.7993170524335614e-05,
+      "loss": 0.4103,
+      "step": 488
+    },
+    {
+      "epoch": 0.5710115311633338,
+      "grad_norm": 0.6407570904743627,
+      "learning_rate": 1.7980895375216282e-05,
+      "loss": 0.4255,
+      "step": 489
+    },
+    {
+      "epoch": 0.5721792439059992,
+      "grad_norm": 0.6292157865273278,
+      "learning_rate": 1.7968587012969606e-05,
+      "loss": 0.4406,
+      "step": 490
+    },
+    {
+      "epoch": 0.5733469566486644,
+      "grad_norm": 0.6238896161852796,
+      "learning_rate": 1.795624548881781e-05,
+      "loss": 0.4274,
+      "step": 491
+    },
+    {
+      "epoch": 0.5745146693913298,
+      "grad_norm": 0.6914193671787252,
+      "learning_rate": 1.7943870854121126e-05,
+      "loss": 0.4722,
+      "step": 492
+    },
+    {
+      "epoch": 0.575682382133995,
+      "grad_norm": 0.5662043717086018,
+      "learning_rate": 1.7931463160377575e-05,
+      "loss": 0.4349,
+      "step": 493
+    },
+    {
+      "epoch": 0.5768500948766604,
+      "grad_norm": 0.6152619085192582,
+      "learning_rate": 1.7919022459222754e-05,
+      "loss": 0.4635,
+      "step": 494
+    },
+    {
+      "epoch": 0.5780178076193256,
+      "grad_norm": 0.8120772339815311,
+      "learning_rate": 1.7906548802429622e-05,
+      "loss": 0.4302,
+      "step": 495
+    },
+    {
+      "epoch": 0.579185520361991,
+      "grad_norm": 0.5768762898277381,
+      "learning_rate": 1.7894042241908293e-05,
+      "loss": 0.4422,
+      "step": 496
+    },
+    {
+      "epoch": 0.5803532331046563,
+      "grad_norm": 0.6994526342336237,
+      "learning_rate": 1.7881502829705808e-05,
+      "loss": 0.4295,
+      "step": 497
+    },
+    {
+      "epoch": 0.5815209458473216,
+      "grad_norm": 0.8169485603921001,
+      "learning_rate": 1.786893061800592e-05,
+      "loss": 0.4266,
+      "step": 498
+    },
+    {
+      "epoch": 0.5826886585899869,
+      "grad_norm": 0.6899972634599797,
+      "learning_rate": 1.785632565912888e-05,
+      "loss": 0.4513,
+      "step": 499
+    },
+    {
+      "epoch": 0.5838563713326521,
+      "grad_norm": 1.2993815254446277,
+      "learning_rate": 1.7843688005531227e-05,
+      "loss": 0.4502,
+      "step": 500
+    },
+    {
+      "epoch": 0.5850240840753175,
+      "grad_norm": 0.6247074197094112,
+      "learning_rate": 1.7831017709805555e-05,
+      "loss": 0.4714,
+      "step": 501
+    },
+    {
+      "epoch": 0.5861917968179827,
+      "grad_norm": 0.8546388118065866,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.4528,
+      "step": 502
+    },
+    {
+      "epoch": 0.5873595095606481,
+      "grad_norm": 0.6840797747988135,
+      "learning_rate": 1.7805579403019523e-05,
+      "loss": 0.4414,
+      "step": 503
+    },
+    {
+      "epoch": 0.5885272223033133,
+      "grad_norm": 0.600248139794138,
+      "learning_rate": 1.779281149782269e-05,
+      "loss": 0.4218,
+      "step": 504
+    },
+    {
+      "epoch": 0.5896949350459787,
+      "grad_norm": 0.628410015467311,
+      "learning_rate": 1.7780011162224444e-05,
+      "loss": 0.4392,
+      "step": 505
+    },
+    {
+      "epoch": 0.590862647788644,
+      "grad_norm": 0.6134045762563667,
+      "learning_rate": 1.7767178449494403e-05,
+      "loss": 0.4535,
+      "step": 506
+    },
+    {
+      "epoch": 0.5920303605313093,
+      "grad_norm": 0.6940516916993587,
+      "learning_rate": 1.7754313413036907e-05,
+      "loss": 0.4297,
+      "step": 507
+    },
+    {
+      "epoch": 0.5931980732739746,
+      "grad_norm": 0.6945035842661087,
+      "learning_rate": 1.7741416106390828e-05,
+      "loss": 0.4671,
+      "step": 508
+    },
+    {
+      "epoch": 0.5943657860166399,
+      "grad_norm": 0.6241523080344628,
+      "learning_rate": 1.7728486583229323e-05,
+      "loss": 0.4322,
+      "step": 509
+    },
+    {
+      "epoch": 0.5955334987593052,
+      "grad_norm": 0.6425851504105187,
+      "learning_rate": 1.771552489735963e-05,
+      "loss": 0.4341,
+      "step": 510
+    },
+    {
+      "epoch": 0.5967012115019705,
+      "grad_norm": 0.7077800383966888,
+      "learning_rate": 1.770253110272283e-05,
+      "loss": 0.4513,
+      "step": 511
+    },
+    {
+      "epoch": 0.5978689242446358,
+      "grad_norm": 0.6483133313241387,
+      "learning_rate": 1.768950525339362e-05,
+      "loss": 0.4357,
+      "step": 512
+    },
+    {
+      "epoch": 0.5990366369873011,
+      "grad_norm": 0.7098496391049285,
+      "learning_rate": 1.7676447403580114e-05,
+      "loss": 0.4191,
+      "step": 513
+    },
+    {
+      "epoch": 0.6002043497299664,
+      "grad_norm": 0.5978977157191546,
+      "learning_rate": 1.7663357607623577e-05,
+      "loss": 0.4307,
+      "step": 514
+    },
+    {
+      "epoch": 0.6013720624726318,
+      "grad_norm": 0.661386576272504,
+      "learning_rate": 1.7650235919998234e-05,
+      "loss": 0.4526,
+      "step": 515
+    },
+    {
+      "epoch": 0.602539775215297,
+      "grad_norm": 0.6095715598796404,
+      "learning_rate": 1.7637082395311024e-05,
+      "loss": 0.4099,
+      "step": 516
+    },
+    {
+      "epoch": 0.6037074879579624,
+      "grad_norm": 0.581080150308715,
+      "learning_rate": 1.7623897088301387e-05,
+      "loss": 0.4193,
+      "step": 517
+    },
+    {
+      "epoch": 0.6048752007006276,
+      "grad_norm": 0.6194829372088866,
+      "learning_rate": 1.7610680053841008e-05,
+      "loss": 0.4315,
+      "step": 518
+    },
+    {
+      "epoch": 0.606042913443293,
+      "grad_norm": 0.66621716697805,
+      "learning_rate": 1.7597431346933633e-05,
+      "loss": 0.4562,
+      "step": 519
+    },
+    {
+      "epoch": 0.6072106261859582,
+      "grad_norm": 0.6222856229769534,
+      "learning_rate": 1.75841510227148e-05,
+      "loss": 0.4365,
+      "step": 520
+    },
+    {
+      "epoch": 0.6083783389286236,
+      "grad_norm": 0.6244741233028591,
+      "learning_rate": 1.757083913645163e-05,
+      "loss": 0.4482,
+      "step": 521
+    },
+    {
+      "epoch": 0.6095460516712888,
+      "grad_norm": 0.5718654434292418,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4155,
+      "step": 522
+    },
+    {
+      "epoch": 0.6107137644139542,
+      "grad_norm": 0.634847643658831,
+      "learning_rate": 1.7544120899517255e-05,
+      "loss": 0.4376,
+      "step": 523
+    },
+    {
+      "epoch": 0.6118814771566194,
+      "grad_norm": 0.6436735753299665,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.4803,
+      "step": 524
+    },
+    {
+      "epoch": 0.6130491898992848,
+      "grad_norm": 0.6361108029042629,
+      "learning_rate": 1.7517277080890277e-05,
+      "loss": 0.4842,
+      "step": 525
+    },
+    {
+      "epoch": 0.6142169026419501,
+      "grad_norm": 0.645330967727173,
+      "learning_rate": 1.7503808218001305e-05,
+      "loss": 0.4365,
+      "step": 526
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.6181717358909106,
+      "learning_rate": 1.7490308127420928e-05,
+      "loss": 0.4649,
+      "step": 527
+    },
+    {
+      "epoch": 0.6165523281272807,
+      "grad_norm": 0.6173586187082203,
+      "learning_rate": 1.7476776865330847e-05,
+      "loss": 0.4153,
+      "step": 528
+    },
+    {
+      "epoch": 0.617720040869946,
+      "grad_norm": 0.5863344279724265,
+      "learning_rate": 1.7463214488042472e-05,
+      "loss": 0.4467,
+      "step": 529
+    },
+    {
+      "epoch": 0.6188877536126113,
+      "grad_norm": 0.6307912634600832,
+      "learning_rate": 1.7449621051996714e-05,
+      "loss": 0.4477,
+      "step": 530
+    },
+    {
+      "epoch": 0.6200554663552766,
+      "grad_norm": 0.638438264477883,
+      "learning_rate": 1.743599661376373e-05,
+      "loss": 0.4841,
+      "step": 531
+    },
+    {
+      "epoch": 0.6212231790979419,
+      "grad_norm": 0.6177280316331522,
+      "learning_rate": 1.74223412300427e-05,
+      "loss": 0.4345,
+      "step": 532
+    },
+    {
+      "epoch": 0.6223908918406073,
+      "grad_norm": 0.6533005781105918,
+      "learning_rate": 1.7408654957661585e-05,
+      "loss": 0.4551,
+      "step": 533
+    },
+    {
+      "epoch": 0.6235586045832725,
+      "grad_norm": 0.618803165322957,
+      "learning_rate": 1.739493785357688e-05,
+      "loss": 0.4306,
+      "step": 534
+    },
+    {
+      "epoch": 0.6247263173259379,
+      "grad_norm": 0.6101715297593397,
+      "learning_rate": 1.738118997487341e-05,
+      "loss": 0.4247,
+      "step": 535
+    },
+    {
+      "epoch": 0.6258940300686031,
+      "grad_norm": 0.6104812355729845,
+      "learning_rate": 1.736741137876405e-05,
+      "loss": 0.4221,
+      "step": 536
+    },
+    {
+      "epoch": 0.6270617428112685,
+      "grad_norm": 0.6152060210595834,
+      "learning_rate": 1.7353602122589528e-05,
+      "loss": 0.4611,
+      "step": 537
+    },
+    {
+      "epoch": 0.6282294555539337,
+      "grad_norm": 0.5815768786131711,
+      "learning_rate": 1.7339762263818144e-05,
+      "loss": 0.4438,
+      "step": 538
+    },
+    {
+      "epoch": 0.6293971682965991,
+      "grad_norm": 0.6147519474922392,
+      "learning_rate": 1.7325891860045577e-05,
+      "loss": 0.4249,
+      "step": 539
+    },
+    {
+      "epoch": 0.6305648810392643,
+      "grad_norm": 0.6222792703357362,
+      "learning_rate": 1.7311990968994598e-05,
+      "loss": 0.4389,
+      "step": 540
+    },
+    {
+      "epoch": 0.6317325937819297,
+      "grad_norm": 0.6128820500402491,
+      "learning_rate": 1.7298059648514876e-05,
+      "loss": 0.4469,
+      "step": 541
+    },
+    {
+      "epoch": 0.632900306524595,
+      "grad_norm": 0.6212383224405283,
+      "learning_rate": 1.7284097956582694e-05,
+      "loss": 0.4077,
+      "step": 542
+    },
+    {
+      "epoch": 0.6340680192672603,
+      "grad_norm": 0.5823272323627529,
+      "learning_rate": 1.727010595130074e-05,
+      "loss": 0.4307,
+      "step": 543
+    },
+    {
+      "epoch": 0.6352357320099256,
+      "grad_norm": 0.6268292318102405,
+      "learning_rate": 1.725608369089785e-05,
+      "loss": 0.4271,
+      "step": 544
+    },
+    {
+      "epoch": 0.6364034447525908,
+      "grad_norm": 0.6132170612932472,
+      "learning_rate": 1.724203123372877e-05,
+      "loss": 0.4282,
+      "step": 545
+    },
+    {
+      "epoch": 0.6375711574952562,
+      "grad_norm": 0.6096705646413837,
+      "learning_rate": 1.7227948638273918e-05,
+      "loss": 0.433,
+      "step": 546
+    },
+    {
+      "epoch": 0.6387388702379214,
+      "grad_norm": 0.5618244231419152,
+      "learning_rate": 1.721383596313912e-05,
+      "loss": 0.4187,
+      "step": 547
+    },
+    {
+      "epoch": 0.6399065829805868,
+      "grad_norm": 0.6271141257634439,
+      "learning_rate": 1.7199693267055392e-05,
+      "loss": 0.4194,
+      "step": 548
+    },
+    {
+      "epoch": 0.641074295723252,
+      "grad_norm": 0.7037277419337425,
+      "learning_rate": 1.7185520608878685e-05,
+      "loss": 0.4537,
+      "step": 549
+    },
+    {
+      "epoch": 0.6422420084659174,
+      "grad_norm": 0.5630967643556134,
+      "learning_rate": 1.7171318047589637e-05,
+      "loss": 0.4541,
+      "step": 550
+    },
+    {
+      "epoch": 0.6434097212085826,
+      "grad_norm": 0.6775977439100137,
+      "learning_rate": 1.7157085642293328e-05,
+      "loss": 0.4123,
+      "step": 551
+    },
+    {
+      "epoch": 0.644577433951248,
+      "grad_norm": 0.6594558379585748,
+      "learning_rate": 1.7142823452219036e-05,
+      "loss": 0.4294,
+      "step": 552
+    },
+    {
+      "epoch": 0.6457451466939133,
+      "grad_norm": 0.6327354034922164,
+      "learning_rate": 1.712853153672e-05,
+      "loss": 0.4381,
+      "step": 553
+    },
+    {
+      "epoch": 0.6469128594365786,
+      "grad_norm": 0.6778633926066358,
+      "learning_rate": 1.7114209955273154e-05,
+      "loss": 0.475,
+      "step": 554
+    },
+    {
+      "epoch": 0.6480805721792439,
+      "grad_norm": 0.6839843686646412,
+      "learning_rate": 1.7099858767478888e-05,
+      "loss": 0.4393,
+      "step": 555
+    },
+    {
+      "epoch": 0.6492482849219092,
+      "grad_norm": 0.676960522487825,
+      "learning_rate": 1.7085478033060808e-05,
+      "loss": 0.4455,
+      "step": 556
+    },
+    {
+      "epoch": 0.6504159976645745,
+      "grad_norm": 0.6887731429957744,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.3983,
+      "step": 557
+    },
+    {
+      "epoch": 0.6515837104072398,
+      "grad_norm": 0.6136689605575111,
+      "learning_rate": 1.705662816386217e-05,
+      "loss": 0.4682,
+      "step": 558
+    },
+    {
+      "epoch": 0.6527514231499051,
+      "grad_norm": 0.6479836859466991,
+      "learning_rate": 1.7042159149142626e-05,
+      "loss": 0.4247,
+      "step": 559
+    },
+    {
+      "epoch": 0.6539191358925704,
+      "grad_norm": 0.7162482427775203,
+      "learning_rate": 1.7027660827920798e-05,
+      "loss": 0.4508,
+      "step": 560
+    },
+    {
+      "epoch": 0.6550868486352357,
+      "grad_norm": 0.5650626422173433,
+      "learning_rate": 1.7013133260532584e-05,
+      "loss": 0.3895,
+      "step": 561
+    },
+    {
+      "epoch": 0.6562545613779011,
+      "grad_norm": 0.6802495623470067,
+      "learning_rate": 1.699857650743562e-05,
+      "loss": 0.4206,
+      "step": 562
+    },
+    {
+      "epoch": 0.6574222741205663,
+      "grad_norm": 0.6147962394206135,
+      "learning_rate": 1.6983990629208973e-05,
+      "loss": 0.404,
+      "step": 563
+    },
+    {
+      "epoch": 0.6585899868632317,
+      "grad_norm": 0.6652214498786506,
+      "learning_rate": 1.696937568655294e-05,
+      "loss": 0.415,
+      "step": 564
+    },
+    {
+      "epoch": 0.6597576996058969,
+      "grad_norm": 0.6714345650602606,
+      "learning_rate": 1.6954731740288744e-05,
+      "loss": 0.4288,
+      "step": 565
+    },
+    {
+      "epoch": 0.6609254123485623,
+      "grad_norm": 0.6040019305145549,
+      "learning_rate": 1.6940058851358344e-05,
+      "loss": 0.419,
+      "step": 566
+    },
+    {
+      "epoch": 0.6620931250912275,
+      "grad_norm": 0.6525769499817149,
+      "learning_rate": 1.692535708082412e-05,
+      "loss": 0.4614,
+      "step": 567
+    },
+    {
+      "epoch": 0.6632608378338929,
+      "grad_norm": 0.6375193505269245,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.4133,
+      "step": 568
+    },
+    {
+      "epoch": 0.6644285505765581,
+      "grad_norm": 0.6083705887919718,
+      "learning_rate": 1.6895867139794455e-05,
+      "loss": 0.4425,
+      "step": 569
+    },
+    {
+      "epoch": 0.6655962633192235,
+      "grad_norm": 0.6150118037592329,
+      "learning_rate": 1.6881079092023742e-05,
+      "loss": 0.4397,
+      "step": 570
+    },
+    {
+      "epoch": 0.6667639760618888,
+      "grad_norm": 0.5887479132528075,
+      "learning_rate": 1.6866262408098134e-05,
+      "loss": 0.4302,
+      "step": 571
+    },
+    {
+      "epoch": 0.6679316888045541,
+      "grad_norm": 0.5789800359421425,
+      "learning_rate": 1.6851417149678442e-05,
+      "loss": 0.4083,
+      "step": 572
+    },
+    {
+      "epoch": 0.6690994015472194,
+      "grad_norm": 0.5731529863759083,
+      "learning_rate": 1.683654337854438e-05,
+      "loss": 0.3964,
+      "step": 573
+    },
+    {
+      "epoch": 0.6702671142898847,
+      "grad_norm": 0.6131922082274642,
+      "learning_rate": 1.682164115659432e-05,
+      "loss": 0.435,
+      "step": 574
+    },
+    {
+      "epoch": 0.67143482703255,
+      "grad_norm": 0.5896696200486451,
+      "learning_rate": 1.6806710545845036e-05,
+      "loss": 0.4249,
+      "step": 575
+    },
+    {
+      "epoch": 0.6726025397752153,
+      "grad_norm": 0.6070157798125584,
+      "learning_rate": 1.679175160843145e-05,
+      "loss": 0.4489,
+      "step": 576
+    },
+    {
+      "epoch": 0.6737702525178806,
+      "grad_norm": 0.5840556176463093,
+      "learning_rate": 1.6776764406606362e-05,
+      "loss": 0.4427,
+      "step": 577
+    },
+    {
+      "epoch": 0.674937965260546,
+      "grad_norm": 0.638669263026154,
+      "learning_rate": 1.6761749002740195e-05,
+      "loss": 0.4354,
+      "step": 578
+    },
+    {
+      "epoch": 0.6761056780032112,
+      "grad_norm": 0.596181689132289,
+      "learning_rate": 1.6746705459320746e-05,
+      "loss": 0.4449,
+      "step": 579
+    },
+    {
+      "epoch": 0.6772733907458766,
+      "grad_norm": 0.6154037015019524,
+      "learning_rate": 1.6731633838952905e-05,
+      "loss": 0.4077,
+      "step": 580
+    },
+    {
+      "epoch": 0.6784411034885418,
+      "grad_norm": 0.6207308882915905,
+      "learning_rate": 1.6716534204358416e-05,
+      "loss": 0.4388,
+      "step": 581
+    },
+    {
+      "epoch": 0.6796088162312072,
+      "grad_norm": 0.6799232632709163,
+      "learning_rate": 1.67014066183756e-05,
+      "loss": 0.4506,
+      "step": 582
+    },
+    {
+      "epoch": 0.6807765289738724,
+      "grad_norm": 0.6456677052802752,
+      "learning_rate": 1.6686251143959096e-05,
+      "loss": 0.4201,
+      "step": 583
+    },
+    {
+      "epoch": 0.6819442417165378,
+      "grad_norm": 0.641086622482457,
+      "learning_rate": 1.6671067844179625e-05,
+      "loss": 0.4548,
+      "step": 584
+    },
+    {
+      "epoch": 0.683111954459203,
+      "grad_norm": 0.6158683660142521,
+      "learning_rate": 1.6655856782223682e-05,
+      "loss": 0.423,
+      "step": 585
+    },
+    {
+      "epoch": 0.6842796672018684,
+      "grad_norm": 0.5991195599952508,
+      "learning_rate": 1.6640618021393305e-05,
+      "loss": 0.4285,
+      "step": 586
+    },
+    {
+      "epoch": 0.6854473799445336,
+      "grad_norm": 0.6322840779693253,
+      "learning_rate": 1.6625351625105796e-05,
+      "loss": 0.4391,
+      "step": 587
+    },
+    {
+      "epoch": 0.686615092687199,
+      "grad_norm": 0.632694775296364,
+      "learning_rate": 1.6610057656893483e-05,
+      "loss": 0.4444,
+      "step": 588
+    },
+    {
+      "epoch": 0.6877828054298643,
+      "grad_norm": 0.567190660582192,
+      "learning_rate": 1.6594736180403412e-05,
+      "loss": 0.4107,
+      "step": 589
+    },
+    {
+      "epoch": 0.6889505181725295,
+      "grad_norm": 0.6352472686869244,
+      "learning_rate": 1.657938725939713e-05,
+      "loss": 0.418,
+      "step": 590
+    },
+    {
+      "epoch": 0.6901182309151949,
+      "grad_norm": 0.5985085328251746,
+      "learning_rate": 1.6564010957750375e-05,
+      "loss": 0.4241,
+      "step": 591
+    },
+    {
+      "epoch": 0.6912859436578601,
+      "grad_norm": 0.6247232926733378,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4524,
+      "step": 592
+    },
+    {
+      "epoch": 0.6924536564005255,
+      "grad_norm": 0.6094537528926182,
+      "learning_rate": 1.653317646860793e-05,
+      "loss": 0.4204,
+      "step": 593
+    },
+    {
+      "epoch": 0.6936213691431907,
+      "grad_norm": 0.6298931439802982,
+      "learning_rate": 1.6517718409432405e-05,
+      "loss": 0.4432,
+      "step": 594
+    },
+    {
+      "epoch": 0.6947890818858561,
+      "grad_norm": 0.6400406363073463,
+      "learning_rate": 1.650223322625621e-05,
+      "loss": 0.4467,
+      "step": 595
+    },
+    {
+      "epoch": 0.6959567946285213,
+      "grad_norm": 0.5941415807085889,
+      "learning_rate": 1.6486720983522156e-05,
+      "loss": 0.417,
+      "step": 596
+    },
+    {
+      "epoch": 0.6971245073711867,
+      "grad_norm": 0.601781920685803,
+      "learning_rate": 1.6471181745785673e-05,
+      "loss": 0.4334,
+      "step": 597
+    },
+    {
+      "epoch": 0.698292220113852,
+      "grad_norm": 0.6239346914818055,
+      "learning_rate": 1.645561557771453e-05,
+      "loss": 0.4363,
+      "step": 598
+    },
+    {
+      "epoch": 0.6994599328565173,
+      "grad_norm": 0.5382457151187909,
+      "learning_rate": 1.6440022544088553e-05,
+      "loss": 0.4125,
+      "step": 599
+    },
+    {
+      "epoch": 0.7006276455991826,
+      "grad_norm": 0.591587416148419,
+      "learning_rate": 1.6424402709799404e-05,
+      "loss": 0.4368,
+      "step": 600
+    },
+    {
+      "epoch": 0.7017953583418479,
+      "grad_norm": 0.6241171329985936,
+      "learning_rate": 1.6408756139850243e-05,
+      "loss": 0.4307,
+      "step": 601
+    },
+    {
+      "epoch": 0.7029630710845132,
+      "grad_norm": 0.5157122517802932,
+      "learning_rate": 1.6393082899355516e-05,
+      "loss": 0.4558,
+      "step": 602
+    },
+    {
+      "epoch": 0.7041307838271785,
+      "grad_norm": 0.5814643533733521,
+      "learning_rate": 1.637738305354065e-05,
+      "loss": 0.4176,
+      "step": 603
+    },
+    {
+      "epoch": 0.7052984965698438,
+      "grad_norm": 0.6361560789071897,
+      "learning_rate": 1.63616566677418e-05,
+      "loss": 0.4708,
+      "step": 604
+    },
+    {
+      "epoch": 0.7064662093125091,
+      "grad_norm": 0.5952345127612693,
+      "learning_rate": 1.634590380740556e-05,
+      "loss": 0.4236,
+      "step": 605
+    },
+    {
+      "epoch": 0.7076339220551744,
+      "grad_norm": 0.6200300089625485,
+      "learning_rate": 1.6330124538088705e-05,
+      "loss": 0.4429,
+      "step": 606
+    },
+    {
+      "epoch": 0.7088016347978398,
+      "grad_norm": 0.6311349994172634,
+      "learning_rate": 1.631431892545791e-05,
+      "loss": 0.4526,
+      "step": 607
+    },
+    {
+      "epoch": 0.709969347540505,
+      "grad_norm": 0.6062415815091826,
+      "learning_rate": 1.629848703528949e-05,
+      "loss": 0.4258,
+      "step": 608
+    },
+    {
+      "epoch": 0.7111370602831704,
+      "grad_norm": 0.5809304099320435,
+      "learning_rate": 1.6282628933469097e-05,
+      "loss": 0.43,
+      "step": 609
+    },
+    {
+      "epoch": 0.7123047730258356,
+      "grad_norm": 0.6368433147467144,
+      "learning_rate": 1.6266744685991487e-05,
+      "loss": 0.4315,
+      "step": 610
+    },
+    {
+      "epoch": 0.713472485768501,
+      "grad_norm": 0.5852663844943176,
+      "learning_rate": 1.6250834358960203e-05,
+      "loss": 0.411,
+      "step": 611
+    },
+    {
+      "epoch": 0.7146401985111662,
+      "grad_norm": 0.5714481038852456,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.4215,
+      "step": 612
+    },
+    {
+      "epoch": 0.7158079112538316,
+      "grad_norm": 0.6152576965893425,
+      "learning_rate": 1.6218935731193223e-05,
+      "loss": 0.4014,
+      "step": 613
+    },
+    {
+      "epoch": 0.7169756239964968,
+      "grad_norm": 0.5691674886924838,
+      "learning_rate": 1.620294756320619e-05,
+      "loss": 0.4069,
+      "step": 614
+    },
+    {
+      "epoch": 0.7181433367391622,
+      "grad_norm": 0.6043504582133349,
+      "learning_rate": 1.6186933581162254e-05,
+      "loss": 0.422,
+      "step": 615
+    },
+    {
+      "epoch": 0.7193110494818274,
+      "grad_norm": 0.6222524955418788,
+      "learning_rate": 1.6170893851704875e-05,
+      "loss": 0.4321,
+      "step": 616
+    },
+    {
+      "epoch": 0.7204787622244928,
+      "grad_norm": 0.6288267766607232,
+      "learning_rate": 1.6154828441584655e-05,
+      "loss": 0.4304,
+      "step": 617
+    },
+    {
+      "epoch": 0.7216464749671581,
+      "grad_norm": 0.5968815322417278,
+      "learning_rate": 1.613873741765907e-05,
+      "loss": 0.4187,
+      "step": 618
+    },
+    {
+      "epoch": 0.7228141877098234,
+      "grad_norm": 0.6055460024411744,
+      "learning_rate": 1.6122620846892182e-05,
+      "loss": 0.4164,
+      "step": 619
+    },
+    {
+      "epoch": 0.7239819004524887,
+      "grad_norm": 0.6544170534374241,
+      "learning_rate": 1.6106478796354382e-05,
+      "loss": 0.4647,
+      "step": 620
+    },
+    {
+      "epoch": 0.725149613195154,
+      "grad_norm": 0.6050834969673194,
+      "learning_rate": 1.6090311333222088e-05,
+      "loss": 0.4349,
+      "step": 621
+    },
+    {
+      "epoch": 0.7263173259378193,
+      "grad_norm": 0.5955409592322323,
+      "learning_rate": 1.607411852477748e-05,
+      "loss": 0.4284,
+      "step": 622
+    },
+    {
+      "epoch": 0.7274850386804846,
+      "grad_norm": 0.5840900534913376,
+      "learning_rate": 1.60579004384082e-05,
+      "loss": 0.4309,
+      "step": 623
+    },
+    {
+      "epoch": 0.7286527514231499,
+      "grad_norm": 0.5621875095154542,
+      "learning_rate": 1.604165714160711e-05,
+      "loss": 0.4165,
+      "step": 624
+    },
+    {
+      "epoch": 0.7298204641658153,
+      "grad_norm": 0.6306844817342947,
+      "learning_rate": 1.6025388701971963e-05,
+      "loss": 0.4508,
+      "step": 625
+    },
+    {
+      "epoch": 0.7309881769084805,
+      "grad_norm": 0.5876860971959289,
+      "learning_rate": 1.600909518720517e-05,
+      "loss": 0.4458,
+      "step": 626
+    },
+    {
+      "epoch": 0.7321558896511459,
+      "grad_norm": 0.6196829932507718,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.4514,
+      "step": 627
+    },
+    {
+      "epoch": 0.7333236023938111,
+      "grad_norm": 0.600212530317224,
+      "learning_rate": 1.597643320360769e-05,
+      "loss": 0.4383,
+      "step": 628
+    },
+    {
+      "epoch": 0.7344913151364765,
+      "grad_norm": 0.7318978804047045,
+      "learning_rate": 1.5960064870702434e-05,
+      "loss": 0.4252,
+      "step": 629
+    },
+    {
+      "epoch": 0.7356590278791417,
+      "grad_norm": 0.6238968545884853,
+      "learning_rate": 1.594367173451582e-05,
+      "loss": 0.4543,
+      "step": 630
+    },
+    {
+      "epoch": 0.7368267406218071,
+      "grad_norm": 0.580281163812814,
+      "learning_rate": 1.5927253863269174e-05,
+      "loss": 0.4323,
+      "step": 631
+    },
+    {
+      "epoch": 0.7379944533644723,
+      "grad_norm": 0.6451496057745228,
+      "learning_rate": 1.5910811325286768e-05,
+      "loss": 0.4374,
+      "step": 632
+    },
+    {
+      "epoch": 0.7391621661071377,
+      "grad_norm": 0.6409906885553216,
+      "learning_rate": 1.5894344188995522e-05,
+      "loss": 0.418,
+      "step": 633
+    },
+    {
+      "epoch": 0.740329878849803,
+      "grad_norm": 0.567986327222503,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.4093,
+      "step": 634
+    },
+    {
+      "epoch": 0.7414975915924683,
+      "grad_norm": 0.6431701966057847,
+      "learning_rate": 1.5861336395705765e-05,
+      "loss": 0.4467,
+      "step": 635
+    },
+    {
+      "epoch": 0.7426653043351336,
+      "grad_norm": 0.6149914303771749,
+      "learning_rate": 1.58447958760718e-05,
+      "loss": 0.4309,
+      "step": 636
+    },
+    {
+      "epoch": 0.7438330170777988,
+      "grad_norm": 0.607998937395883,
+      "learning_rate": 1.5828231032857503e-05,
+      "loss": 0.4371,
+      "step": 637
+    },
+    {
+      "epoch": 0.7450007298204642,
+      "grad_norm": 0.5982391595953834,
+      "learning_rate": 1.581164193499879e-05,
+      "loss": 0.433,
+      "step": 638
+    },
+    {
+      "epoch": 0.7461684425631294,
+      "grad_norm": 0.5821758578815989,
+      "learning_rate": 1.5795028651532502e-05,
+      "loss": 0.4195,
+      "step": 639
+    },
+    {
+      "epoch": 0.7473361553057948,
+      "grad_norm": 0.7008130081492723,
+      "learning_rate": 1.577839125159613e-05,
+      "loss": 0.4325,
+      "step": 640
+    },
+    {
+      "epoch": 0.74850386804846,
+      "grad_norm": 0.5837423474235958,
+      "learning_rate": 1.576172980442753e-05,
+      "loss": 0.422,
+      "step": 641
+    },
+    {
+      "epoch": 0.7496715807911254,
+      "grad_norm": 0.5969738301531713,
+      "learning_rate": 1.5745044379364637e-05,
+      "loss": 0.4113,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508392935337906,
+      "grad_norm": 0.6288920046297114,
+      "learning_rate": 1.572833504584516e-05,
+      "loss": 0.4591,
+      "step": 643
+    },
+    {
+      "epoch": 0.752007006276456,
+      "grad_norm": 0.6277912154619338,
+      "learning_rate": 1.5711601873406314e-05,
+      "loss": 0.3899,
+      "step": 644
+    },
+    {
+      "epoch": 0.7531747190191213,
+      "grad_norm": 0.6111849493442082,
+      "learning_rate": 1.569484493168452e-05,
+      "loss": 0.4562,
+      "step": 645
+    },
+    {
+      "epoch": 0.7543424317617866,
+      "grad_norm": 0.6243502669627149,
+      "learning_rate": 1.5678064290415122e-05,
+      "loss": 0.4147,
+      "step": 646
+    },
+    {
+      "epoch": 0.7555101445044519,
+      "grad_norm": 0.601637089411234,
+      "learning_rate": 1.5661260019432077e-05,
+      "loss": 0.4706,
+      "step": 647
+    },
+    {
+      "epoch": 0.7566778572471172,
+      "grad_norm": 0.5794229378045771,
+      "learning_rate": 1.5644432188667695e-05,
+      "loss": 0.4342,
+      "step": 648
+    },
+    {
+      "epoch": 0.7578455699897825,
+      "grad_norm": 0.6207008274773017,
+      "learning_rate": 1.5627580868152316e-05,
+      "loss": 0.4117,
+      "step": 649
+    },
+    {
+      "epoch": 0.7590132827324478,
+      "grad_norm": 0.5856518459795237,
+      "learning_rate": 1.5610706128014054e-05,
+      "loss": 0.4296,
+      "step": 650
+    },
+    {
+      "epoch": 0.7601809954751131,
+      "grad_norm": 0.5886903394808722,
+      "learning_rate": 1.5593808038478465e-05,
+      "loss": 0.4056,
+      "step": 651
+    },
+    {
+      "epoch": 0.7613487082177784,
+      "grad_norm": 0.6666377743210237,
+      "learning_rate": 1.5576886669868297e-05,
+      "loss": 0.4253,
+      "step": 652
+    },
+    {
+      "epoch": 0.7625164209604437,
+      "grad_norm": 0.591321969672488,
+      "learning_rate": 1.5559942092603152e-05,
+      "loss": 0.4201,
+      "step": 653
+    },
+    {
+      "epoch": 0.7636841337031091,
+      "grad_norm": 0.5544237144966251,
+      "learning_rate": 1.554297437719923e-05,
+      "loss": 0.4225,
+      "step": 654
+    },
+    {
+      "epoch": 0.7648518464457743,
+      "grad_norm": 0.6529183766324635,
+      "learning_rate": 1.5525983594269026e-05,
+      "loss": 0.4243,
+      "step": 655
+    },
+    {
+      "epoch": 0.7660195591884397,
+      "grad_norm": 0.5421852672343195,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.4178,
+      "step": 656
+    },
+    {
+      "epoch": 0.7671872719311049,
+      "grad_norm": 0.6270445841788007,
+      "learning_rate": 1.5491933108759417e-05,
+      "loss": 0.417,
+      "step": 657
+    },
+    {
+      "epoch": 0.7683549846737703,
+      "grad_norm": 0.6139147900354028,
+      "learning_rate": 1.5474873547883803e-05,
+      "loss": 0.437,
+      "step": 658
+    },
+    {
+      "epoch": 0.7695226974164355,
+      "grad_norm": 0.5568284571749346,
+      "learning_rate": 1.545779120288889e-05,
+      "loss": 0.4133,
+      "step": 659
+    },
+    {
+      "epoch": 0.7706904101591009,
+      "grad_norm": 0.6226283646843738,
+      "learning_rate": 1.5440686144864207e-05,
+      "loss": 0.4238,
+      "step": 660
+    },
+    {
+      "epoch": 0.7718581229017661,
+      "grad_norm": 0.627184893064876,
+      "learning_rate": 1.5423558444993808e-05,
+      "loss": 0.417,
+      "step": 661
+    },
+    {
+      "epoch": 0.7730258356444315,
+      "grad_norm": 0.5746511458382713,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4094,
+      "step": 662
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.6328910400806285,
+      "learning_rate": 1.5389235404922912e-05,
+      "loss": 0.4224,
+      "step": 663
+    },
+    {
+      "epoch": 0.7753612611297621,
+      "grad_norm": 0.5950883842295158,
+      "learning_rate": 1.5372040207560457e-05,
+      "loss": 0.4473,
+      "step": 664
+    },
+    {
+      "epoch": 0.7765289738724274,
+      "grad_norm": 0.5948288216098573,
+      "learning_rate": 1.5354822654027787e-05,
+      "loss": 0.4595,
+      "step": 665
+    },
+    {
+      "epoch": 0.7776966866150927,
+      "grad_norm": 0.6178195818360889,
+      "learning_rate": 1.5337582815977102e-05,
+      "loss": 0.4078,
+      "step": 666
+    },
+    {
+      "epoch": 0.778864399357758,
+      "grad_norm": 0.6368464925375956,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.4376,
+      "step": 667
+    },
+    {
+      "epoch": 0.7800321121004233,
+      "grad_norm": 0.650427870481684,
+      "learning_rate": 1.5303036573393964e-05,
+      "loss": 0.4333,
+      "step": 668
+    },
+    {
+      "epoch": 0.7811998248430886,
+      "grad_norm": 0.6809056810495064,
+      "learning_rate": 1.528573031262842e-05,
+      "loss": 0.4519,
+      "step": 669
+    },
+    {
+      "epoch": 0.782367537585754,
+      "grad_norm": 0.5883228386504503,
+      "learning_rate": 1.5268402054878116e-05,
+      "loss": 0.4144,
+      "step": 670
+    },
+    {
+      "epoch": 0.7835352503284192,
+      "grad_norm": 0.7244958151779798,
+      "learning_rate": 1.5251051872255963e-05,
+      "loss": 0.4319,
+      "step": 671
+    },
+    {
+      "epoch": 0.7847029630710846,
+      "grad_norm": 0.6400922716376285,
+      "learning_rate": 1.5233679836966122e-05,
+      "loss": 0.455,
+      "step": 672
+    },
+    {
+      "epoch": 0.7858706758137498,
+      "grad_norm": 0.5985112329199814,
+      "learning_rate": 1.5216286021303685e-05,
+      "loss": 0.4536,
+      "step": 673
+    },
+    {
+      "epoch": 0.7870383885564152,
+      "grad_norm": 0.7159502385394264,
+      "learning_rate": 1.5198870497654397e-05,
+      "loss": 0.4764,
+      "step": 674
+    },
+    {
+      "epoch": 0.7882061012990804,
+      "grad_norm": 0.6055486382930659,
+      "learning_rate": 1.5181433338494337e-05,
+      "loss": 0.4236,
+      "step": 675
+    },
+    {
+      "epoch": 0.7893738140417458,
+      "grad_norm": 0.5819176655951349,
+      "learning_rate": 1.5163974616389621e-05,
+      "loss": 0.4166,
+      "step": 676
+    },
+    {
+      "epoch": 0.790541526784411,
+      "grad_norm": 0.6738474843469546,
+      "learning_rate": 1.5146494403996101e-05,
+      "loss": 0.4265,
+      "step": 677
+    },
+    {
+      "epoch": 0.7917092395270764,
+      "grad_norm": 0.6753310502178633,
+      "learning_rate": 1.5128992774059063e-05,
+      "loss": 0.4314,
+      "step": 678
+    },
+    {
+      "epoch": 0.7928769522697416,
+      "grad_norm": 0.6492754240424418,
+      "learning_rate": 1.5111469799412927e-05,
+      "loss": 0.4176,
+      "step": 679
+    },
+    {
+      "epoch": 0.794044665012407,
+      "grad_norm": 0.6726627458239045,
+      "learning_rate": 1.5093925552980934e-05,
+      "loss": 0.4361,
+      "step": 680
+    },
+    {
+      "epoch": 0.7952123777550723,
+      "grad_norm": 0.6166082026373936,
+      "learning_rate": 1.5076360107774855e-05,
+      "loss": 0.4399,
+      "step": 681
+    },
+    {
+      "epoch": 0.7963800904977375,
+      "grad_norm": 0.6401179780694632,
+      "learning_rate": 1.5058773536894685e-05,
+      "loss": 0.4239,
+      "step": 682
+    },
+    {
+      "epoch": 0.7975478032404029,
+      "grad_norm": 0.650281254949028,
+      "learning_rate": 1.504116591352832e-05,
+      "loss": 0.4208,
+      "step": 683
+    },
+    {
+      "epoch": 0.7987155159830681,
+      "grad_norm": 0.6392511895860209,
+      "learning_rate": 1.5023537310951284e-05,
+      "loss": 0.4282,
+      "step": 684
+    },
+    {
+      "epoch": 0.7998832287257335,
+      "grad_norm": 0.6046422945718788,
+      "learning_rate": 1.5005887802526398e-05,
+      "loss": 0.4657,
+      "step": 685
+    },
+    {
+      "epoch": 0.8010509414683987,
+      "grad_norm": 0.576513120962615,
+      "learning_rate": 1.4988217461703488e-05,
+      "loss": 0.402,
+      "step": 686
+    },
+    {
+      "epoch": 0.8022186542110641,
+      "grad_norm": 0.610097829578813,
+      "learning_rate": 1.4970526362019077e-05,
+      "loss": 0.4614,
+      "step": 687
+    },
+    {
+      "epoch": 0.8033863669537293,
+      "grad_norm": 0.5785820776185246,
+      "learning_rate": 1.4952814577096073e-05,
+      "loss": 0.4066,
+      "step": 688
+    },
+    {
+      "epoch": 0.8045540796963947,
+      "grad_norm": 0.5942194205124953,
+      "learning_rate": 1.493508218064347e-05,
+      "loss": 0.3974,
+      "step": 689
+    },
+    {
+      "epoch": 0.80572179243906,
+      "grad_norm": 0.6229039418472488,
+      "learning_rate": 1.491732924645604e-05,
+      "loss": 0.4413,
+      "step": 690
+    },
+    {
+      "epoch": 0.8068895051817253,
+      "grad_norm": 0.615320226144004,
+      "learning_rate": 1.489955584841402e-05,
+      "loss": 0.4348,
+      "step": 691
+    },
+    {
+      "epoch": 0.8080572179243906,
+      "grad_norm": 0.5923430708739851,
+      "learning_rate": 1.4881762060482814e-05,
+      "loss": 0.4031,
+      "step": 692
+    },
+    {
+      "epoch": 0.8092249306670559,
+      "grad_norm": 0.5684305753446505,
+      "learning_rate": 1.4863947956712679e-05,
+      "loss": 0.4227,
+      "step": 693
+    },
+    {
+      "epoch": 0.8103926434097212,
+      "grad_norm": 0.7002076874792758,
+      "learning_rate": 1.4846113611238413e-05,
+      "loss": 0.402,
+      "step": 694
+    },
+    {
+      "epoch": 0.8115603561523865,
+      "grad_norm": 0.6370531689583191,
+      "learning_rate": 1.4828259098279059e-05,
+      "loss": 0.4687,
+      "step": 695
+    },
+    {
+      "epoch": 0.8127280688950518,
+      "grad_norm": 0.5537025393042143,
+      "learning_rate": 1.4810384492137582e-05,
+      "loss": 0.4192,
+      "step": 696
+    },
+    {
+      "epoch": 0.8138957816377171,
+      "grad_norm": 0.6413370037958742,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.4396,
+      "step": 697
+    },
+    {
+      "epoch": 0.8150634943803824,
+      "grad_norm": 0.6170911513634153,
+      "learning_rate": 1.4774575297937921e-05,
+      "loss": 0.4112,
+      "step": 698
+    },
+    {
+      "epoch": 0.8162312071230478,
+      "grad_norm": 0.5963356341190809,
+      "learning_rate": 1.4756640858902531e-05,
+      "loss": 0.4354,
+      "step": 699
+    },
+    {
+      "epoch": 0.817398919865713,
+      "grad_norm": 0.5685912734128744,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.4126,
+      "step": 700
+    },
+    {
+      "epoch": 0.8185666326083784,
+      "grad_norm": 0.6169284360113975,
+      "learning_rate": 1.4720712670138255e-05,
+      "loss": 0.419,
+      "step": 701
+    },
+    {
+      "epoch": 0.8197343453510436,
+      "grad_norm": 0.5881423358539065,
+      "learning_rate": 1.4702719069927368e-05,
+      "loss": 0.4131,
+      "step": 702
+    },
+    {
+      "epoch": 0.820902058093709,
+      "grad_norm": 0.537296622358385,
+      "learning_rate": 1.4684705898979119e-05,
+      "loss": 0.4062,
+      "step": 703
+    },
+    {
+      "epoch": 0.8220697708363742,
+      "grad_norm": 0.5910135264771016,
+      "learning_rate": 1.4666673232256738e-05,
+      "loss": 0.4114,
+      "step": 704
+    },
+    {
+      "epoch": 0.8232374835790396,
+      "grad_norm": 0.6263091753621015,
+      "learning_rate": 1.46486211448046e-05,
+      "loss": 0.4276,
+      "step": 705
+    },
+    {
+      "epoch": 0.8244051963217048,
+      "grad_norm": 0.5995024511431367,
+      "learning_rate": 1.4630549711747888e-05,
+      "loss": 0.4081,
+      "step": 706
+    },
+    {
+      "epoch": 0.8255729090643702,
+      "grad_norm": 0.6588567847444406,
+      "learning_rate": 1.4612459008292307e-05,
+      "loss": 0.4057,
+      "step": 707
+    },
+    {
+      "epoch": 0.8267406218070354,
+      "grad_norm": 0.6174320368022668,
+      "learning_rate": 1.4594349109723744e-05,
+      "loss": 0.4117,
+      "step": 708
+    },
+    {
+      "epoch": 0.8279083345497008,
+      "grad_norm": 0.5752563455386189,
+      "learning_rate": 1.4576220091407981e-05,
+      "loss": 0.4463,
+      "step": 709
+    },
+    {
+      "epoch": 0.8290760472923661,
+      "grad_norm": 0.6487059639784346,
+      "learning_rate": 1.4558072028790356e-05,
+      "loss": 0.4178,
+      "step": 710
+    },
+    {
+      "epoch": 0.8302437600350314,
+      "grad_norm": 0.5526172083915128,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.4236,
+      "step": 711
+    },
+    {
+      "epoch": 0.8314114727776967,
+      "grad_norm": 0.5678179128437394,
+      "learning_rate": 1.4521719072826858e-05,
+      "loss": 0.398,
+      "step": 712
+    },
+    {
+      "epoch": 0.832579185520362,
+      "grad_norm": 0.6014865518827358,
+      "learning_rate": 1.4503514330766689e-05,
+      "loss": 0.4225,
+      "step": 713
+    },
+    {
+      "epoch": 0.8337468982630273,
+      "grad_norm": 0.5634352123001315,
+      "learning_rate": 1.4485290846975431e-05,
+      "loss": 0.4195,
+      "step": 714
+    },
+    {
+      "epoch": 0.8349146110056926,
+      "grad_norm": 0.5955614386510253,
+      "learning_rate": 1.4467048697291554e-05,
+      "loss": 0.4244,
+      "step": 715
+    },
+    {
+      "epoch": 0.8360823237483579,
+      "grad_norm": 0.6246610764060357,
+      "learning_rate": 1.4448787957631209e-05,
+      "loss": 0.4325,
+      "step": 716
+    },
+    {
+      "epoch": 0.8372500364910233,
+      "grad_norm": 0.5989440227809293,
+      "learning_rate": 1.4430508703987914e-05,
+      "loss": 0.4252,
+      "step": 717
+    },
+    {
+      "epoch": 0.8384177492336885,
+      "grad_norm": 0.567243340330328,
+      "learning_rate": 1.4412211012432213e-05,
+      "loss": 0.4212,
+      "step": 718
+    },
+    {
+      "epoch": 0.8395854619763539,
+      "grad_norm": 0.6108726152989146,
+      "learning_rate": 1.4393894959111415e-05,
+      "loss": 0.4407,
+      "step": 719
+    },
+    {
+      "epoch": 0.8407531747190191,
+      "grad_norm": 0.5984973714627059,
+      "learning_rate": 1.4375560620249209e-05,
+      "loss": 0.3997,
+      "step": 720
+    },
+    {
+      "epoch": 0.8419208874616845,
+      "grad_norm": 0.602766451738862,
+      "learning_rate": 1.4357208072145405e-05,
+      "loss": 0.4157,
+      "step": 721
+    },
+    {
+      "epoch": 0.8430886002043497,
+      "grad_norm": 0.6584782229027845,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.3862,
+      "step": 722
+    },
+    {
+      "epoch": 0.8442563129470151,
+      "grad_norm": 0.5763481222289392,
+      "learning_rate": 1.4320448653790782e-05,
+      "loss": 0.4299,
+      "step": 723
+    },
+    {
+      "epoch": 0.8454240256896803,
+      "grad_norm": 0.6670960215460137,
+      "learning_rate": 1.430204193651719e-05,
+      "loss": 0.4607,
+      "step": 724
+    },
+    {
+      "epoch": 0.8465917384323457,
+      "grad_norm": 0.6280880451402773,
+      "learning_rate": 1.4283617315955815e-05,
+      "loss": 0.4306,
+      "step": 725
+    },
+    {
+      "epoch": 0.847759451175011,
+      "grad_norm": 0.5946538431848621,
+      "learning_rate": 1.4265174868782172e-05,
+      "loss": 0.4259,
+      "step": 726
+    },
+    {
+      "epoch": 0.8489271639176762,
+      "grad_norm": 0.5948729744103158,
+      "learning_rate": 1.4246714671745964e-05,
+      "loss": 0.4302,
+      "step": 727
+    },
+    {
+      "epoch": 0.8500948766603416,
+      "grad_norm": 0.6332147467756055,
+      "learning_rate": 1.4228236801670762e-05,
+      "loss": 0.4264,
+      "step": 728
+    },
+    {
+      "epoch": 0.8512625894030068,
+      "grad_norm": 0.5921843870688323,
+      "learning_rate": 1.4209741335453688e-05,
+      "loss": 0.4173,
+      "step": 729
+    },
+    {
+      "epoch": 0.8524303021456722,
+      "grad_norm": 0.6209430507447861,
+      "learning_rate": 1.4191228350065079e-05,
+      "loss": 0.4164,
+      "step": 730
+    },
+    {
+      "epoch": 0.8535980148883374,
+      "grad_norm": 0.6226768979843091,
+      "learning_rate": 1.417269792254819e-05,
+      "loss": 0.4342,
+      "step": 731
+    },
+    {
+      "epoch": 0.8547657276310028,
+      "grad_norm": 0.6330427410366011,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4138,
+      "step": 732
+    },
+    {
+      "epoch": 0.855933440373668,
+      "grad_norm": 0.606268222308281,
+      "learning_rate": 1.4135585049665207e-05,
+      "loss": 0.4361,
+      "step": 733
+    },
+    {
+      "epoch": 0.8571011531163334,
+      "grad_norm": 0.6042131133476056,
+      "learning_rate": 1.4117002758747268e-05,
+      "loss": 0.4184,
+      "step": 734
+    },
+    {
+      "epoch": 0.8582688658589986,
+      "grad_norm": 0.6096946354132102,
+      "learning_rate": 1.409840333459672e-05,
+      "loss": 0.4042,
+      "step": 735
+    },
+    {
+      "epoch": 0.859436578601664,
+      "grad_norm": 0.6033482390431533,
+      "learning_rate": 1.4079786854616537e-05,
+      "loss": 0.4397,
+      "step": 736
+    },
+    {
+      "epoch": 0.8606042913443293,
+      "grad_norm": 0.6074665547028244,
+      "learning_rate": 1.4061153396280673e-05,
+      "loss": 0.414,
+      "step": 737
+    },
+    {
+      "epoch": 0.8617720040869946,
+      "grad_norm": 0.5546228196016744,
+      "learning_rate": 1.4042503037133736e-05,
+      "loss": 0.4036,
+      "step": 738
+    },
+    {
+      "epoch": 0.8629397168296599,
+      "grad_norm": 0.571807041507257,
+      "learning_rate": 1.4023835854790682e-05,
+      "loss": 0.4256,
+      "step": 739
+    },
+    {
+      "epoch": 0.8641074295723252,
+      "grad_norm": 0.658832459081279,
+      "learning_rate": 1.400515192693645e-05,
+      "loss": 0.3926,
+      "step": 740
+    },
+    {
+      "epoch": 0.8652751423149905,
+      "grad_norm": 0.5934072039883117,
+      "learning_rate": 1.3986451331325697e-05,
+      "loss": 0.4129,
+      "step": 741
+    },
+    {
+      "epoch": 0.8664428550576558,
+      "grad_norm": 0.6560373310607097,
+      "learning_rate": 1.3967734145782425e-05,
+      "loss": 0.4717,
+      "step": 742
+    },
+    {
+      "epoch": 0.8676105678003211,
+      "grad_norm": 0.5518385704958986,
+      "learning_rate": 1.3949000448199686e-05,
+      "loss": 0.4057,
+      "step": 743
+    },
+    {
+      "epoch": 0.8687782805429864,
+      "grad_norm": 0.6102042571990014,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.4211,
+      "step": 744
+    },
+    {
+      "epoch": 0.8699459932856517,
+      "grad_norm": 0.6094883250865959,
+      "learning_rate": 1.391148382883124e-05,
+      "loss": 0.4193,
+      "step": 745
+    },
+    {
+      "epoch": 0.8711137060283171,
+      "grad_norm": 0.5893650344230299,
+      "learning_rate": 1.3892701063173917e-05,
+      "loss": 0.4381,
+      "step": 746
+    },
+    {
+      "epoch": 0.8722814187709823,
+      "grad_norm": 0.6249512908345096,
+      "learning_rate": 1.3873902097733231e-05,
+      "loss": 0.4425,
+      "step": 747
+    },
+    {
+      "epoch": 0.8734491315136477,
+      "grad_norm": 0.5569629213629326,
+      "learning_rate": 1.3855087010742563e-05,
+      "loss": 0.4121,
+      "step": 748
+    },
+    {
+      "epoch": 0.8746168442563129,
+      "grad_norm": 0.5827146431416513,
+      "learning_rate": 1.3836255880502389e-05,
+      "loss": 0.4198,
+      "step": 749
+    },
+    {
+      "epoch": 0.8757845569989783,
+      "grad_norm": 0.6043296229117744,
+      "learning_rate": 1.3817408785379944e-05,
+      "loss": 0.4064,
+      "step": 750
+    },
+    {
+      "epoch": 0.8769522697416435,
+      "grad_norm": 0.5506717666247877,
+      "learning_rate": 1.379854580380891e-05,
+      "loss": 0.4228,
+      "step": 751
+    },
+    {
+      "epoch": 0.8781199824843089,
+      "grad_norm": 0.6123798390890531,
+      "learning_rate": 1.3779667014289067e-05,
+      "loss": 0.4331,
+      "step": 752
+    },
+    {
+      "epoch": 0.8792876952269741,
+      "grad_norm": 0.6794102241050839,
+      "learning_rate": 1.3760772495385998e-05,
+      "loss": 0.4463,
+      "step": 753
+    },
+    {
+      "epoch": 0.8804554079696395,
+      "grad_norm": 0.5457829554536283,
+      "learning_rate": 1.3741862325730739e-05,
+      "loss": 0.3987,
+      "step": 754
+    },
+    {
+      "epoch": 0.8816231207123048,
+      "grad_norm": 0.5690716360505756,
+      "learning_rate": 1.3722936584019453e-05,
+      "loss": 0.4548,
+      "step": 755
+    },
+    {
+      "epoch": 0.8827908334549701,
+      "grad_norm": 0.5760191102601327,
+      "learning_rate": 1.3703995349013113e-05,
+      "loss": 0.4028,
+      "step": 756
+    },
+    {
+      "epoch": 0.8839585461976354,
+      "grad_norm": 0.5836619386107424,
+      "learning_rate": 1.3685038699537165e-05,
+      "loss": 0.4654,
+      "step": 757
+    },
+    {
+      "epoch": 0.8851262589403007,
+      "grad_norm": 0.5717008862983266,
+      "learning_rate": 1.3666066714481205e-05,
+      "loss": 0.41,
+      "step": 758
+    },
+    {
+      "epoch": 0.886293971682966,
+      "grad_norm": 0.5628128032263553,
+      "learning_rate": 1.3647079472798652e-05,
+      "loss": 0.4162,
+      "step": 759
+    },
+    {
+      "epoch": 0.8874616844256313,
+      "grad_norm": 0.5876854641765936,
+      "learning_rate": 1.362807705350641e-05,
+      "loss": 0.4621,
+      "step": 760
+    },
+    {
+      "epoch": 0.8886293971682966,
+      "grad_norm": 0.5599562886818765,
+      "learning_rate": 1.3609059535684557e-05,
+      "loss": 0.4096,
+      "step": 761
+    },
+    {
+      "epoch": 0.889797109910962,
+      "grad_norm": 0.6471027054523758,
+      "learning_rate": 1.3590026998475987e-05,
+      "loss": 0.4338,
+      "step": 762
+    },
+    {
+      "epoch": 0.8909648226536272,
+      "grad_norm": 0.5880307730030826,
+      "learning_rate": 1.3570979521086118e-05,
+      "loss": 0.4121,
+      "step": 763
+    },
+    {
+      "epoch": 0.8921325353962926,
+      "grad_norm": 0.5170849213683774,
+      "learning_rate": 1.355191718278253e-05,
+      "loss": 0.4083,
+      "step": 764
+    },
+    {
+      "epoch": 0.8933002481389578,
+      "grad_norm": 0.5973972340714483,
+      "learning_rate": 1.3532840062894654e-05,
+      "loss": 0.4012,
+      "step": 765
+    },
+    {
+      "epoch": 0.8944679608816232,
+      "grad_norm": 0.7002774338064259,
+      "learning_rate": 1.3513748240813429e-05,
+      "loss": 0.4271,
+      "step": 766
+    },
+    {
+      "epoch": 0.8956356736242884,
+      "grad_norm": 0.5641739244204268,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.4023,
+      "step": 767
+    },
+    {
+      "epoch": 0.8968033863669538,
+      "grad_norm": 0.5520876464021133,
+      "learning_rate": 1.3475520807940303e-05,
+      "loss": 0.3876,
+      "step": 768
+    },
+    {
+      "epoch": 0.897971099109619,
+      "grad_norm": 0.6476806540588114,
+      "learning_rate": 1.345638535623489e-05,
+      "loss": 0.4037,
+      "step": 769
+    },
+    {
+      "epoch": 0.8991388118522844,
+      "grad_norm": 0.5548494662299304,
+      "learning_rate": 1.3437235520508434e-05,
+      "loss": 0.3992,
+      "step": 770
+    },
+    {
+      "epoch": 0.9003065245949496,
+      "grad_norm": 0.5478624889860081,
+      "learning_rate": 1.3418071380454501e-05,
+      "loss": 0.3647,
+      "step": 771
+    },
+    {
+      "epoch": 0.9014742373376149,
+      "grad_norm": 0.5988272710508927,
+      "learning_rate": 1.3398893015826166e-05,
+      "loss": 0.41,
+      "step": 772
+    },
+    {
+      "epoch": 0.9026419500802803,
+      "grad_norm": 0.6817993616809158,
+      "learning_rate": 1.337970050643572e-05,
+      "loss": 0.4402,
+      "step": 773
+    },
+    {
+      "epoch": 0.9038096628229455,
+      "grad_norm": 0.5616562432885858,
+      "learning_rate": 1.3360493932154301e-05,
+      "loss": 0.4461,
+      "step": 774
+    },
+    {
+      "epoch": 0.9049773755656109,
+      "grad_norm": 0.5928316549383963,
+      "learning_rate": 1.3341273372911591e-05,
+      "loss": 0.4152,
+      "step": 775
+    },
+    {
+      "epoch": 0.9061450883082761,
+      "grad_norm": 0.6082213299743613,
+      "learning_rate": 1.3322038908695466e-05,
+      "loss": 0.4045,
+      "step": 776
+    },
+    {
+      "epoch": 0.9073128010509415,
+      "grad_norm": 0.6082380767418752,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.4272,
+      "step": 777
+    },
+    {
+      "epoch": 0.9084805137936067,
+      "grad_norm": 0.5935113777300552,
+      "learning_rate": 1.3283528585583484e-05,
+      "loss": 0.4169,
+      "step": 778
+    },
+    {
+      "epoch": 0.9096482265362721,
+      "grad_norm": 0.6457630234454999,
+      "learning_rate": 1.326425288695138e-05,
+      "loss": 0.4431,
+      "step": 779
+    },
+    {
+      "epoch": 0.9108159392789373,
+      "grad_norm": 0.581329682979585,
+      "learning_rate": 1.3244963603872707e-05,
+      "loss": 0.415,
+      "step": 780
+    },
+    {
+      "epoch": 0.9119836520216027,
+      "grad_norm": 0.5671491201970689,
+      "learning_rate": 1.3225660816621342e-05,
+      "loss": 0.4123,
+      "step": 781
+    },
+    {
+      "epoch": 0.913151364764268,
+      "grad_norm": 0.5555280397310436,
+      "learning_rate": 1.3206344605527355e-05,
+      "loss": 0.4409,
+      "step": 782
+    },
+    {
+      "epoch": 0.9143190775069333,
+      "grad_norm": 0.6755762576582353,
+      "learning_rate": 1.3187015050976696e-05,
+      "loss": 0.4272,
+      "step": 783
+    },
+    {
+      "epoch": 0.9154867902495986,
+      "grad_norm": 0.6190896438996474,
+      "learning_rate": 1.3167672233410826e-05,
+      "loss": 0.425,
+      "step": 784
+    },
+    {
+      "epoch": 0.9166545029922639,
+      "grad_norm": 0.5677512728296183,
+      "learning_rate": 1.3148316233326414e-05,
+      "loss": 0.4302,
+      "step": 785
+    },
+    {
+      "epoch": 0.9178222157349292,
+      "grad_norm": 0.6704941579384655,
+      "learning_rate": 1.312894713127499e-05,
+      "loss": 0.4115,
+      "step": 786
+    },
+    {
+      "epoch": 0.9189899284775945,
+      "grad_norm": 0.6046880657802479,
+      "learning_rate": 1.3109565007862597e-05,
+      "loss": 0.4361,
+      "step": 787
+    },
+    {
+      "epoch": 0.9201576412202598,
+      "grad_norm": 0.6163765093950883,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.4031,
+      "step": 788
+    },
+    {
+      "epoch": 0.9213253539629251,
+      "grad_norm": 0.5666346224998433,
+      "learning_rate": 1.3070762019649722e-05,
+      "loss": 0.4227,
+      "step": 789
+    },
+    {
+      "epoch": 0.9224930667055904,
+      "grad_norm": 0.5631919948018209,
+      "learning_rate": 1.3051341316330946e-05,
+      "loss": 0.4259,
+      "step": 790
+    },
+    {
+      "epoch": 0.9236607794482558,
+      "grad_norm": 0.5933431439860977,
+      "learning_rate": 1.303190791461394e-05,
+      "loss": 0.4575,
+      "step": 791
+    },
+    {
+      "epoch": 0.924828492190921,
+      "grad_norm": 0.5805252977306451,
+      "learning_rate": 1.3012461895372343e-05,
+      "loss": 0.4034,
+      "step": 792
+    },
+    {
+      "epoch": 0.9259962049335864,
+      "grad_norm": 0.577306703661629,
+      "learning_rate": 1.2993003339532304e-05,
+      "loss": 0.3958,
+      "step": 793
+    },
+    {
+      "epoch": 0.9271639176762516,
+      "grad_norm": 0.5406376549072268,
+      "learning_rate": 1.2973532328072138e-05,
+      "loss": 0.3955,
+      "step": 794
+    },
+    {
+      "epoch": 0.928331630418917,
+      "grad_norm": 0.5548080233906243,
+      "learning_rate": 1.2954048942022002e-05,
+      "loss": 0.4291,
+      "step": 795
+    },
+    {
+      "epoch": 0.9294993431615822,
+      "grad_norm": 0.5631226352440963,
+      "learning_rate": 1.2934553262463548e-05,
+      "loss": 0.4361,
+      "step": 796
+    },
+    {
+      "epoch": 0.9306670559042476,
+      "grad_norm": 0.5557173769341457,
+      "learning_rate": 1.2915045370529586e-05,
+      "loss": 0.4133,
+      "step": 797
+    },
+    {
+      "epoch": 0.9318347686469128,
+      "grad_norm": 0.5648221510582228,
+      "learning_rate": 1.2895525347403757e-05,
+      "loss": 0.4164,
+      "step": 798
+    },
+    {
+      "epoch": 0.9330024813895782,
+      "grad_norm": 0.5573280827345388,
+      "learning_rate": 1.2875993274320173e-05,
+      "loss": 0.4066,
+      "step": 799
+    },
+    {
+      "epoch": 0.9341701941322434,
+      "grad_norm": 0.5858156073077699,
+      "learning_rate": 1.285644923256311e-05,
+      "loss": 0.4185,
+      "step": 800
+    },
+    {
+      "epoch": 0.9353379068749088,
+      "grad_norm": 0.5871992148534396,
+      "learning_rate": 1.2836893303466636e-05,
+      "loss": 0.4219,
+      "step": 801
+    },
+    {
+      "epoch": 0.9365056196175741,
+      "grad_norm": 0.6086992636970817,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.4549,
+      "step": 802
+    },
+    {
+      "epoch": 0.9376733323602394,
+      "grad_norm": 0.5933560117019424,
+      "learning_rate": 1.2797746108838777e-05,
+      "loss": 0.425,
+      "step": 803
+    },
+    {
+      "epoch": 0.9388410451029047,
+      "grad_norm": 0.546062510457357,
+      "learning_rate": 1.277815500622154e-05,
+      "loss": 0.3948,
+      "step": 804
+    },
+    {
+      "epoch": 0.94000875784557,
+      "grad_norm": 0.5581866780877199,
+      "learning_rate": 1.275855234209251e-05,
+      "loss": 0.3965,
+      "step": 805
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.5522470019669262,
+      "learning_rate": 1.2738938198029725e-05,
+      "loss": 0.3706,
+      "step": 806
+    },
+    {
+      "epoch": 0.9423441833309006,
+      "grad_norm": 0.5578029652636521,
+      "learning_rate": 1.2719312655658995e-05,
+      "loss": 0.4293,
+      "step": 807
+    },
+    {
+      "epoch": 0.9435118960735659,
+      "grad_norm": 0.5547459105149019,
+      "learning_rate": 1.269967579665357e-05,
+      "loss": 0.3894,
+      "step": 808
+    },
+    {
+      "epoch": 0.9446796088162313,
+      "grad_norm": 0.595609667626807,
+      "learning_rate": 1.2680027702733791e-05,
+      "loss": 0.4393,
+      "step": 809
+    },
+    {
+      "epoch": 0.9458473215588965,
+      "grad_norm": 0.5949164412809517,
+      "learning_rate": 1.2660368455666752e-05,
+      "loss": 0.4167,
+      "step": 810
+    },
+    {
+      "epoch": 0.9470150343015619,
+      "grad_norm": 0.5743695587976126,
+      "learning_rate": 1.2640698137265967e-05,
+      "loss": 0.4171,
+      "step": 811
+    },
+    {
+      "epoch": 0.9481827470442271,
+      "grad_norm": 0.5655124301060094,
+      "learning_rate": 1.2621016829391022e-05,
+      "loss": 0.4154,
+      "step": 812
+    },
+    {
+      "epoch": 0.9493504597868925,
+      "grad_norm": 0.5869535402152061,
+      "learning_rate": 1.2601324613947239e-05,
+      "loss": 0.4112,
+      "step": 813
+    },
+    {
+      "epoch": 0.9505181725295577,
+      "grad_norm": 0.603657955306491,
+      "learning_rate": 1.2581621572885324e-05,
+      "loss": 0.4194,
+      "step": 814
+    },
+    {
+      "epoch": 0.9516858852722231,
+      "grad_norm": 0.5706949699595998,
+      "learning_rate": 1.2561907788201042e-05,
+      "loss": 0.4005,
+      "step": 815
+    },
+    {
+      "epoch": 0.9528535980148883,
+      "grad_norm": 0.6336218874875033,
+      "learning_rate": 1.2542183341934873e-05,
+      "loss": 0.4224,
+      "step": 816
+    },
+    {
+      "epoch": 0.9540213107575536,
+      "grad_norm": 0.6217078883859312,
+      "learning_rate": 1.2522448316171651e-05,
+      "loss": 0.4372,
+      "step": 817
+    },
+    {
+      "epoch": 0.955189023500219,
+      "grad_norm": 0.5821634156659715,
+      "learning_rate": 1.2502702793040257e-05,
+      "loss": 0.4172,
+      "step": 818
+    },
+    {
+      "epoch": 0.9563567362428842,
+      "grad_norm": 0.66416005208682,
+      "learning_rate": 1.2482946854713238e-05,
+      "loss": 0.4441,
+      "step": 819
+    },
+    {
+      "epoch": 0.9575244489855496,
+      "grad_norm": 0.6107049089533688,
+      "learning_rate": 1.24631805834065e-05,
+      "loss": 0.4223,
+      "step": 820
+    },
+    {
+      "epoch": 0.9586921617282148,
+      "grad_norm": 0.5859349942821666,
+      "learning_rate": 1.2443404061378941e-05,
+      "loss": 0.4369,
+      "step": 821
+    },
+    {
+      "epoch": 0.9598598744708802,
+      "grad_norm": 0.6040567243865704,
+      "learning_rate": 1.2423617370932127e-05,
+      "loss": 0.4166,
+      "step": 822
+    },
+    {
+      "epoch": 0.9610275872135454,
+      "grad_norm": 0.5764627103940527,
+      "learning_rate": 1.2403820594409926e-05,
+      "loss": 0.413,
+      "step": 823
+    },
+    {
+      "epoch": 0.9621952999562108,
+      "grad_norm": 0.5792520187180212,
+      "learning_rate": 1.2384013814198197e-05,
+      "loss": 0.405,
+      "step": 824
+    },
+    {
+      "epoch": 0.963363012698876,
+      "grad_norm": 0.5649912947600113,
+      "learning_rate": 1.2364197112724421e-05,
+      "loss": 0.4245,
+      "step": 825
+    },
+    {
+      "epoch": 0.9645307254415414,
+      "grad_norm": 0.5307384259958211,
+      "learning_rate": 1.2344370572457368e-05,
+      "loss": 0.3992,
+      "step": 826
+    },
+    {
+      "epoch": 0.9656984381842066,
+      "grad_norm": 0.5462868768223944,
+      "learning_rate": 1.2324534275906747e-05,
+      "loss": 0.4018,
+      "step": 827
+    },
+    {
+      "epoch": 0.966866150926872,
+      "grad_norm": 0.5952569857248627,
+      "learning_rate": 1.2304688305622889e-05,
+      "loss": 0.4328,
+      "step": 828
+    },
+    {
+      "epoch": 0.9680338636695373,
+      "grad_norm": 0.5994001932232093,
+      "learning_rate": 1.2284832744196355e-05,
+      "loss": 0.4178,
+      "step": 829
+    },
+    {
+      "epoch": 0.9692015764122026,
+      "grad_norm": 0.5239688171965969,
+      "learning_rate": 1.2264967674257647e-05,
+      "loss": 0.389,
+      "step": 830
+    },
+    {
+      "epoch": 0.9703692891548679,
+      "grad_norm": 0.5918131457640353,
+      "learning_rate": 1.2245093178476813e-05,
+      "loss": 0.461,
+      "step": 831
+    },
+    {
+      "epoch": 0.9715370018975332,
+      "grad_norm": 0.6469328966728428,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.444,
+      "step": 832
+    },
+    {
+      "epoch": 0.9727047146401985,
+      "grad_norm": 0.5607351794874871,
+      "learning_rate": 1.2205316240264812e-05,
+      "loss": 0.4021,
+      "step": 833
+    },
+    {
+      "epoch": 0.9738724273828638,
+      "grad_norm": 0.5761737794435746,
+      "learning_rate": 1.218541396336852e-05,
+      "loss": 0.4326,
+      "step": 834
+    },
+    {
+      "epoch": 0.9750401401255291,
+      "grad_norm": 0.5437029663444433,
+      "learning_rate": 1.2165502591699166e-05,
+      "loss": 0.4468,
+      "step": 835
+    },
+    {
+      "epoch": 0.9762078528681944,
+      "grad_norm": 0.5494040179689569,
+      "learning_rate": 1.2145582208119497e-05,
+      "loss": 0.408,
+      "step": 836
+    },
+    {
+      "epoch": 0.9773755656108597,
+      "grad_norm": 0.5838529972390478,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.4288,
+      "step": 837
+    },
+    {
+      "epoch": 0.9785432783535251,
+      "grad_norm": 0.5387080450194842,
+      "learning_rate": 1.2105714736867391e-05,
+      "loss": 0.395,
+      "step": 838
+    },
+    {
+      "epoch": 0.9797109910961903,
+      "grad_norm": 0.5681512145254656,
+      "learning_rate": 1.2085767815106583e-05,
+      "loss": 0.4286,
+      "step": 839
+    },
+    {
+      "epoch": 0.9808787038388557,
+      "grad_norm": 0.6036030796795506,
+      "learning_rate": 1.2065812213258051e-05,
+      "loss": 0.4248,
+      "step": 840
+    },
+    {
+      "epoch": 0.9820464165815209,
+      "grad_norm": 0.554795678017312,
+      "learning_rate": 1.2045848014368591e-05,
+      "loss": 0.4343,
+      "step": 841
+    },
+    {
+      "epoch": 0.9832141293241863,
+      "grad_norm": 0.5491220354136169,
+      "learning_rate": 1.2025875301520811e-05,
+      "loss": 0.4428,
+      "step": 842
+    },
+    {
+      "epoch": 0.9843818420668515,
+      "grad_norm": 0.551186597876788,
+      "learning_rate": 1.200589415783273e-05,
+      "loss": 0.4139,
+      "step": 843
+    },
+    {
+      "epoch": 0.9855495548095169,
+      "grad_norm": 0.576839341392374,
+      "learning_rate": 1.1985904666457455e-05,
+      "loss": 0.4036,
+      "step": 844
+    },
+    {
+      "epoch": 0.9867172675521821,
+      "grad_norm": 0.5881613459508244,
+      "learning_rate": 1.196590691058284e-05,
+      "loss": 0.4194,
+      "step": 845
+    },
+    {
+      "epoch": 0.9878849802948475,
+      "grad_norm": 0.5434905376024451,
+      "learning_rate": 1.1945900973431127e-05,
+      "loss": 0.3869,
+      "step": 846
+    },
+    {
+      "epoch": 0.9890526930375128,
+      "grad_norm": 0.5751320351193693,
+      "learning_rate": 1.1925886938258606e-05,
+      "loss": 0.3993,
+      "step": 847
+    },
+    {
+      "epoch": 0.9902204057801781,
+      "grad_norm": 0.5927512161654048,
+      "learning_rate": 1.1905864888355264e-05,
+      "loss": 0.4276,
+      "step": 848
+    },
+    {
+      "epoch": 0.9913881185228434,
+      "grad_norm": 0.5637040518655385,
+      "learning_rate": 1.1885834907044444e-05,
+      "loss": 0.425,
+      "step": 849
+    },
+    {
+      "epoch": 0.9925558312655087,
+      "grad_norm": 0.5602759285010025,
+      "learning_rate": 1.1865797077682509e-05,
+      "loss": 0.3884,
+      "step": 850
+    },
+    {
+      "epoch": 0.993723544008174,
+      "grad_norm": 0.5937662909217749,
+      "learning_rate": 1.1845751483658454e-05,
+      "loss": 0.3942,
+      "step": 851
+    },
+    {
+      "epoch": 0.9948912567508393,
+      "grad_norm": 0.5528135440074239,
+      "learning_rate": 1.182569820839362e-05,
+      "loss": 0.3943,
+      "step": 852
+    },
+    {
+      "epoch": 0.9960589694935046,
+      "grad_norm": 0.579345456446773,
+      "learning_rate": 1.1805637335341285e-05,
+      "loss": 0.4213,
+      "step": 853
+    },
+    {
+      "epoch": 0.99722668223617,
+      "grad_norm": 0.5410731674208072,
+      "learning_rate": 1.1785568947986368e-05,
+      "loss": 0.399,
+      "step": 854
+    },
+    {
+      "epoch": 0.9983943949788352,
+      "grad_norm": 0.5438299912738765,
+      "learning_rate": 1.1765493129845047e-05,
+      "loss": 0.3921,
+      "step": 855
+    },
+    {
+      "epoch": 0.9995621077215006,
+      "grad_norm": 0.5379049196861156,
+      "learning_rate": 1.1745409964464425e-05,
+      "loss": 0.4005,
+      "step": 856
+    },
+    {
+      "epoch": 1.000729820464166,
+      "grad_norm": 0.5733027865299759,
+      "learning_rate": 1.1725319535422189e-05,
+      "loss": 0.353,
+      "step": 857
+    },
+    {
+      "epoch": 1.0018975332068312,
+      "grad_norm": 0.7080075762479725,
+      "learning_rate": 1.170522192632624e-05,
+      "loss": 0.3423,
+      "step": 858
+    },
+    {
+      "epoch": 1.0030652459494964,
+      "grad_norm": 0.6328555541418939,
+      "learning_rate": 1.1685117220814373e-05,
+      "loss": 0.3302,
+      "step": 859
+    },
+    {
+      "epoch": 1.0042329586921617,
+      "grad_norm": 0.5534983415264147,
+      "learning_rate": 1.1665005502553912e-05,
+      "loss": 0.3419,
+      "step": 860
+    },
+    {
+      "epoch": 1.0054006714348271,
+      "grad_norm": 0.6265783457734877,
+      "learning_rate": 1.164488685524136e-05,
+      "loss": 0.319,
+      "step": 861
+    },
+    {
+      "epoch": 1.0065683841774924,
+      "grad_norm": 0.6498334807993773,
+      "learning_rate": 1.1624761362602061e-05,
+      "loss": 0.3485,
+      "step": 862
+    },
+    {
+      "epoch": 1.0077360969201576,
+      "grad_norm": 0.6089129622442091,
+      "learning_rate": 1.160462910838984e-05,
+      "loss": 0.3395,
+      "step": 863
+    },
+    {
+      "epoch": 1.008903809662823,
+      "grad_norm": 0.6232977956123427,
+      "learning_rate": 1.1584490176386671e-05,
+      "loss": 0.3612,
+      "step": 864
+    },
+    {
+      "epoch": 1.0100715224054881,
+      "grad_norm": 0.6493759559652524,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.3718,
+      "step": 865
+    },
+    {
+      "epoch": 1.0112392351481536,
+      "grad_norm": 0.6162224908869947,
+      "learning_rate": 1.1544192614273957e-05,
+      "loss": 0.3387,
+      "step": 866
+    },
+    {
+      "epoch": 1.0124069478908189,
+      "grad_norm": 0.5984698123256187,
+      "learning_rate": 1.1524034151865904e-05,
+      "loss": 0.3368,
+      "step": 867
+    },
+    {
+      "epoch": 1.0135746606334841,
+      "grad_norm": 0.5976824051216212,
+      "learning_rate": 1.1503869347069186e-05,
+      "loss": 0.3231,
+      "step": 868
+    },
+    {
+      "epoch": 1.0147423733761494,
+      "grad_norm": 0.5852660307198161,
+      "learning_rate": 1.1483698283801233e-05,
+      "loss": 0.3313,
+      "step": 869
+    },
+    {
+      "epoch": 1.0159100861188148,
+      "grad_norm": 0.6415260152477733,
+      "learning_rate": 1.1463521046005524e-05,
+      "loss": 0.3615,
+      "step": 870
+    },
+    {
+      "epoch": 1.01707779886148,
+      "grad_norm": 0.6159579849248676,
+      "learning_rate": 1.1443337717651221e-05,
+      "loss": 0.3413,
+      "step": 871
+    },
+    {
+      "epoch": 1.0182455116041453,
+      "grad_norm": 0.5852095741831175,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.351,
+      "step": 872
+    },
+    {
+      "epoch": 1.0194132243468106,
+      "grad_norm": 0.5646105863812684,
+      "learning_rate": 1.1402953125269927e-05,
+      "loss": 0.3311,
+      "step": 873
+    },
+    {
+      "epoch": 1.020580937089476,
+      "grad_norm": 0.6067007125213353,
+      "learning_rate": 1.1382752029306606e-05,
+      "loss": 0.3513,
+      "step": 874
+    },
+    {
+      "epoch": 1.0217486498321413,
+      "grad_norm": 0.6281928775013575,
+      "learning_rate": 1.1362545178911346e-05,
+      "loss": 0.3697,
+      "step": 875
+    },
+    {
+      "epoch": 1.0229163625748066,
+      "grad_norm": 0.5946560155206285,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.3307,
+      "step": 876
+    },
+    {
+      "epoch": 1.0240840753174718,
+      "grad_norm": 0.6018703466523194,
+      "learning_rate": 1.132211455121824e-05,
+      "loss": 0.3576,
+      "step": 877
+    },
+    {
+      "epoch": 1.0252517880601373,
+      "grad_norm": 0.6028503636631866,
+      "learning_rate": 1.1301890942175649e-05,
+      "loss": 0.3584,
+      "step": 878
+    },
+    {
+      "epoch": 1.0264195008028025,
+      "grad_norm": 0.5828255636555915,
+      "learning_rate": 1.1281661915210931e-05,
+      "loss": 0.3476,
+      "step": 879
+    },
+    {
+      "epoch": 1.0275872135454678,
+      "grad_norm": 0.5700214914638962,
+      "learning_rate": 1.126142755450878e-05,
+      "loss": 0.3205,
+      "step": 880
+    },
+    {
+      "epoch": 1.028754926288133,
+      "grad_norm": 0.5949531418153122,
+      "learning_rate": 1.1241187944276088e-05,
+      "loss": 0.3326,
+      "step": 881
+    },
+    {
+      "epoch": 1.0299226390307985,
+      "grad_norm": 0.5878359530713991,
+      "learning_rate": 1.12209431687416e-05,
+      "loss": 0.3228,
+      "step": 882
+    },
+    {
+      "epoch": 1.0310903517734638,
+      "grad_norm": 0.5940781431266537,
+      "learning_rate": 1.1200693312155545e-05,
+      "loss": 0.3381,
+      "step": 883
+    },
+    {
+      "epoch": 1.032258064516129,
+      "grad_norm": 0.6215991777370997,
+      "learning_rate": 1.1180438458789305e-05,
+      "loss": 0.3506,
+      "step": 884
+    },
+    {
+      "epoch": 1.0334257772587943,
+      "grad_norm": 0.6256448479053209,
+      "learning_rate": 1.1160178692935054e-05,
+      "loss": 0.37,
+      "step": 885
+    },
+    {
+      "epoch": 1.0345934900014597,
+      "grad_norm": 0.5944215932616467,
+      "learning_rate": 1.1139914098905406e-05,
+      "loss": 0.3155,
+      "step": 886
+    },
+    {
+      "epoch": 1.035761202744125,
+      "grad_norm": 0.6170217956300591,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 0.3343,
+      "step": 887
+    },
+    {
+      "epoch": 1.0369289154867902,
+      "grad_norm": 0.5794626141207991,
+      "learning_rate": 1.1099370763670523e-05,
+      "loss": 0.3245,
+      "step": 888
+    },
+    {
+      "epoch": 1.0380966282294555,
+      "grad_norm": 0.6074912158923124,
+      "learning_rate": 1.107909219118958e-05,
+      "loss": 0.323,
+      "step": 889
+    },
+    {
+      "epoch": 1.039264340972121,
+      "grad_norm": 0.6324321730856886,
+      "learning_rate": 1.1058809127981133e-05,
+      "loss": 0.3432,
+      "step": 890
+    },
+    {
+      "epoch": 1.0404320537147862,
+      "grad_norm": 0.6206385510892002,
+      "learning_rate": 1.1038521658454753e-05,
+      "loss": 0.3221,
+      "step": 891
+    },
+    {
+      "epoch": 1.0415997664574514,
+      "grad_norm": 0.5917754284032108,
+      "learning_rate": 1.1018229867038358e-05,
+      "loss": 0.3122,
+      "step": 892
+    },
+    {
+      "epoch": 1.0427674792001167,
+      "grad_norm": 0.6175102650734178,
+      "learning_rate": 1.0997933838177828e-05,
+      "loss": 0.3303,
+      "step": 893
+    },
+    {
+      "epoch": 1.0439351919427822,
+      "grad_norm": 0.6217787764546113,
+      "learning_rate": 1.0977633656336706e-05,
+      "loss": 0.3533,
+      "step": 894
+    },
+    {
+      "epoch": 1.0451029046854474,
+      "grad_norm": 0.5865760208319161,
+      "learning_rate": 1.0957329405995794e-05,
+      "loss": 0.3202,
+      "step": 895
+    },
+    {
+      "epoch": 1.0462706174281127,
+      "grad_norm": 0.6089001470699003,
+      "learning_rate": 1.0937021171652842e-05,
+      "loss": 0.3275,
+      "step": 896
+    },
+    {
+      "epoch": 1.047438330170778,
+      "grad_norm": 0.6260424311828812,
+      "learning_rate": 1.0916709037822172e-05,
+      "loss": 0.3218,
+      "step": 897
+    },
+    {
+      "epoch": 1.0486060429134434,
+      "grad_norm": 0.5607339357412251,
+      "learning_rate": 1.0896393089034336e-05,
+      "loss": 0.3286,
+      "step": 898
+    },
+    {
+      "epoch": 1.0497737556561086,
+      "grad_norm": 0.5821439614393185,
+      "learning_rate": 1.0876073409835762e-05,
+      "loss": 0.3373,
+      "step": 899
+    },
+    {
+      "epoch": 1.050941468398774,
+      "grad_norm": 0.6248458162949915,
+      "learning_rate": 1.08557500847884e-05,
+      "loss": 0.36,
+      "step": 900
+    },
+    {
+      "epoch": 1.0521091811414391,
+      "grad_norm": 0.5717444011085775,
+      "learning_rate": 1.0835423198469373e-05,
+      "loss": 0.3027,
+      "step": 901
+    },
+    {
+      "epoch": 1.0532768938841046,
+      "grad_norm": 0.6417179563546759,
+      "learning_rate": 1.0815092835470633e-05,
+      "loss": 0.3321,
+      "step": 902
+    },
+    {
+      "epoch": 1.0544446066267699,
+      "grad_norm": 0.598830455252831,
+      "learning_rate": 1.0794759080398587e-05,
+      "loss": 0.3448,
+      "step": 903
+    },
+    {
+      "epoch": 1.0556123193694351,
+      "grad_norm": 0.6151107774416679,
+      "learning_rate": 1.077442201787377e-05,
+      "loss": 0.3241,
+      "step": 904
+    },
+    {
+      "epoch": 1.0567800321121004,
+      "grad_norm": 0.5659879301818891,
+      "learning_rate": 1.0754081732530473e-05,
+      "loss": 0.3045,
+      "step": 905
+    },
+    {
+      "epoch": 1.0579477448547658,
+      "grad_norm": 0.5924130275973172,
+      "learning_rate": 1.0733738309016402e-05,
+      "loss": 0.3415,
+      "step": 906
+    },
+    {
+      "epoch": 1.059115457597431,
+      "grad_norm": 0.607018119373463,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.3258,
+      "step": 907
+    },
+    {
+      "epoch": 1.0602831703400963,
+      "grad_norm": 0.6143763832842917,
+      "learning_rate": 1.0693042386131713e-05,
+      "loss": 0.3396,
+      "step": 908
+    },
+    {
+      "epoch": 1.0614508830827616,
+      "grad_norm": 0.6156400050342027,
+      "learning_rate": 1.0672690056120398e-05,
+      "loss": 0.3415,
+      "step": 909
+    },
+    {
+      "epoch": 1.0626185958254268,
+      "grad_norm": 0.6098468012355265,
+      "learning_rate": 1.0652334926656209e-05,
+      "loss": 0.3292,
+      "step": 910
+    },
+    {
+      "epoch": 1.0637863085680923,
+      "grad_norm": 0.6146898035799918,
+      "learning_rate": 1.0631977082448626e-05,
+      "loss": 0.3332,
+      "step": 911
+    },
+    {
+      "epoch": 1.0649540213107576,
+      "grad_norm": 0.6283156318848911,
+      "learning_rate": 1.0611616608218429e-05,
+      "loss": 0.333,
+      "step": 912
+    },
+    {
+      "epoch": 1.0661217340534228,
+      "grad_norm": 0.6013428141389501,
+      "learning_rate": 1.059125358869734e-05,
+      "loss": 0.3338,
+      "step": 913
+    },
+    {
+      "epoch": 1.067289446796088,
+      "grad_norm": 0.5873718021343223,
+      "learning_rate": 1.0570888108627682e-05,
+      "loss": 0.3481,
+      "step": 914
+    },
+    {
+      "epoch": 1.0684571595387535,
+      "grad_norm": 0.7256666069680738,
+      "learning_rate": 1.0550520252762002e-05,
+      "loss": 0.3529,
+      "step": 915
+    },
+    {
+      "epoch": 1.0696248722814188,
+      "grad_norm": 0.6235951145839256,
+      "learning_rate": 1.0530150105862748e-05,
+      "loss": 0.3537,
+      "step": 916
+    },
+    {
+      "epoch": 1.070792585024084,
+      "grad_norm": 0.5612166777996952,
+      "learning_rate": 1.05097777527019e-05,
+      "loss": 0.3205,
+      "step": 917
+    },
+    {
+      "epoch": 1.0719602977667493,
+      "grad_norm": 0.6444947184171602,
+      "learning_rate": 1.0489403278060613e-05,
+      "loss": 0.348,
+      "step": 918
+    },
+    {
+      "epoch": 1.0731280105094148,
+      "grad_norm": 0.6344924315240748,
+      "learning_rate": 1.0469026766728876e-05,
+      "loss": 0.3239,
+      "step": 919
+    },
+    {
+      "epoch": 1.07429572325208,
+      "grad_norm": 0.6543395748815153,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.3458,
+      "step": 920
+    },
+    {
+      "epoch": 1.0754634359947453,
+      "grad_norm": 0.6075253298307579,
+      "learning_rate": 1.0428267973196027e-05,
+      "loss": 0.3453,
+      "step": 921
+    },
+    {
+      "epoch": 1.0766311487374105,
+      "grad_norm": 0.6374760075786919,
+      "learning_rate": 1.0407885860615859e-05,
+      "loss": 0.3317,
+      "step": 922
+    },
+    {
+      "epoch": 1.077798861480076,
+      "grad_norm": 0.6363762631216306,
+      "learning_rate": 1.0387502050586418e-05,
+      "loss": 0.3321,
+      "step": 923
+    },
+    {
+      "epoch": 1.0789665742227412,
+      "grad_norm": 0.5894293026751711,
+      "learning_rate": 1.0367116627936549e-05,
+      "loss": 0.3474,
+      "step": 924
+    },
+    {
+      "epoch": 1.0801342869654065,
+      "grad_norm": 0.6608241954781574,
+      "learning_rate": 1.0346729677501793e-05,
+      "loss": 0.3405,
+      "step": 925
+    },
+    {
+      "epoch": 1.0813019997080717,
+      "grad_norm": 0.6598572966117401,
+      "learning_rate": 1.0326341284124061e-05,
+      "loss": 0.3103,
+      "step": 926
+    },
+    {
+      "epoch": 1.0824697124507372,
+      "grad_norm": 0.6102134904585165,
+      "learning_rate": 1.0305951532651262e-05,
+      "loss": 0.3501,
+      "step": 927
+    },
+    {
+      "epoch": 1.0836374251934024,
+      "grad_norm": 0.5977073065923448,
+      "learning_rate": 1.0285560507936962e-05,
+      "loss": 0.3254,
+      "step": 928
+    },
+    {
+      "epoch": 1.0848051379360677,
+      "grad_norm": 0.6795525935447657,
+      "learning_rate": 1.0265168294840025e-05,
+      "loss": 0.3571,
+      "step": 929
+    },
+    {
+      "epoch": 1.085972850678733,
+      "grad_norm": 0.61832934176594,
+      "learning_rate": 1.0244774978224254e-05,
+      "loss": 0.3234,
+      "step": 930
+    },
+    {
+      "epoch": 1.0871405634213984,
+      "grad_norm": 0.5509767683858018,
+      "learning_rate": 1.0224380642958052e-05,
+      "loss": 0.312,
+      "step": 931
+    },
+    {
+      "epoch": 1.0883082761640637,
+      "grad_norm": 0.5901625478132777,
+      "learning_rate": 1.0203985373914056e-05,
+      "loss": 0.3486,
+      "step": 932
+    },
+    {
+      "epoch": 1.089475988906729,
+      "grad_norm": 0.5982698708013823,
+      "learning_rate": 1.018358925596879e-05,
+      "loss": 0.3316,
+      "step": 933
+    },
+    {
+      "epoch": 1.0906437016493942,
+      "grad_norm": 0.6371610615938184,
+      "learning_rate": 1.016319237400232e-05,
+      "loss": 0.3289,
+      "step": 934
+    },
+    {
+      "epoch": 1.0918114143920596,
+      "grad_norm": 0.5705755499029778,
+      "learning_rate": 1.0142794812897874e-05,
+      "loss": 0.3504,
+      "step": 935
+    },
+    {
+      "epoch": 1.092979127134725,
+      "grad_norm": 0.61135403705729,
+      "learning_rate": 1.0122396657541522e-05,
+      "loss": 0.3486,
+      "step": 936
+    },
+    {
+      "epoch": 1.0941468398773901,
+      "grad_norm": 0.6525911087993839,
+      "learning_rate": 1.0101997992821798e-05,
+      "loss": 0.3395,
+      "step": 937
+    },
+    {
+      "epoch": 1.0953145526200554,
+      "grad_norm": 0.5963327931884257,
+      "learning_rate": 1.0081598903629359e-05,
+      "loss": 0.3338,
+      "step": 938
+    },
+    {
+      "epoch": 1.0964822653627209,
+      "grad_norm": 0.6077513530366386,
+      "learning_rate": 1.0061199474856628e-05,
+      "loss": 0.3384,
+      "step": 939
+    },
+    {
+      "epoch": 1.0976499781053861,
+      "grad_norm": 0.586673160312449,
+      "learning_rate": 1.0040799791397444e-05,
+      "loss": 0.3547,
+      "step": 940
+    },
+    {
+      "epoch": 1.0988176908480514,
+      "grad_norm": 0.5858089513355673,
+      "learning_rate": 1.0020399938146701e-05,
+      "loss": 0.3351,
+      "step": 941
+    },
+    {
+      "epoch": 1.0999854035907166,
+      "grad_norm": 0.5870950212238221,
+      "learning_rate": 1e-05,
+      "loss": 0.3228,
+      "step": 942
+    },
+    {
+      "epoch": 1.101153116333382,
+      "grad_norm": 0.5600594671812871,
+      "learning_rate": 9.979600061853302e-06,
+      "loss": 0.3358,
+      "step": 943
+    },
+    {
+      "epoch": 1.1023208290760473,
+      "grad_norm": 0.7124925371559063,
+      "learning_rate": 9.95920020860256e-06,
+      "loss": 0.3344,
+      "step": 944
+    },
+    {
+      "epoch": 1.1034885418187126,
+      "grad_norm": 0.6507189640732297,
+      "learning_rate": 9.938800525143374e-06,
+      "loss": 0.3368,
+      "step": 945
+    },
+    {
+      "epoch": 1.1046562545613778,
+      "grad_norm": 0.6324612111857144,
+      "learning_rate": 9.918401096370644e-06,
+      "loss": 0.3255,
+      "step": 946
+    },
+    {
+      "epoch": 1.1058239673040433,
+      "grad_norm": 0.5981877989726946,
+      "learning_rate": 9.898002007178204e-06,
+      "loss": 0.3461,
+      "step": 947
+    },
+    {
+      "epoch": 1.1069916800467086,
+      "grad_norm": 0.5950334569246378,
+      "learning_rate": 9.877603342458483e-06,
+      "loss": 0.3386,
+      "step": 948
+    },
+    {
+      "epoch": 1.1081593927893738,
+      "grad_norm": 0.6290079377280343,
+      "learning_rate": 9.85720518710213e-06,
+      "loss": 0.3325,
+      "step": 949
+    },
+    {
+      "epoch": 1.109327105532039,
+      "grad_norm": 0.6144124280687013,
+      "learning_rate": 9.836807625997683e-06,
+      "loss": 0.3452,
+      "step": 950
+    },
+    {
+      "epoch": 1.1104948182747045,
+      "grad_norm": 0.6028477884040625,
+      "learning_rate": 9.81641074403121e-06,
+      "loss": 0.3331,
+      "step": 951
+    },
+    {
+      "epoch": 1.1116625310173698,
+      "grad_norm": 0.5963226162574251,
+      "learning_rate": 9.79601462608595e-06,
+      "loss": 0.3302,
+      "step": 952
+    },
+    {
+      "epoch": 1.112830243760035,
+      "grad_norm": 0.660836450498669,
+      "learning_rate": 9.775619357041952e-06,
+      "loss": 0.3628,
+      "step": 953
+    },
+    {
+      "epoch": 1.1139979565027003,
+      "grad_norm": 0.5953020911139357,
+      "learning_rate": 9.75522502177575e-06,
+      "loss": 0.3174,
+      "step": 954
+    },
+    {
+      "epoch": 1.1151656692453655,
+      "grad_norm": 0.6444604126266348,
+      "learning_rate": 9.734831705159975e-06,
+      "loss": 0.3445,
+      "step": 955
+    },
+    {
+      "epoch": 1.116333381988031,
+      "grad_norm": 0.6307061608611486,
+      "learning_rate": 9.71443949206304e-06,
+      "loss": 0.3449,
+      "step": 956
+    },
+    {
+      "epoch": 1.1175010947306963,
+      "grad_norm": 0.6070113942247372,
+      "learning_rate": 9.69404846734874e-06,
+      "loss": 0.3208,
+      "step": 957
+    },
+    {
+      "epoch": 1.1186688074733615,
+      "grad_norm": 0.5849714162216884,
+      "learning_rate": 9.67365871587594e-06,
+      "loss": 0.3363,
+      "step": 958
+    },
+    {
+      "epoch": 1.1198365202160268,
+      "grad_norm": 0.578634894508863,
+      "learning_rate": 9.653270322498209e-06,
+      "loss": 0.3266,
+      "step": 959
+    },
+    {
+      "epoch": 1.1210042329586922,
+      "grad_norm": 0.6165941962485161,
+      "learning_rate": 9.632883372063458e-06,
+      "loss": 0.3464,
+      "step": 960
+    },
+    {
+      "epoch": 1.1221719457013575,
+      "grad_norm": 0.5866504448166627,
+      "learning_rate": 9.612497949413586e-06,
+      "loss": 0.3297,
+      "step": 961
+    },
+    {
+      "epoch": 1.1233396584440227,
+      "grad_norm": 0.5958484690927442,
+      "learning_rate": 9.592114139384146e-06,
+      "loss": 0.3305,
+      "step": 962
+    },
+    {
+      "epoch": 1.124507371186688,
+      "grad_norm": 0.5768128672731234,
+      "learning_rate": 9.571732026803978e-06,
+      "loss": 0.3242,
+      "step": 963
+    },
+    {
+      "epoch": 1.1256750839293534,
+      "grad_norm": 0.6066907669322806,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.3266,
+      "step": 964
+    },
+    {
+      "epoch": 1.1268427966720187,
+      "grad_norm": 0.5641717626760581,
+      "learning_rate": 9.530973233271127e-06,
+      "loss": 0.3554,
+      "step": 965
+    },
+    {
+      "epoch": 1.128010509414684,
+      "grad_norm": 0.5893972630202221,
+      "learning_rate": 9.51059672193939e-06,
+      "loss": 0.3168,
+      "step": 966
+    },
+    {
+      "epoch": 1.1291782221573492,
+      "grad_norm": 0.6142090607274365,
+      "learning_rate": 9.4902222472981e-06,
+      "loss": 0.3535,
+      "step": 967
+    },
+    {
+      "epoch": 1.1303459349000147,
+      "grad_norm": 0.5891123176262837,
+      "learning_rate": 9.469849894137254e-06,
+      "loss": 0.3324,
+      "step": 968
+    },
+    {
+      "epoch": 1.13151364764268,
+      "grad_norm": 0.5741054931137111,
+      "learning_rate": 9.449479747238001e-06,
+      "loss": 0.3345,
+      "step": 969
+    },
+    {
+      "epoch": 1.1326813603853452,
+      "grad_norm": 0.5995540158447479,
+      "learning_rate": 9.42911189137232e-06,
+      "loss": 0.3339,
+      "step": 970
+    },
+    {
+      "epoch": 1.1338490731280104,
+      "grad_norm": 0.5950089905462422,
+      "learning_rate": 9.408746411302661e-06,
+      "loss": 0.3374,
+      "step": 971
+    },
+    {
+      "epoch": 1.135016785870676,
+      "grad_norm": 0.5617901300179063,
+      "learning_rate": 9.388383391781576e-06,
+      "loss": 0.303,
+      "step": 972
+    },
+    {
+      "epoch": 1.1361844986133411,
+      "grad_norm": 0.6534592388381217,
+      "learning_rate": 9.368022917551377e-06,
+      "loss": 0.3505,
+      "step": 973
+    },
+    {
+      "epoch": 1.1373522113560064,
+      "grad_norm": 0.6077336951039518,
+      "learning_rate": 9.347665073343795e-06,
+      "loss": 0.3417,
+      "step": 974
+    },
+    {
+      "epoch": 1.1385199240986716,
+      "grad_norm": 0.6263419588967151,
+      "learning_rate": 9.327309943879604e-06,
+      "loss": 0.3439,
+      "step": 975
+    },
+    {
+      "epoch": 1.1396876368413371,
+      "grad_norm": 0.5902459193634356,
+      "learning_rate": 9.306957613868292e-06,
+      "loss": 0.3295,
+      "step": 976
+    },
+    {
+      "epoch": 1.1408553495840024,
+      "grad_norm": 0.6292663164493003,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.3434,
+      "step": 977
+    },
+    {
+      "epoch": 1.1420230623266676,
+      "grad_norm": 0.6164127486093949,
+      "learning_rate": 9.266261690983601e-06,
+      "loss": 0.3298,
+      "step": 978
+    },
+    {
+      "epoch": 1.1431907750693329,
+      "grad_norm": 0.6294228264975905,
+      "learning_rate": 9.245918267469528e-06,
+      "loss": 0.3366,
+      "step": 979
+    },
+    {
+      "epoch": 1.1443584878119983,
+      "grad_norm": 0.6121276459519236,
+      "learning_rate": 9.225577982126234e-06,
+      "loss": 0.3634,
+      "step": 980
+    },
+    {
+      "epoch": 1.1455262005546636,
+      "grad_norm": 0.6092795230748526,
+      "learning_rate": 9.205240919601417e-06,
+      "loss": 0.3243,
+      "step": 981
+    },
+    {
+      "epoch": 1.1466939132973288,
+      "grad_norm": 0.5972900901692023,
+      "learning_rate": 9.184907164529369e-06,
+      "loss": 0.3394,
+      "step": 982
+    },
+    {
+      "epoch": 1.147861626039994,
+      "grad_norm": 0.5775769128688568,
+      "learning_rate": 9.164576801530628e-06,
+      "loss": 0.3356,
+      "step": 983
+    },
+    {
+      "epoch": 1.1490293387826596,
+      "grad_norm": 0.5617155498389623,
+      "learning_rate": 9.144249915211605e-06,
+      "loss": 0.3433,
+      "step": 984
+    },
+    {
+      "epoch": 1.1501970515253248,
+      "grad_norm": 0.5912607665897911,
+      "learning_rate": 9.123926590164241e-06,
+      "loss": 0.329,
+      "step": 985
+    },
+    {
+      "epoch": 1.15136476426799,
+      "grad_norm": 0.5868787612558205,
+      "learning_rate": 9.103606910965666e-06,
+      "loss": 0.3442,
+      "step": 986
+    },
+    {
+      "epoch": 1.1525324770106553,
+      "grad_norm": 0.5662769305659202,
+      "learning_rate": 9.083290962177828e-06,
+      "loss": 0.3288,
+      "step": 987
+    },
+    {
+      "epoch": 1.1537001897533208,
+      "grad_norm": 0.5990820465792899,
+      "learning_rate": 9.06297882834716e-06,
+      "loss": 0.3246,
+      "step": 988
+    },
+    {
+      "epoch": 1.154867902495986,
+      "grad_norm": 0.5683989615589246,
+      "learning_rate": 9.04267059400421e-06,
+      "loss": 0.3372,
+      "step": 989
+    },
+    {
+      "epoch": 1.1560356152386513,
+      "grad_norm": 0.6055634053408364,
+      "learning_rate": 9.022366343663299e-06,
+      "loss": 0.3247,
+      "step": 990
+    },
+    {
+      "epoch": 1.1572033279813165,
+      "grad_norm": 0.5964207966349285,
+      "learning_rate": 9.002066161822174e-06,
+      "loss": 0.3262,
+      "step": 991
+    },
+    {
+      "epoch": 1.1583710407239818,
+      "grad_norm": 0.6629138350822035,
+      "learning_rate": 8.981770132961649e-06,
+      "loss": 0.3734,
+      "step": 992
+    },
+    {
+      "epoch": 1.1595387534666473,
+      "grad_norm": 0.636560323236049,
+      "learning_rate": 8.961478341545248e-06,
+      "loss": 0.3422,
+      "step": 993
+    },
+    {
+      "epoch": 1.1607064662093125,
+      "grad_norm": 0.5958217050485245,
+      "learning_rate": 8.94119087201887e-06,
+      "loss": 0.3243,
+      "step": 994
+    },
+    {
+      "epoch": 1.1618741789519778,
+      "grad_norm": 0.6312548465205183,
+      "learning_rate": 8.920907808810423e-06,
+      "loss": 0.323,
+      "step": 995
+    },
+    {
+      "epoch": 1.1630418916946432,
+      "grad_norm": 0.6170904077375025,
+      "learning_rate": 8.900629236329482e-06,
+      "loss": 0.3475,
+      "step": 996
+    },
+    {
+      "epoch": 1.1642096044373085,
+      "grad_norm": 0.6131219365161665,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.3309,
+      "step": 997
+    },
+    {
+      "epoch": 1.1653773171799737,
+      "grad_norm": 0.5940029008750473,
+      "learning_rate": 8.860085901094595e-06,
+      "loss": 0.3625,
+      "step": 998
+    },
+    {
+      "epoch": 1.166545029922639,
+      "grad_norm": 0.5943790566831245,
+      "learning_rate": 8.839821307064948e-06,
+      "loss": 0.3071,
+      "step": 999
+    },
+    {
+      "epoch": 1.1677127426653042,
+      "grad_norm": 0.5670468172152491,
+      "learning_rate": 8.819561541210698e-06,
+      "loss": 0.3086,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1688804554079697,
+      "grad_norm": 0.5908413065344994,
+      "learning_rate": 8.799306687844459e-06,
+      "loss": 0.3374,
+      "step": 1001
+    },
+    {
+      "epoch": 1.170048168150635,
+      "grad_norm": 0.6214200008853535,
+      "learning_rate": 8.779056831258402e-06,
+      "loss": 0.3291,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1712158808933002,
+      "grad_norm": 0.5926956997086258,
+      "learning_rate": 8.758812055723912e-06,
+      "loss": 0.3228,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1723835936359657,
+      "grad_norm": 0.5953115870613941,
+      "learning_rate": 8.738572445491225e-06,
+      "loss": 0.3185,
+      "step": 1004
+    },
+    {
+      "epoch": 1.173551306378631,
+      "grad_norm": 0.6266954865060391,
+      "learning_rate": 8.718338084789074e-06,
+      "loss": 0.3552,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1747190191212962,
+      "grad_norm": 0.5753742037339852,
+      "learning_rate": 8.698109057824354e-06,
+      "loss": 0.3239,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1758867318639614,
+      "grad_norm": 0.6235860357730405,
+      "learning_rate": 8.67788544878176e-06,
+      "loss": 0.3344,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1770544446066267,
+      "grad_norm": 0.6038406787340044,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.3401,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1782221573492921,
+      "grad_norm": 0.5624715586221942,
+      "learning_rate": 8.637454821088657e-06,
+      "loss": 0.3224,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1793898700919574,
+      "grad_norm": 0.5867024920584354,
+      "learning_rate": 8.6172479706934e-06,
+      "loss": 0.3377,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1805575828346226,
+      "grad_norm": 0.5973966825756316,
+      "learning_rate": 8.597046874730074e-06,
+      "loss": 0.3524,
+      "step": 1011
+    },
+    {
+      "epoch": 1.181725295577288,
+      "grad_norm": 0.5960907175884259,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.336,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1828930083199534,
+      "grad_norm": 0.6588409927078461,
+      "learning_rate": 8.556662282348782e-06,
+      "loss": 0.328,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1840607210626186,
+      "grad_norm": 0.6314110469314091,
+      "learning_rate": 8.53647895399448e-06,
+      "loss": 0.3305,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1852284338052839,
+      "grad_norm": 0.6514862396021054,
+      "learning_rate": 8.516301716198769e-06,
+      "loss": 0.3694,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1863961465479491,
+      "grad_norm": 0.6472640790696207,
+      "learning_rate": 8.496130652930818e-06,
+      "loss": 0.3388,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1875638592906146,
+      "grad_norm": 0.646981324532655,
+      "learning_rate": 8.4759658481341e-06,
+      "loss": 0.327,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1887315720332798,
+      "grad_norm": 0.6309625992156652,
+      "learning_rate": 8.455807385726047e-06,
+      "loss": 0.3257,
+      "step": 1018
+    },
+    {
+      "epoch": 1.189899284775945,
+      "grad_norm": 0.6257421903704458,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.3264,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1910669975186103,
+      "grad_norm": 0.6492303529522874,
+      "learning_rate": 8.415509823613332e-06,
+      "loss": 0.3388,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1922347102612758,
+      "grad_norm": 0.5618618555963276,
+      "learning_rate": 8.395370891610162e-06,
+      "loss": 0.3275,
+      "step": 1021
+    },
+    {
+      "epoch": 1.193402423003941,
+      "grad_norm": 0.6237450753682882,
+      "learning_rate": 8.375238637397942e-06,
+      "loss": 0.314,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1945701357466063,
+      "grad_norm": 0.6721716259703175,
+      "learning_rate": 8.355113144758642e-06,
+      "loss": 0.3239,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1957378484892716,
+      "grad_norm": 0.5985205799013982,
+      "learning_rate": 8.334994497446091e-06,
+      "loss": 0.3023,
+      "step": 1024
+    },
+    {
+      "epoch": 1.196905561231937,
+      "grad_norm": 0.6964777278833361,
+      "learning_rate": 8.31488277918563e-06,
+      "loss": 0.3348,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1980732739746023,
+      "grad_norm": 0.6320833354408427,
+      "learning_rate": 8.294778073673762e-06,
+      "loss": 0.3331,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1992409867172675,
+      "grad_norm": 0.615812910791454,
+      "learning_rate": 8.274680464577816e-06,
+      "loss": 0.3338,
+      "step": 1027
+    },
+    {
+      "epoch": 1.2004086994599328,
+      "grad_norm": 0.6106656822771644,
+      "learning_rate": 8.25459003553558e-06,
+      "loss": 0.3374,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2015764122025983,
+      "grad_norm": 0.5982107542048682,
+      "learning_rate": 8.234506870154956e-06,
+      "loss": 0.317,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2027441249452635,
+      "grad_norm": 0.62646050112161,
+      "learning_rate": 8.214431052013636e-06,
+      "loss": 0.3199,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2039118376879288,
+      "grad_norm": 0.5961714158873452,
+      "learning_rate": 8.194362664658715e-06,
+      "loss": 0.3151,
+      "step": 1031
+    },
+    {
+      "epoch": 1.205079550430594,
+      "grad_norm": 0.6417779105235183,
+      "learning_rate": 8.174301791606384e-06,
+      "loss": 0.3652,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2062472631732595,
+      "grad_norm": 0.596372288058658,
+      "learning_rate": 8.154248516341547e-06,
+      "loss": 0.3367,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2074149759159247,
+      "grad_norm": 0.5820849142379285,
+      "learning_rate": 8.134202922317496e-06,
+      "loss": 0.3297,
+      "step": 1034
+    },
+    {
+      "epoch": 1.20858268865859,
+      "grad_norm": 0.6340622610386579,
+      "learning_rate": 8.114165092955556e-06,
+      "loss": 0.3495,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2097504014012552,
+      "grad_norm": 0.5884596626971791,
+      "learning_rate": 8.094135111644741e-06,
+      "loss": 0.3398,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2109181141439205,
+      "grad_norm": 0.6437852474713643,
+      "learning_rate": 8.074113061741397e-06,
+      "loss": 0.3273,
+      "step": 1037
+    },
+    {
+      "epoch": 1.212085826886586,
+      "grad_norm": 0.5943772331855418,
+      "learning_rate": 8.054099026568874e-06,
+      "loss": 0.3163,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2132535396292512,
+      "grad_norm": 0.6107523463233316,
+      "learning_rate": 8.03409308941716e-06,
+      "loss": 0.3497,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2144212523719164,
+      "grad_norm": 0.6308505632693739,
+      "learning_rate": 8.014095333542548e-06,
+      "loss": 0.338,
+      "step": 1040
+    },
+    {
+      "epoch": 1.215588965114582,
+      "grad_norm": 0.5982654981839908,
+      "learning_rate": 7.994105842167274e-06,
+      "loss": 0.332,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2167566778572472,
+      "grad_norm": 0.5567215414019999,
+      "learning_rate": 7.974124698479192e-06,
+      "loss": 0.3182,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2179243905999124,
+      "grad_norm": 0.5798350034631515,
+      "learning_rate": 7.954151985631409e-06,
+      "loss": 0.3416,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2190921033425777,
+      "grad_norm": 0.6351782594820321,
+      "learning_rate": 7.934187786741956e-06,
+      "loss": 0.3505,
+      "step": 1044
+    },
+    {
+      "epoch": 1.220259816085243,
+      "grad_norm": 0.5727010104186758,
+      "learning_rate": 7.914232184893418e-06,
+      "loss": 0.36,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2214275288279084,
+      "grad_norm": 0.6162702060613414,
+      "learning_rate": 7.894285263132612e-06,
+      "loss": 0.3556,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2225952415705736,
+      "grad_norm": 0.6259763730803493,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.335,
+      "step": 1047
+    },
+    {
+      "epoch": 1.223762954313239,
+      "grad_norm": 0.5982093824773319,
+      "learning_rate": 7.854417791880508e-06,
+      "loss": 0.3331,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2249306670559044,
+      "grad_norm": 0.5757024155882702,
+      "learning_rate": 7.834497408300839e-06,
+      "loss": 0.3308,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2260983797985696,
+      "grad_norm": 0.558588709161812,
+      "learning_rate": 7.814586036631485e-06,
+      "loss": 0.2958,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2272660925412349,
+      "grad_norm": 0.655838271892713,
+      "learning_rate": 7.79468375973519e-06,
+      "loss": 0.3567,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2284338052839001,
+      "grad_norm": 0.5868680727282953,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.3344,
+      "step": 1052
+    },
+    {
+      "epoch": 1.2296015180265654,
+      "grad_norm": 0.5912263684716998,
+      "learning_rate": 7.75490682152319e-06,
+      "loss": 0.3225,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.6259535681555469,
+      "learning_rate": 7.735032325742355e-06,
+      "loss": 0.3512,
+      "step": 1054
+    },
+    {
+      "epoch": 1.231936943511896,
+      "grad_norm": 0.5932552652851689,
+      "learning_rate": 7.715167255803643e-06,
+      "loss": 0.3322,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2331046562545613,
+      "grad_norm": 0.6360858535511377,
+      "learning_rate": 7.695311694377116e-06,
+      "loss": 0.3284,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2342723689972266,
+      "grad_norm": 0.6146679787519743,
+      "learning_rate": 7.675465724093254e-06,
+      "loss": 0.3417,
+      "step": 1057
+    },
+    {
+      "epoch": 1.235440081739892,
+      "grad_norm": 0.6462905154423072,
+      "learning_rate": 7.655629427542636e-06,
+      "loss": 0.325,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2366077944825573,
+      "grad_norm": 0.5828667642505818,
+      "learning_rate": 7.635802887275582e-06,
+      "loss": 0.3457,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2377755072252226,
+      "grad_norm": 0.6041465418353412,
+      "learning_rate": 7.615986185801807e-06,
+      "loss": 0.3258,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2389432199678878,
+      "grad_norm": 0.6574120913464946,
+      "learning_rate": 7.596179405590076e-06,
+      "loss": 0.3291,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2401109327105533,
+      "grad_norm": 0.573589106790915,
+      "learning_rate": 7.576382629067877e-06,
+      "loss": 0.3232,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2412786454532185,
+      "grad_norm": 0.6124530590171058,
+      "learning_rate": 7.556595938621058e-06,
+      "loss": 0.3373,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2424463581958838,
+      "grad_norm": 0.5626325500427822,
+      "learning_rate": 7.536819416593504e-06,
+      "loss": 0.3041,
+      "step": 1064
+    },
+    {
+      "epoch": 1.243614070938549,
+      "grad_norm": 0.6062292830360265,
+      "learning_rate": 7.517053145286766e-06,
+      "loss": 0.3366,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2447817836812145,
+      "grad_norm": 0.6140622254990218,
+      "learning_rate": 7.497297206959746e-06,
+      "loss": 0.3443,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2459494964238798,
+      "grad_norm": 0.6054457922624256,
+      "learning_rate": 7.47755168382835e-06,
+      "loss": 0.3208,
+      "step": 1067
+    },
+    {
+      "epoch": 1.247117209166545,
+      "grad_norm": 0.5909393790467525,
+      "learning_rate": 7.4578166580651335e-06,
+      "loss": 0.3069,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2482849219092103,
+      "grad_norm": 0.6361243718739771,
+      "learning_rate": 7.438092211798959e-06,
+      "loss": 0.3228,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2494526346518757,
+      "grad_norm": 0.5874874683276726,
+      "learning_rate": 7.4183784271146796e-06,
+      "loss": 0.353,
+      "step": 1070
+    },
+    {
+      "epoch": 1.250620347394541,
+      "grad_norm": 0.5980587604177432,
+      "learning_rate": 7.398675386052767e-06,
+      "loss": 0.3455,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2517880601372062,
+      "grad_norm": 0.6057437067673217,
+      "learning_rate": 7.378983170608982e-06,
+      "loss": 0.3265,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2529557728798715,
+      "grad_norm": 0.6601481451418033,
+      "learning_rate": 7.359301862734034e-06,
+      "loss": 0.316,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2541234856225367,
+      "grad_norm": 0.6211197054527257,
+      "learning_rate": 7.33963154433325e-06,
+      "loss": 0.3366,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2552911983652022,
+      "grad_norm": 0.6080541502927,
+      "learning_rate": 7.319972297266215e-06,
+      "loss": 0.3368,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2564589111078674,
+      "grad_norm": 0.5968375050135117,
+      "learning_rate": 7.3003242033464314e-06,
+      "loss": 0.3325,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2576266238505327,
+      "grad_norm": 0.5731920071765817,
+      "learning_rate": 7.280687344341006e-06,
+      "loss": 0.3019,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2587943365931982,
+      "grad_norm": 0.584127905082153,
+      "learning_rate": 7.2610618019702776e-06,
+      "loss": 0.3474,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2599620493358634,
+      "grad_norm": 0.5892867769138471,
+      "learning_rate": 7.241447657907494e-06,
+      "loss": 0.3114,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2611297620785287,
+      "grad_norm": 0.5793348903051772,
+      "learning_rate": 7.221844993778464e-06,
+      "loss": 0.3206,
+      "step": 1080
+    },
+    {
+      "epoch": 1.262297474821194,
+      "grad_norm": 0.5771670535270963,
+      "learning_rate": 7.202253891161225e-06,
+      "loss": 0.3168,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2634651875638592,
+      "grad_norm": 0.6122228213055287,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3275,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2646329003065246,
+      "grad_norm": 0.6404075664280797,
+      "learning_rate": 7.1631066965333705e-06,
+      "loss": 0.3421,
+      "step": 1083
+    },
+    {
+      "epoch": 1.26580061304919,
+      "grad_norm": 0.609716537923668,
+      "learning_rate": 7.143550767436894e-06,
+      "loss": 0.3201,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2669683257918551,
+      "grad_norm": 0.611536894706131,
+      "learning_rate": 7.124006725679828e-06,
+      "loss": 0.3363,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2681360385345206,
+      "grad_norm": 0.6661356568772632,
+      "learning_rate": 7.1044746525962446e-06,
+      "loss": 0.3375,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2693037512771859,
+      "grad_norm": 0.6004055197924362,
+      "learning_rate": 7.084954629470417e-06,
+      "loss": 0.3334,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2704714640198511,
+      "grad_norm": 0.6218677085980978,
+      "learning_rate": 7.065446737536455e-06,
+      "loss": 0.3107,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2716391767625164,
+      "grad_norm": 0.6178993158521935,
+      "learning_rate": 7.045951057978001e-06,
+      "loss": 0.3248,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2728068895051816,
+      "grad_norm": 0.5822294319150657,
+      "learning_rate": 7.026467671927863e-06,
+      "loss": 0.3429,
+      "step": 1090
+    },
+    {
+      "epoch": 1.273974602247847,
+      "grad_norm": 0.6233345485726064,
+      "learning_rate": 7.0069966604676996e-06,
+      "loss": 0.3282,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2751423149905123,
+      "grad_norm": 0.658434366279278,
+      "learning_rate": 6.9875381046276605e-06,
+      "loss": 0.3087,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2763100277331776,
+      "grad_norm": 0.5813184935032588,
+      "learning_rate": 6.968092085386063e-06,
+      "loss": 0.3322,
+      "step": 1093
+    },
+    {
+      "epoch": 1.277477740475843,
+      "grad_norm": 0.6088502646484204,
+      "learning_rate": 6.948658683669056e-06,
+      "loss": 0.3313,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2786454532185083,
+      "grad_norm": 0.5993645340506873,
+      "learning_rate": 6.929237980350283e-06,
+      "loss": 0.3544,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2798131659611736,
+      "grad_norm": 0.6549874864192505,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3374,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2809808787038388,
+      "grad_norm": 0.6228200460862363,
+      "learning_rate": 6.8904349921374065e-06,
+      "loss": 0.3423,
+      "step": 1097
+    },
+    {
+      "epoch": 1.282148591446504,
+      "grad_norm": 0.5826660303993216,
+      "learning_rate": 6.871052868725012e-06,
+      "loss": 0.3316,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2833163041891695,
+      "grad_norm": 0.618892881602154,
+      "learning_rate": 6.851683766673587e-06,
+      "loss": 0.3226,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2844840169318348,
+      "grad_norm": 0.6108590895794173,
+      "learning_rate": 6.832327766589177e-06,
+      "loss": 0.3232,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2856517296745,
+      "grad_norm": 0.5940752921628405,
+      "learning_rate": 6.812984949023307e-06,
+      "loss": 0.3531,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2868194424171655,
+      "grad_norm": 0.5629042761644147,
+      "learning_rate": 6.793655394472645e-06,
+      "loss": 0.3136,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2879871551598308,
+      "grad_norm": 0.5684490892422774,
+      "learning_rate": 6.774339183378663e-06,
+      "loss": 0.3351,
+      "step": 1103
+    },
+    {
+      "epoch": 1.289154867902496,
+      "grad_norm": 0.6807253278530063,
+      "learning_rate": 6.755036396127297e-06,
+      "loss": 0.3412,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2903225806451613,
+      "grad_norm": 0.5868680173497199,
+      "learning_rate": 6.73574711304862e-06,
+      "loss": 0.3201,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2914902933878265,
+      "grad_norm": 0.5803187407353924,
+      "learning_rate": 6.716471414416519e-06,
+      "loss": 0.3473,
+      "step": 1106
+    },
+    {
+      "epoch": 1.292658006130492,
+      "grad_norm": 0.6270826115324022,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 0.3653,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2938257188731572,
+      "grad_norm": 0.6049412730737163,
+      "learning_rate": 6.6779610913045344e-06,
+      "loss": 0.3486,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2949934316158225,
+      "grad_norm": 0.5701044221700858,
+      "learning_rate": 6.6587266270884095e-06,
+      "loss": 0.3143,
+      "step": 1109
+    },
+    {
+      "epoch": 1.296161144358488,
+      "grad_norm": 0.537664690519685,
+      "learning_rate": 6.639506067845698e-06,
+      "loss": 0.3167,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2973288571011532,
+      "grad_norm": 0.6169007174499044,
+      "learning_rate": 6.620299493564283e-06,
+      "loss": 0.3294,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2984965698438184,
+      "grad_norm": 0.6058032789420769,
+      "learning_rate": 6.601106984173835e-06,
+      "loss": 0.3491,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2996642825864837,
+      "grad_norm": 0.5923530977116124,
+      "learning_rate": 6.581928619545501e-06,
+      "loss": 0.2939,
+      "step": 1113
+    },
+    {
+      "epoch": 1.300831995329149,
+      "grad_norm": 0.5919977740056183,
+      "learning_rate": 6.562764479491566e-06,
+      "loss": 0.3271,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3019997080718144,
+      "grad_norm": 0.5496457275630284,
+      "learning_rate": 6.543614643765116e-06,
+      "loss": 0.3402,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3031674208144797,
+      "grad_norm": 0.575208073385335,
+      "learning_rate": 6.524479192059699e-06,
+      "loss": 0.3454,
+      "step": 1116
+    },
+    {
+      "epoch": 1.304335133557145,
+      "grad_norm": 0.6228226415583826,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.3302,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3055028462998102,
+      "grad_norm": 0.5906624210895624,
+      "learning_rate": 6.486251759186573e-06,
+      "loss": 0.2974,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3066705590424754,
+      "grad_norm": 0.5862395646213521,
+      "learning_rate": 6.467159937105352e-06,
+      "loss": 0.3473,
+      "step": 1119
+    },
+    {
+      "epoch": 1.307838271785141,
+      "grad_norm": 0.5944999404633945,
+      "learning_rate": 6.4480828172174714e-06,
+      "loss": 0.3299,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3090059845278061,
+      "grad_norm": 0.6319278539936934,
+      "learning_rate": 6.429020478913885e-06,
+      "loss": 0.3136,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3101736972704714,
+      "grad_norm": 0.5986201346114853,
+      "learning_rate": 6.409973001524012e-06,
+      "loss": 0.3383,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3113414100131369,
+      "grad_norm": 0.6252682725159169,
+      "learning_rate": 6.390940464315448e-06,
+      "loss": 0.3338,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3125091227558021,
+      "grad_norm": 0.5776714546980365,
+      "learning_rate": 6.3719229464935915e-06,
+      "loss": 0.34,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3136768354984674,
+      "grad_norm": 0.6163419640340055,
+      "learning_rate": 6.352920527201349e-06,
+      "loss": 0.3276,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3148445482411326,
+      "grad_norm": 0.604631667272429,
+      "learning_rate": 6.333933285518797e-06,
+      "loss": 0.3305,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3160122609837979,
+      "grad_norm": 0.5288265145994206,
+      "learning_rate": 6.314961300462841e-06,
+      "loss": 0.3054,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3171799737264633,
+      "grad_norm": 0.5820364501735351,
+      "learning_rate": 6.29600465098689e-06,
+      "loss": 0.3209,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3183476864691286,
+      "grad_norm": 0.5890177570556451,
+      "learning_rate": 6.277063415980549e-06,
+      "loss": 0.3262,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3195153992117938,
+      "grad_norm": 0.5815716657772748,
+      "learning_rate": 6.258137674269262e-06,
+      "loss": 0.3144,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3206831119544593,
+      "grad_norm": 0.5944447907596753,
+      "learning_rate": 6.239227504614004e-06,
+      "loss": 0.322,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3218508246971246,
+      "grad_norm": 0.5849643955235955,
+      "learning_rate": 6.220332985710936e-06,
+      "loss": 0.3428,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3230185374397898,
+      "grad_norm": 0.5725346813660577,
+      "learning_rate": 6.201454196191095e-06,
+      "loss": 0.3204,
+      "step": 1133
+    },
+    {
+      "epoch": 1.324186250182455,
+      "grad_norm": 0.6036037932697815,
+      "learning_rate": 6.182591214620057e-06,
+      "loss": 0.3134,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3253539629251203,
+      "grad_norm": 0.6209697460039231,
+      "learning_rate": 6.1637441194976145e-06,
+      "loss": 0.2942,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3265216756677858,
+      "grad_norm": 0.5842586684382056,
+      "learning_rate": 6.144912989257441e-06,
+      "loss": 0.3399,
+      "step": 1136
+    },
+    {
+      "epoch": 1.327689388410451,
+      "grad_norm": 0.5955613196951506,
+      "learning_rate": 6.126097902266772e-06,
+      "loss": 0.3558,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3288571011531163,
+      "grad_norm": 0.5871599696794811,
+      "learning_rate": 6.107298936826086e-06,
+      "loss": 0.3208,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3300248138957818,
+      "grad_norm": 0.5970172661369455,
+      "learning_rate": 6.088516171168763e-06,
+      "loss": 0.3186,
+      "step": 1139
+    },
+    {
+      "epoch": 1.331192526638447,
+      "grad_norm": 0.6266830822988185,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.3232,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3323602393811123,
+      "grad_norm": 0.5866874673710601,
+      "learning_rate": 6.050999551800318e-06,
+      "loss": 0.3282,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3335279521237775,
+      "grad_norm": 0.5876280822789759,
+      "learning_rate": 6.032265854217574e-06,
+      "loss": 0.3392,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3346956648664428,
+      "grad_norm": 0.6068777014712615,
+      "learning_rate": 6.013548668674305e-06,
+      "loss": 0.3348,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3358633776091082,
+      "grad_norm": 0.6175872115768598,
+      "learning_rate": 5.994848073063552e-06,
+      "loss": 0.3102,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3370310903517735,
+      "grad_norm": 0.6442494168622542,
+      "learning_rate": 5.9761641452093225e-06,
+      "loss": 0.3367,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3381988030944387,
+      "grad_norm": 0.580675369792476,
+      "learning_rate": 5.957496962866263e-06,
+      "loss": 0.3136,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3393665158371042,
+      "grad_norm": 0.5550109096811938,
+      "learning_rate": 5.938846603719333e-06,
+      "loss": 0.3333,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3405342285797694,
+      "grad_norm": 0.6178657399433503,
+      "learning_rate": 5.9202131453834664e-06,
+      "loss": 0.3224,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3417019413224347,
+      "grad_norm": 0.6003105551389183,
+      "learning_rate": 5.901596665403284e-06,
+      "loss": 0.344,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3428696540651,
+      "grad_norm": 0.5798762414113412,
+      "learning_rate": 5.8829972412527325e-06,
+      "loss": 0.3371,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3440373668077652,
+      "grad_norm": 0.5882182110893537,
+      "learning_rate": 5.864414950334796e-06,
+      "loss": 0.3391,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3452050795504307,
+      "grad_norm": 0.630904205528619,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3297,
+      "step": 1152
+    },
+    {
+      "epoch": 1.346372792293096,
+      "grad_norm": 0.6302405777322921,
+      "learning_rate": 5.827302077451813e-06,
+      "loss": 0.339,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3475405050357612,
+      "grad_norm": 0.6208194488861715,
+      "learning_rate": 5.808771649934924e-06,
+      "loss": 0.3492,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3487082177784266,
+      "grad_norm": 0.6028391382030713,
+      "learning_rate": 5.790258664546318e-06,
+      "loss": 0.3407,
+      "step": 1155
+    },
+    {
+      "epoch": 1.349875930521092,
+      "grad_norm": 0.6061151632991786,
+      "learning_rate": 5.7717631983292375e-06,
+      "loss": 0.3671,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3510436432637571,
+      "grad_norm": 0.6071301188168144,
+      "learning_rate": 5.753285328254039e-06,
+      "loss": 0.3198,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3522113560064224,
+      "grad_norm": 0.5909503729335845,
+      "learning_rate": 5.73482513121783e-06,
+      "loss": 0.3076,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3533790687490876,
+      "grad_norm": 0.6598295465856969,
+      "learning_rate": 5.716382684044191e-06,
+      "loss": 0.3374,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3545467814917531,
+      "grad_norm": 0.6119963817252112,
+      "learning_rate": 5.6979580634828125e-06,
+      "loss": 0.3251,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3557144942344184,
+      "grad_norm": 0.6115172886160264,
+      "learning_rate": 5.679551346209218e-06,
+      "loss": 0.3227,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3568822069770836,
+      "grad_norm": 0.6088087817175172,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.3255,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3580499197197489,
+      "grad_norm": 0.5693137437725768,
+      "learning_rate": 5.642791927854599e-06,
+      "loss": 0.305,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3592176324624141,
+      "grad_norm": 0.6124884776695871,
+      "learning_rate": 5.6244393797507944e-06,
+      "loss": 0.3182,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3603853452050796,
+      "grad_norm": 0.6169799908697504,
+      "learning_rate": 5.60610504088859e-06,
+      "loss": 0.3266,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3615530579477448,
+      "grad_norm": 0.6059107165046778,
+      "learning_rate": 5.587788987567785e-06,
+      "loss": 0.3359,
+      "step": 1166
+    },
+    {
+      "epoch": 1.36272077069041,
+      "grad_norm": 0.6181334512916642,
+      "learning_rate": 5.569491296012095e-06,
+      "loss": 0.3615,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3638884834330756,
+      "grad_norm": 0.5977931044082839,
+      "learning_rate": 5.551212042368792e-06,
+      "loss": 0.3203,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3650561961757408,
+      "grad_norm": 0.6283901686222585,
+      "learning_rate": 5.532951302708447e-06,
+      "loss": 0.3388,
+      "step": 1169
+    },
+    {
+      "epoch": 1.366223908918406,
+      "grad_norm": 0.6118168466506736,
+      "learning_rate": 5.514709153024571e-06,
+      "loss": 0.3226,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3673916216610713,
+      "grad_norm": 0.6009569479370827,
+      "learning_rate": 5.496485669233315e-06,
+      "loss": 0.3435,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3685593344037366,
+      "grad_norm": 0.6197364504919728,
+      "learning_rate": 5.478280927173145e-06,
+      "loss": 0.3088,
+      "step": 1172
+    },
+    {
+      "epoch": 1.369727047146402,
+      "grad_norm": 0.6236985763738885,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.3396,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3708947598890673,
+      "grad_norm": 0.5681695931480297,
+      "learning_rate": 5.441927971209644e-06,
+      "loss": 0.3094,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3720624726317325,
+      "grad_norm": 0.5639744543273071,
+      "learning_rate": 5.423779908592022e-06,
+      "loss": 0.3407,
+      "step": 1175
+    },
+    {
+      "epoch": 1.373230185374398,
+      "grad_norm": 0.6204374144804307,
+      "learning_rate": 5.405650890276255e-06,
+      "loss": 0.3187,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3743978981170633,
+      "grad_norm": 0.5939583531444291,
+      "learning_rate": 5.387540991707697e-06,
+      "loss": 0.3117,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3755656108597285,
+      "grad_norm": 0.5993870289183485,
+      "learning_rate": 5.369450288252113e-06,
+      "loss": 0.3301,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3767333236023938,
+      "grad_norm": 0.5532300776504113,
+      "learning_rate": 5.351378855195405e-06,
+      "loss": 0.3156,
+      "step": 1179
+    },
+    {
+      "epoch": 1.377901036345059,
+      "grad_norm": 0.645750905843748,
+      "learning_rate": 5.333326767743263e-06,
+      "loss": 0.3675,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3790687490877245,
+      "grad_norm": 0.5593133283732139,
+      "learning_rate": 5.315294101020882e-06,
+      "loss": 0.3289,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3802364618303897,
+      "grad_norm": 0.6114178522390488,
+      "learning_rate": 5.297280930072632e-06,
+      "loss": 0.32,
+      "step": 1182
+    },
+    {
+      "epoch": 1.381404174573055,
+      "grad_norm": 0.5807979719000798,
+      "learning_rate": 5.279287329861749e-06,
+      "loss": 0.32,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3825718873157204,
+      "grad_norm": 0.5524368438400991,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.3172,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3837396000583857,
+      "grad_norm": 0.5846052309639777,
+      "learning_rate": 5.243359141097469e-06,
+      "loss": 0.305,
+      "step": 1185
+    },
+    {
+      "epoch": 1.384907312801051,
+      "grad_norm": 0.6273474631706062,
+      "learning_rate": 5.225424702062082e-06,
+      "loss": 0.3353,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3860750255437162,
+      "grad_norm": 0.5546408313824248,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.32,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3872427382863814,
+      "grad_norm": 0.6141795649929311,
+      "learning_rate": 5.1896155078624225e-06,
+      "loss": 0.3234,
+      "step": 1188
+    },
+    {
+      "epoch": 1.388410451029047,
+      "grad_norm": 0.6351542981244239,
+      "learning_rate": 5.171740901720943e-06,
+      "loss": 0.3761,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3895781637717122,
+      "grad_norm": 0.5588059057789255,
+      "learning_rate": 5.153886388761586e-06,
+      "loss": 0.3147,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3907458765143774,
+      "grad_norm": 0.5608855905257788,
+      "learning_rate": 5.136052043287324e-06,
+      "loss": 0.2957,
+      "step": 1191
+    },
+    {
+      "epoch": 1.391913589257043,
+      "grad_norm": 0.5914771936297953,
+      "learning_rate": 5.11823793951719e-06,
+      "loss": 0.3243,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3930813019997081,
+      "grad_norm": 0.5727054250890954,
+      "learning_rate": 5.100444151585982e-06,
+      "loss": 0.3234,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3942490147423734,
+      "grad_norm": 0.5673268524499282,
+      "learning_rate": 5.082670753543961e-06,
+      "loss": 0.3286,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3954167274850386,
+      "grad_norm": 0.5582299164641882,
+      "learning_rate": 5.064917819356532e-06,
+      "loss": 0.3198,
+      "step": 1195
+    },
+    {
+      "epoch": 1.396584440227704,
+      "grad_norm": 0.5375163994433056,
+      "learning_rate": 5.0471854229039286e-06,
+      "loss": 0.308,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3977521529703694,
+      "grad_norm": 0.6395484571450342,
+      "learning_rate": 5.029473637980926e-06,
+      "loss": 0.3228,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3989198657130346,
+      "grad_norm": 0.6089509636131114,
+      "learning_rate": 5.011782538296512e-06,
+      "loss": 0.3145,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4000875784556999,
+      "grad_norm": 0.6115267269810053,
+      "learning_rate": 4.994112197473605e-06,
+      "loss": 0.3195,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4012552911983653,
+      "grad_norm": 0.5737258077484382,
+      "learning_rate": 4.976462689048718e-06,
+      "loss": 0.3056,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4024230039410306,
+      "grad_norm": 0.5849232570345975,
+      "learning_rate": 4.958834086471683e-06,
+      "loss": 0.3137,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4035907166836958,
+      "grad_norm": 0.5904320055617874,
+      "learning_rate": 4.941226463105322e-06,
+      "loss": 0.2986,
+      "step": 1202
+    },
+    {
+      "epoch": 1.404758429426361,
+      "grad_norm": 0.5649539306563756,
+      "learning_rate": 4.923639892225147e-06,
+      "loss": 0.3246,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4059261421690263,
+      "grad_norm": 0.5722870647508127,
+      "learning_rate": 4.906074447019068e-06,
+      "loss": 0.3361,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4070938549116918,
+      "grad_norm": 0.5603498599746488,
+      "learning_rate": 4.888530200587075e-06,
+      "loss": 0.319,
+      "step": 1205
+    },
+    {
+      "epoch": 1.408261567654357,
+      "grad_norm": 0.5554428885813543,
+      "learning_rate": 4.87100722594094e-06,
+      "loss": 0.3291,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4094292803970223,
+      "grad_norm": 0.5429565085665191,
+      "learning_rate": 4.8535055960039046e-06,
+      "loss": 0.3171,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4105969931396876,
+      "grad_norm": 0.5815392789330235,
+      "learning_rate": 4.836025383610382e-06,
+      "loss": 0.322,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.5730375220086111,
+      "learning_rate": 4.818566661505663e-06,
+      "loss": 0.2993,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4129324186250183,
+      "grad_norm": 0.6055992924658231,
+      "learning_rate": 4.801129502345605e-06,
+      "loss": 0.3472,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4141001313676835,
+      "grad_norm": 0.6212653332419882,
+      "learning_rate": 4.783713978696315e-06,
+      "loss": 0.3334,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4152678441103488,
+      "grad_norm": 0.6120767106179542,
+      "learning_rate": 4.766320163033882e-06,
+      "loss": 0.3292,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4164355568530143,
+      "grad_norm": 0.5727386715863528,
+      "learning_rate": 4.748948127744037e-06,
+      "loss": 0.313,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4176032695956795,
+      "grad_norm": 0.5708924499550864,
+      "learning_rate": 4.7315979451218866e-06,
+      "loss": 0.3196,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4187709823383448,
+      "grad_norm": 0.5744418673884892,
+      "learning_rate": 4.714269687371581e-06,
+      "loss": 0.3422,
+      "step": 1215
+    },
+    {
+      "epoch": 1.41993869508101,
+      "grad_norm": 0.5790859719097946,
+      "learning_rate": 4.696963426606041e-06,
+      "loss": 0.2982,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4211064078236753,
+      "grad_norm": 0.5798497823852907,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.3126,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4222741205663407,
+      "grad_norm": 0.6342074623156888,
+      "learning_rate": 4.6624171840229005e-06,
+      "loss": 0.335,
+      "step": 1218
+    },
+    {
+      "epoch": 1.423441833309006,
+      "grad_norm": 0.5651357607094127,
+      "learning_rate": 4.645177345972218e-06,
+      "loss": 0.3189,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4246095460516712,
+      "grad_norm": 0.5706700280233543,
+      "learning_rate": 4.6279597924395434e-06,
+      "loss": 0.3073,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4257772587943367,
+      "grad_norm": 0.5886598573356556,
+      "learning_rate": 4.61076459507709e-06,
+      "loss": 0.3357,
+      "step": 1221
+    },
+    {
+      "epoch": 1.426944971537002,
+      "grad_norm": 0.6456500441452383,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.3756,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4281126842796672,
+      "grad_norm": 0.6806225478811585,
+      "learning_rate": 4.576441555006193e-06,
+      "loss": 0.3327,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4292803970223324,
+      "grad_norm": 0.6015730245901494,
+      "learning_rate": 4.559313855135795e-06,
+      "loss": 0.2985,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4304481097649977,
+      "grad_norm": 0.5813066188860718,
+      "learning_rate": 4.542208797111113e-06,
+      "loss": 0.295,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4316158225076632,
+      "grad_norm": 0.6257262084995834,
+      "learning_rate": 4.525126452116201e-06,
+      "loss": 0.3528,
+      "step": 1226
+    },
+    {
+      "epoch": 1.4327835352503284,
+      "grad_norm": 0.5633491296364481,
+      "learning_rate": 4.508066891240586e-06,
+      "loss": 0.2978,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4339512479929937,
+      "grad_norm": 0.6212375977512035,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.3296,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4351189607356591,
+      "grad_norm": 0.5859827851195061,
+      "learning_rate": 4.474016405730973e-06,
+      "loss": 0.3372,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4362866734783244,
+      "grad_norm": 0.5557858504928047,
+      "learning_rate": 4.457025622800771e-06,
+      "loss": 0.2922,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4374543862209896,
+      "grad_norm": 0.6140248839904036,
+      "learning_rate": 4.440057907396854e-06,
+      "loss": 0.3271,
+      "step": 1231
+    },
+    {
+      "epoch": 1.438622098963655,
+      "grad_norm": 0.596875750894977,
+      "learning_rate": 4.423113330131708e-06,
+      "loss": 0.3084,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4397898117063201,
+      "grad_norm": 0.6293436654920895,
+      "learning_rate": 4.406191961521533e-06,
+      "loss": 0.3101,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4409575244489856,
+      "grad_norm": 0.5758330163641427,
+      "learning_rate": 4.389293871985949e-06,
+      "loss": 0.3255,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4421252371916509,
+      "grad_norm": 0.6204033102699582,
+      "learning_rate": 4.372419131847684e-06,
+      "loss": 0.3329,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4432929499343161,
+      "grad_norm": 0.5852603479501702,
+      "learning_rate": 4.355567811332311e-06,
+      "loss": 0.3193,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4444606626769816,
+      "grad_norm": 0.6118048368824077,
+      "learning_rate": 4.338739980567926e-06,
+      "loss": 0.3376,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4456283754196468,
+      "grad_norm": 0.5791945601687869,
+      "learning_rate": 4.321935709584884e-06,
+      "loss": 0.3385,
+      "step": 1238
+    },
+    {
+      "epoch": 1.446796088162312,
+      "grad_norm": 0.5817207683262648,
+      "learning_rate": 4.305155068315481e-06,
+      "loss": 0.3071,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4479638009049773,
+      "grad_norm": 0.5656605926835451,
+      "learning_rate": 4.2883981265936884e-06,
+      "loss": 0.3137,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4491315136476426,
+      "grad_norm": 0.5983082681884239,
+      "learning_rate": 4.271664954154845e-06,
+      "loss": 0.3323,
+      "step": 1241
+    },
+    {
+      "epoch": 1.450299226390308,
+      "grad_norm": 0.6446215786897959,
+      "learning_rate": 4.254955620635371e-06,
+      "loss": 0.3455,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4514669391329733,
+      "grad_norm": 0.6128806595507833,
+      "learning_rate": 4.2382701955724724e-06,
+      "loss": 0.3244,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4526346518756386,
+      "grad_norm": 0.6202905317711209,
+      "learning_rate": 4.221608748403872e-06,
+      "loss": 0.3383,
+      "step": 1244
+    },
+    {
+      "epoch": 1.453802364618304,
+      "grad_norm": 0.6811412113321582,
+      "learning_rate": 4.204971348467498e-06,
+      "loss": 0.3266,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4549700773609693,
+      "grad_norm": 0.6464852814888112,
+      "learning_rate": 4.188358065001215e-06,
+      "loss": 0.3238,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4561377901036345,
+      "grad_norm": 0.6363092064143095,
+      "learning_rate": 4.1717689671425e-06,
+      "loss": 0.3268,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4573055028462998,
+      "grad_norm": 0.5922239464547726,
+      "learning_rate": 4.155204123928205e-06,
+      "loss": 0.3389,
+      "step": 1248
+    },
+    {
+      "epoch": 1.458473215588965,
+      "grad_norm": 0.584761179928322,
+      "learning_rate": 4.138663604294233e-06,
+      "loss": 0.3303,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4596409283316305,
+      "grad_norm": 0.6159373628673407,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.3074,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4608086410742958,
+      "grad_norm": 0.5843997761677447,
+      "learning_rate": 4.105655811004481e-06,
+      "loss": 0.3354,
+      "step": 1251
+    },
+    {
+      "epoch": 1.461976353816961,
+      "grad_norm": 0.5956145740177085,
+      "learning_rate": 4.0891886747132356e-06,
+      "loss": 0.3055,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4631440665596263,
+      "grad_norm": 0.6071611758648626,
+      "learning_rate": 4.072746136730828e-06,
+      "loss": 0.3287,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4643117793022915,
+      "grad_norm": 0.575198997015534,
+      "learning_rate": 4.056328265484184e-06,
+      "loss": 0.3248,
+      "step": 1254
+    },
+    {
+      "epoch": 1.465479492044957,
+      "grad_norm": 0.6096022216778926,
+      "learning_rate": 4.039935129297565e-06,
+      "loss": 0.3465,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4666472047876222,
+      "grad_norm": 0.5849672757796176,
+      "learning_rate": 4.023566796392313e-06,
+      "loss": 0.3501,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4678149175302875,
+      "grad_norm": 0.5995002632957724,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.3209,
+      "step": 1257
+    },
+    {
+      "epoch": 1.468982630272953,
+      "grad_norm": 0.6111758327025332,
+      "learning_rate": 3.990904812794835e-06,
+      "loss": 0.348,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4701503430156182,
+      "grad_norm": 0.5681653966266241,
+      "learning_rate": 3.974611298028037e-06,
+      "loss": 0.3218,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4713180557582834,
+      "grad_norm": 0.5860849112247218,
+      "learning_rate": 3.958342858392893e-06,
+      "loss": 0.309,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4724857685009487,
+      "grad_norm": 0.5526670179457326,
+      "learning_rate": 3.942099561591802e-06,
+      "loss": 0.3093,
+      "step": 1261
+    },
+    {
+      "epoch": 1.473653481243614,
+      "grad_norm": 0.5741280217532451,
+      "learning_rate": 3.925881475222528e-06,
+      "loss": 0.3221,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4748211939862794,
+      "grad_norm": 0.5510481085492939,
+      "learning_rate": 3.909688666777914e-06,
+      "loss": 0.3096,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4759889067289447,
+      "grad_norm": 0.5925696323642109,
+      "learning_rate": 3.893521203645618e-06,
+      "loss": 0.3022,
+      "step": 1264
+    },
+    {
+      "epoch": 1.47715661947161,
+      "grad_norm": 0.5695587957263079,
+      "learning_rate": 3.877379153107821e-06,
+      "loss": 0.3376,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4783243322142754,
+      "grad_norm": 0.5912283176701106,
+      "learning_rate": 3.8612625823409365e-06,
+      "loss": 0.324,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4794920449569406,
+      "grad_norm": 0.580557677298745,
+      "learning_rate": 3.845171558415348e-06,
+      "loss": 0.3168,
+      "step": 1267
+    },
+    {
+      "epoch": 1.480659757699606,
+      "grad_norm": 0.6092482737379171,
+      "learning_rate": 3.829106148295127e-06,
+      "loss": 0.3266,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4818274704422711,
+      "grad_norm": 0.5832787621408019,
+      "learning_rate": 3.8130664188377464e-06,
+      "loss": 0.3322,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4829951831849364,
+      "grad_norm": 0.5689662156544624,
+      "learning_rate": 3.7970524367938146e-06,
+      "loss": 0.3486,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4841628959276019,
+      "grad_norm": 0.5613609452024878,
+      "learning_rate": 3.78106426880678e-06,
+      "loss": 0.3325,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4853306086702671,
+      "grad_norm": 0.5983676952096137,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.3182,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4864983214129324,
+      "grad_norm": 0.5769910612964714,
+      "learning_rate": 3.749165641039796e-06,
+      "loss": 0.287,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4876660341555978,
+      "grad_norm": 0.5711560038132463,
+      "learning_rate": 3.733255314008516e-06,
+      "loss": 0.3075,
+      "step": 1274
+    },
+    {
+      "epoch": 1.488833746898263,
+      "grad_norm": 0.5400209752477543,
+      "learning_rate": 3.7173710665309017e-06,
+      "loss": 0.3088,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4900014596409283,
+      "grad_norm": 0.5678570040004282,
+      "learning_rate": 3.701512964710513e-06,
+      "loss": 0.3122,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4911691723835936,
+      "grad_norm": 0.581573775633068,
+      "learning_rate": 3.685681074542089e-06,
+      "loss": 0.3101,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4923368851262588,
+      "grad_norm": 0.6354338270611891,
+      "learning_rate": 3.6698754619112974e-06,
+      "loss": 0.3877,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4935045978689243,
+      "grad_norm": 0.5895314372596601,
+      "learning_rate": 3.654096192594442e-06,
+      "loss": 0.3121,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4946723106115896,
+      "grad_norm": 0.5851413952965162,
+      "learning_rate": 3.6383433322582028e-06,
+      "loss": 0.3085,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4958400233542548,
+      "grad_norm": 0.531933498790779,
+      "learning_rate": 3.62261694645935e-06,
+      "loss": 0.3385,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4970077360969203,
+      "grad_norm": 0.581332593203495,
+      "learning_rate": 3.6069171006444882e-06,
+      "loss": 0.3033,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4981754488395855,
+      "grad_norm": 0.5819803498755591,
+      "learning_rate": 3.591243860149759e-06,
+      "loss": 0.3152,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4993431615822508,
+      "grad_norm": 0.5924918056542192,
+      "learning_rate": 3.575597290200599e-06,
+      "loss": 0.3384,
+      "step": 1284
+    },
+    {
+      "epoch": 1.500510874324916,
+      "grad_norm": 0.6203361661490724,
+      "learning_rate": 3.5599774559114475e-06,
+      "loss": 0.3408,
+      "step": 1285
+    },
+    {
+      "epoch": 1.5016785870675813,
+      "grad_norm": 0.5680461307525456,
+      "learning_rate": 3.544384422285477e-06,
+      "loss": 0.3029,
+      "step": 1286
+    },
+    {
+      "epoch": 1.5028462998102468,
+      "grad_norm": 0.6033496137055872,
+      "learning_rate": 3.528818254214329e-06,
+      "loss": 0.3314,
+      "step": 1287
+    },
+    {
+      "epoch": 1.504014012552912,
+      "grad_norm": 0.5705440251254968,
+      "learning_rate": 3.513279016477844e-06,
+      "loss": 0.3244,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5051817252955773,
+      "grad_norm": 0.6445671691824884,
+      "learning_rate": 3.4977667737437905e-06,
+      "loss": 0.3337,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5063494380382427,
+      "grad_norm": 0.5975444724533477,
+      "learning_rate": 3.4822815905675955e-06,
+      "loss": 0.322,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5075171507809078,
+      "grad_norm": 0.5529879267272394,
+      "learning_rate": 3.4668235313920707e-06,
+      "loss": 0.2986,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5086848635235732,
+      "grad_norm": 0.5349198620149784,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.3103,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5098525762662385,
+      "grad_norm": 0.5994718546571468,
+      "learning_rate": 3.4359890422496245e-06,
+      "loss": 0.3271,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5110202890089037,
+      "grad_norm": 0.5379614556451655,
+      "learning_rate": 3.4206127406028744e-06,
+      "loss": 0.3038,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5121880017515692,
+      "grad_norm": 0.5909774397736689,
+      "learning_rate": 3.4052638195965924e-06,
+      "loss": 0.3271,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5133557144942344,
+      "grad_norm": 0.561549142937376,
+      "learning_rate": 3.3899423431065215e-06,
+      "loss": 0.2989,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5145234272368997,
+      "grad_norm": 0.5725997782063574,
+      "learning_rate": 3.374648374894204e-06,
+      "loss": 0.3257,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5156911399795652,
+      "grad_norm": 0.5997540083597338,
+      "learning_rate": 3.3593819786067005e-06,
+      "loss": 0.295,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5168588527222302,
+      "grad_norm": 0.5741947771388433,
+      "learning_rate": 3.344143217776319e-06,
+      "loss": 0.3215,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5180265654648957,
+      "grad_norm": 0.558829942015401,
+      "learning_rate": 3.3289321558203767e-06,
+      "loss": 0.3141,
+      "step": 1300
+    },
+    {
+      "epoch": 1.519194278207561,
+      "grad_norm": 0.6551273867262426,
+      "learning_rate": 3.3137488560409027e-06,
+      "loss": 0.3442,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5203619909502262,
+      "grad_norm": 0.6116018367592759,
+      "learning_rate": 3.2985933816244065e-06,
+      "loss": 0.3297,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5215297036928916,
+      "grad_norm": 0.6034216207270496,
+      "learning_rate": 3.283465795641588e-06,
+      "loss": 0.326,
+      "step": 1303
+    },
+    {
+      "epoch": 1.522697416435557,
+      "grad_norm": 0.5604621675600183,
+      "learning_rate": 3.268366161047096e-06,
+      "loss": 0.3138,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5238651291782221,
+      "grad_norm": 0.5859539215217422,
+      "learning_rate": 3.2532945406792573e-06,
+      "loss": 0.3129,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5250328419208876,
+      "grad_norm": 0.6109817420582964,
+      "learning_rate": 3.2382509972598087e-06,
+      "loss": 0.3272,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5262005546635526,
+      "grad_norm": 0.5559562069533613,
+      "learning_rate": 3.223235593393642e-06,
+      "loss": 0.3383,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5273682674062181,
+      "grad_norm": 0.5643710837959152,
+      "learning_rate": 3.2082483915685526e-06,
+      "loss": 0.306,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5285359801488834,
+      "grad_norm": 0.567945516214042,
+      "learning_rate": 3.1932894541549643e-06,
+      "loss": 0.3214,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5297036928915486,
+      "grad_norm": 0.6207749849205965,
+      "learning_rate": 3.1783588434056845e-06,
+      "loss": 0.3307,
+      "step": 1310
+    },
+    {
+      "epoch": 1.530871405634214,
+      "grad_norm": 0.6110296110766635,
+      "learning_rate": 3.163456621455625e-06,
+      "loss": 0.3052,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5320391183768793,
+      "grad_norm": 0.6054726619933559,
+      "learning_rate": 3.1485828503215588e-06,
+      "loss": 0.3299,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5332068311195446,
+      "grad_norm": 0.6023805198217944,
+      "learning_rate": 3.133737591901864e-06,
+      "loss": 0.3287,
+      "step": 1313
+    },
+    {
+      "epoch": 1.53437454386221,
+      "grad_norm": 0.5751534519981539,
+      "learning_rate": 3.1189209079762606e-06,
+      "loss": 0.3181,
+      "step": 1314
+    },
+    {
+      "epoch": 1.535542256604875,
+      "grad_norm": 0.5758581211025763,
+      "learning_rate": 3.104132860205549e-06,
+      "loss": 0.3174,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5367099693475406,
+      "grad_norm": 0.5870153529077993,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.3174,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5378776820902058,
+      "grad_norm": 0.5749327377407726,
+      "learning_rate": 3.074642919175883e-06,
+      "loss": 0.3169,
+      "step": 1317
+    },
+    {
+      "epoch": 1.539045394832871,
+      "grad_norm": 0.5954987730777995,
+      "learning_rate": 3.0599411486416587e-06,
+      "loss": 0.3352,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5402131075755365,
+      "grad_norm": 0.6234015999203151,
+      "learning_rate": 3.045268259711256e-06,
+      "loss": 0.3352,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5413808203182016,
+      "grad_norm": 0.6101118336720753,
+      "learning_rate": 3.0306243134470668e-06,
+      "loss": 0.3397,
+      "step": 1320
+    },
+    {
+      "epoch": 1.542548533060867,
+      "grad_norm": 0.6250909118985923,
+      "learning_rate": 3.0160093707910275e-06,
+      "loss": 0.3276,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5437162458035323,
+      "grad_norm": 0.5995975487599562,
+      "learning_rate": 3.001423492564384e-06,
+      "loss": 0.3089,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5448839585461975,
+      "grad_norm": 0.6248195813420062,
+      "learning_rate": 2.9868667394674143e-06,
+      "loss": 0.3238,
+      "step": 1323
+    },
+    {
+      "epoch": 1.546051671288863,
+      "grad_norm": 0.6036995165883402,
+      "learning_rate": 2.972339172079204e-06,
+      "loss": 0.3557,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5472193840315283,
+      "grad_norm": 0.5800880820377491,
+      "learning_rate": 2.957840850857373e-06,
+      "loss": 0.3251,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5483870967741935,
+      "grad_norm": 0.6261273983898644,
+      "learning_rate": 2.9433718361378326e-06,
+      "loss": 0.323,
+      "step": 1326
+    },
+    {
+      "epoch": 1.549554809516859,
+      "grad_norm": 0.5876798216581178,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.3223,
+      "step": 1327
+    },
+    {
+      "epoch": 1.550722522259524,
+      "grad_norm": 0.5688458626774547,
+      "learning_rate": 2.9145219669391944e-06,
+      "loss": 0.2938,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5518902350021895,
+      "grad_norm": 0.5702756991656618,
+      "learning_rate": 2.9001412325211163e-06,
+      "loss": 0.3135,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5530579477448547,
+      "grad_norm": 0.5833549781595837,
+      "learning_rate": 2.885790044726853e-06,
+      "loss": 0.3342,
+      "step": 1330
+    },
+    {
+      "epoch": 1.55422566048752,
+      "grad_norm": 0.5752582282593391,
+      "learning_rate": 2.871468463280003e-06,
+      "loss": 0.3376,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5553933732301854,
+      "grad_norm": 0.6108623027194546,
+      "learning_rate": 2.8571765477809645e-06,
+      "loss": 0.3054,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5565610859728507,
+      "grad_norm": 0.5552043254592485,
+      "learning_rate": 2.8429143577066764e-06,
+      "loss": 0.3165,
+      "step": 1333
+    },
+    {
+      "epoch": 1.557728798715516,
+      "grad_norm": 0.6279203327629294,
+      "learning_rate": 2.8286819524103657e-06,
+      "loss": 0.3368,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5588965114581814,
+      "grad_norm": 0.6113433042275115,
+      "learning_rate": 2.8144793911213174e-06,
+      "loss": 0.3271,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5600642242008464,
+      "grad_norm": 0.5903306836412779,
+      "learning_rate": 2.80030673294461e-06,
+      "loss": 0.3118,
+      "step": 1336
+    },
+    {
+      "epoch": 1.561231936943512,
+      "grad_norm": 0.6032969121715787,
+      "learning_rate": 2.7861640368608846e-06,
+      "loss": 0.3264,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5623996496861772,
+      "grad_norm": 0.5895406665873117,
+      "learning_rate": 2.7720513617260857e-06,
+      "loss": 0.3361,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5635673624288424,
+      "grad_norm": 0.6092491881349318,
+      "learning_rate": 2.7579687662712285e-06,
+      "loss": 0.3258,
+      "step": 1339
+    },
+    {
+      "epoch": 1.564735075171508,
+      "grad_norm": 0.5669173423231283,
+      "learning_rate": 2.7439163091021525e-06,
+      "loss": 0.3114,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5659027879141731,
+      "grad_norm": 0.6112467040299236,
+      "learning_rate": 2.7298940486992654e-06,
+      "loss": 0.3298,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5670705006568384,
+      "grad_norm": 0.5665018675863405,
+      "learning_rate": 2.71590204341731e-06,
+      "loss": 0.3222,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5682382133995039,
+      "grad_norm": 0.5564855028044847,
+      "learning_rate": 2.7019403514851274e-06,
+      "loss": 0.3261,
+      "step": 1343
+    },
+    {
+      "epoch": 1.569405926142169,
+      "grad_norm": 0.6300692460330899,
+      "learning_rate": 2.688009031005403e-06,
+      "loss": 0.3306,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5705736388848344,
+      "grad_norm": 0.5563914801408777,
+      "learning_rate": 2.674108139954429e-06,
+      "loss": 0.291,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5717413516274996,
+      "grad_norm": 0.562697069851264,
+      "learning_rate": 2.6602377361818576e-06,
+      "loss": 0.339,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5729090643701649,
+      "grad_norm": 0.6096119386501435,
+      "learning_rate": 2.646397877410476e-06,
+      "loss": 0.3434,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5740767771128303,
+      "grad_norm": 0.5720081463045557,
+      "learning_rate": 2.6325886212359496e-06,
+      "loss": 0.3369,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5752444898554956,
+      "grad_norm": 0.6094521700589048,
+      "learning_rate": 2.6188100251265947e-06,
+      "loss": 0.3404,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5764122025981608,
+      "grad_norm": 0.5654146314699804,
+      "learning_rate": 2.6050621464231243e-06,
+      "loss": 0.3115,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5775799153408263,
+      "grad_norm": 0.5601293115326924,
+      "learning_rate": 2.591345042338421e-06,
+      "loss": 0.3034,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5787476280834913,
+      "grad_norm": 0.5909441124918923,
+      "learning_rate": 2.5776587699573007e-06,
+      "loss": 0.318,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5799153408261568,
+      "grad_norm": 0.5534500130895792,
+      "learning_rate": 2.564003386236271e-06,
+      "loss": 0.3117,
+      "step": 1353
+    },
+    {
+      "epoch": 1.581083053568822,
+      "grad_norm": 0.5778377805541672,
+      "learning_rate": 2.550378948003287e-06,
+      "loss": 0.3293,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5822507663114873,
+      "grad_norm": 0.5762502187870111,
+      "learning_rate": 2.5367855119575314e-06,
+      "loss": 0.333,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5834184790541528,
+      "grad_norm": 0.6049219578174849,
+      "learning_rate": 2.523223134669157e-06,
+      "loss": 0.3386,
+      "step": 1356
+    },
+    {
+      "epoch": 1.584586191796818,
+      "grad_norm": 0.600600279698782,
+      "learning_rate": 2.5096918725790754e-06,
+      "loss": 0.3166,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5857539045394833,
+      "grad_norm": 0.6303320123657141,
+      "learning_rate": 2.496191781998698e-06,
+      "loss": 0.3318,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5869216172821488,
+      "grad_norm": 0.5775713192727807,
+      "learning_rate": 2.4827229191097236e-06,
+      "loss": 0.3253,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5880893300248138,
+      "grad_norm": 0.6317194811227936,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.3187,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5892570427674793,
+      "grad_norm": 0.5236742248627521,
+      "learning_rate": 2.4558791004827487e-06,
+      "loss": 0.2989,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5904247555101445,
+      "grad_norm": 0.5950610963835087,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.3475,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5915924682528098,
+      "grad_norm": 0.5997137555567007,
+      "learning_rate": 2.4291608635483733e-06,
+      "loss": 0.3262,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5927601809954752,
+      "grad_norm": 0.5627342130445236,
+      "learning_rate": 2.4158489772852035e-06,
+      "loss": 0.3116,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5939278937381403,
+      "grad_norm": 0.5800984281523751,
+      "learning_rate": 2.4025686530663716e-06,
+      "loss": 0.3187,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5950956064808057,
+      "grad_norm": 0.5477993821943232,
+      "learning_rate": 2.3893199461589943e-06,
+      "loss": 0.3035,
+      "step": 1366
+    },
+    {
+      "epoch": 1.596263319223471,
+      "grad_norm": 0.6474731444350481,
+      "learning_rate": 2.376102911698618e-06,
+      "loss": 0.3454,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5974310319661362,
+      "grad_norm": 0.6131620307492133,
+      "learning_rate": 2.3629176046889755e-06,
+      "loss": 0.3165,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5985987447088017,
+      "grad_norm": 0.5756228754074724,
+      "learning_rate": 2.3497640800017687e-06,
+      "loss": 0.3207,
+      "step": 1369
+    },
+    {
+      "epoch": 1.599766457451467,
+      "grad_norm": 0.5945548860109273,
+      "learning_rate": 2.3366423923764268e-06,
+      "loss": 0.3236,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6009341701941322,
+      "grad_norm": 0.6426139847655684,
+      "learning_rate": 2.323552596419889e-06,
+      "loss": 0.3457,
+      "step": 1371
+    },
+    {
+      "epoch": 1.6021018829367977,
+      "grad_norm": 0.5651118017197341,
+      "learning_rate": 2.3104947466063785e-06,
+      "loss": 0.3161,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6032695956794627,
+      "grad_norm": 0.5927901098052901,
+      "learning_rate": 2.2974688972771743e-06,
+      "loss": 0.3213,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6044373084221282,
+      "grad_norm": 0.5663522444906705,
+      "learning_rate": 2.284475102640371e-06,
+      "loss": 0.3164,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6056050211647934,
+      "grad_norm": 0.5492405480015172,
+      "learning_rate": 2.27151341677068e-06,
+      "loss": 0.3205,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6067727339074587,
+      "grad_norm": 0.5575194855720007,
+      "learning_rate": 2.2585838936091753e-06,
+      "loss": 0.2967,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6079404466501241,
+      "grad_norm": 0.5910752592705985,
+      "learning_rate": 2.2456865869630974e-06,
+      "loss": 0.3167,
+      "step": 1377
+    },
+    {
+      "epoch": 1.6091081593927894,
+      "grad_norm": 0.6000589069325849,
+      "learning_rate": 2.2328215505056004e-06,
+      "loss": 0.2968,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6102758721354546,
+      "grad_norm": 0.5766325868645013,
+      "learning_rate": 2.2199888377755573e-06,
+      "loss": 0.3141,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6114435848781201,
+      "grad_norm": 0.5858978695231664,
+      "learning_rate": 2.207188502177313e-06,
+      "loss": 0.3142,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6126112976207851,
+      "grad_norm": 0.560276655384442,
+      "learning_rate": 2.194420596980481e-06,
+      "loss": 0.3051,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6137790103634506,
+      "grad_norm": 0.6204093285772938,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.3107,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6149467231061159,
+      "grad_norm": 0.574902812077174,
+      "learning_rate": 2.1689822901944456e-06,
+      "loss": 0.3099,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6161144358487811,
+      "grad_norm": 0.5832866464667151,
+      "learning_rate": 2.156311994468774e-06,
+      "loss": 0.2978,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6172821485914466,
+      "grad_norm": 0.5620181017884901,
+      "learning_rate": 2.1436743408711234e-06,
+      "loss": 0.3368,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6184498613341118,
+      "grad_norm": 0.5479010748144405,
+      "learning_rate": 2.1310693819940842e-06,
+      "loss": 0.3221,
+      "step": 1386
+    },
+    {
+      "epoch": 1.619617574076777,
+      "grad_norm": 0.5800419759080474,
+      "learning_rate": 2.1184971702941947e-06,
+      "loss": 0.3199,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6207852868194426,
+      "grad_norm": 0.5814472198609866,
+      "learning_rate": 2.1059577580917067e-06,
+      "loss": 0.3072,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6219529995621076,
+      "grad_norm": 0.5864635240505053,
+      "learning_rate": 2.0934511975703795e-06,
+      "loss": 0.3136,
+      "step": 1389
+    },
+    {
+      "epoch": 1.623120712304773,
+      "grad_norm": 0.5920195610303148,
+      "learning_rate": 2.0809775407772505e-06,
+      "loss": 0.3122,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6242884250474383,
+      "grad_norm": 0.5262022870991767,
+      "learning_rate": 2.068536839622427e-06,
+      "loss": 0.2998,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6254561377901036,
+      "grad_norm": 0.5835377050888029,
+      "learning_rate": 2.0561291458788736e-06,
+      "loss": 0.3484,
+      "step": 1392
+    },
+    {
+      "epoch": 1.626623850532769,
+      "grad_norm": 0.5701666985410108,
+      "learning_rate": 2.043754511182191e-06,
+      "loss": 0.3087,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6277915632754343,
+      "grad_norm": 0.6064248075808701,
+      "learning_rate": 2.031412987030398e-06,
+      "loss": 0.3225,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6289592760180995,
+      "grad_norm": 0.5772031808064267,
+      "learning_rate": 2.01910462478372e-06,
+      "loss": 0.3162,
+      "step": 1395
+    },
+    {
+      "epoch": 1.630126988760765,
+      "grad_norm": 0.5872215150464203,
+      "learning_rate": 2.0068294756643846e-06,
+      "loss": 0.3125,
+      "step": 1396
+    },
+    {
+      "epoch": 1.63129470150343,
+      "grad_norm": 0.6176859382906766,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.3268,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6324624142460955,
+      "grad_norm": 0.596475261039043,
+      "learning_rate": 1.982379021005325e-06,
+      "loss": 0.3355,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6336301269887608,
+      "grad_norm": 0.5825645348920312,
+      "learning_rate": 1.9702038172181016e-06,
+      "loss": 0.3094,
+      "step": 1399
+    },
+    {
+      "epoch": 1.634797839731426,
+      "grad_norm": 0.5593629484205526,
+      "learning_rate": 1.958062030062795e-06,
+      "loss": 0.3168,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6359655524740915,
+      "grad_norm": 0.6317795005626535,
+      "learning_rate": 1.94595371006842e-06,
+      "loss": 0.3185,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6371332652167567,
+      "grad_norm": 0.5900065811391266,
+      "learning_rate": 1.9338789076247e-06,
+      "loss": 0.3189,
+      "step": 1402
+    },
+    {
+      "epoch": 1.638300977959422,
+      "grad_norm": 0.6108489047567803,
+      "learning_rate": 1.9218376729818842e-06,
+      "loss": 0.3279,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6394686907020875,
+      "grad_norm": 0.5859803871102922,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.331,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6406364034447525,
+      "grad_norm": 0.5706918085336459,
+      "learning_rate": 1.8978561074012702e-06,
+      "loss": 0.322,
+      "step": 1405
+    },
+    {
+      "epoch": 1.641804116187418,
+      "grad_norm": 0.630085710994678,
+      "learning_rate": 1.8859158762646468e-06,
+      "loss": 0.3195,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6429718289300832,
+      "grad_norm": 0.6306568789161201,
+      "learning_rate": 1.8740094125308773e-06,
+      "loss": 0.3414,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6441395416727485,
+      "grad_norm": 0.5547557063378834,
+      "learning_rate": 1.8621367657496504e-06,
+      "loss": 0.3261,
+      "step": 1408
+    },
+    {
+      "epoch": 1.645307254415414,
+      "grad_norm": 0.552573271660822,
+      "learning_rate": 1.850297985329932e-06,
+      "loss": 0.3066,
+      "step": 1409
+    },
+    {
+      "epoch": 1.646474967158079,
+      "grad_norm": 0.602033127962172,
+      "learning_rate": 1.8384931205397305e-06,
+      "loss": 0.3376,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6476426799007444,
+      "grad_norm": 0.5923204837082142,
+      "learning_rate": 1.826722220505931e-06,
+      "loss": 0.3219,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6488103926434097,
+      "grad_norm": 0.5699031183092619,
+      "learning_rate": 1.8149853342140644e-06,
+      "loss": 0.3401,
+      "step": 1412
+    },
+    {
+      "epoch": 1.649978105386075,
+      "grad_norm": 0.6059467494327366,
+      "learning_rate": 1.8032825105081141e-06,
+      "loss": 0.2994,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6511458181287404,
+      "grad_norm": 0.5501478121443165,
+      "learning_rate": 1.7916137980903047e-06,
+      "loss": 0.3084,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6523135308714056,
+      "grad_norm": 0.6022160940202319,
+      "learning_rate": 1.7799792455209019e-06,
+      "loss": 0.3176,
+      "step": 1415
+    },
+    {
+      "epoch": 1.653481243614071,
+      "grad_norm": 0.5720555818245034,
+      "learning_rate": 1.7683789012180196e-06,
+      "loss": 0.3139,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6546489563567364,
+      "grad_norm": 0.5842474366296986,
+      "learning_rate": 1.7568128134574114e-06,
+      "loss": 0.3254,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6558166690994014,
+      "grad_norm": 0.5750315325254499,
+      "learning_rate": 1.74528103037226e-06,
+      "loss": 0.3226,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6569843818420669,
+      "grad_norm": 0.5964663295452539,
+      "learning_rate": 1.7337835999529995e-06,
+      "loss": 0.312,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6581520945847321,
+      "grad_norm": 0.5858065349910327,
+      "learning_rate": 1.722320570047089e-06,
+      "loss": 0.3297,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6593198073273974,
+      "grad_norm": 0.6527947800112359,
+      "learning_rate": 1.7108919883588404e-06,
+      "loss": 0.3377,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6604875200700628,
+      "grad_norm": 0.5915285962414885,
+      "learning_rate": 1.6994979024491943e-06,
+      "loss": 0.3296,
+      "step": 1422
+    },
+    {
+      "epoch": 1.661655232812728,
+      "grad_norm": 0.5804019971010231,
+      "learning_rate": 1.6881383597355438e-06,
+      "loss": 0.3473,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6628229455553933,
+      "grad_norm": 0.6195804351364163,
+      "learning_rate": 1.6768134074915277e-06,
+      "loss": 0.3134,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6639906582980588,
+      "grad_norm": 0.5839206833403029,
+      "learning_rate": 1.6655230928468257e-06,
+      "loss": 0.3263,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6651583710407238,
+      "grad_norm": 0.5713779993099194,
+      "learning_rate": 1.6542674627869738e-06,
+      "loss": 0.3147,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6663260837833893,
+      "grad_norm": 0.610108850876121,
+      "learning_rate": 1.6430465641531701e-06,
+      "loss": 0.2961,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6674937965260546,
+      "grad_norm": 0.5947036576058492,
+      "learning_rate": 1.6318604436420738e-06,
+      "loss": 0.3199,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6686615092687198,
+      "grad_norm": 0.5650687650249682,
+      "learning_rate": 1.6207091478056059e-06,
+      "loss": 0.3031,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6698292220113853,
+      "grad_norm": 0.5553698450706024,
+      "learning_rate": 1.609592723050767e-06,
+      "loss": 0.286,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6709969347540505,
+      "grad_norm": 0.5651337528912361,
+      "learning_rate": 1.5985112156394368e-06,
+      "loss": 0.3241,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6721646474967158,
+      "grad_norm": 0.6244646635427129,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.3106,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6733323602393813,
+      "grad_norm": 0.5899729856878284,
+      "learning_rate": 1.5764531371680859e-06,
+      "loss": 0.2905,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6745000729820463,
+      "grad_norm": 0.5620844871062467,
+      "learning_rate": 1.5654766579045033e-06,
+      "loss": 0.3242,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6756677857247118,
+      "grad_norm": 0.6001532980233398,
+      "learning_rate": 1.5545352795769231e-06,
+      "loss": 0.3335,
+      "step": 1435
+    },
+    {
+      "epoch": 1.676835498467377,
+      "grad_norm": 0.6831688768535135,
+      "learning_rate": 1.5436290477187589e-06,
+      "loss": 0.3483,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6780032112100423,
+      "grad_norm": 0.6108595158389439,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.3353,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6791709239527077,
+      "grad_norm": 0.5946360042380636,
+      "learning_rate": 1.5219222048128125e-06,
+      "loss": 0.3366,
+      "step": 1438
+    },
+    {
+      "epoch": 1.680338636695373,
+      "grad_norm": 0.6199463567659799,
+      "learning_rate": 1.5111216840997745e-06,
+      "loss": 0.3067,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6815063494380382,
+      "grad_norm": 0.5640851868207025,
+      "learning_rate": 1.500356490525261e-06,
+      "loss": 0.3124,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6826740621807037,
+      "grad_norm": 0.5549759822781462,
+      "learning_rate": 1.4896266688894823e-06,
+      "loss": 0.2886,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6838417749233687,
+      "grad_norm": 0.6053433767610317,
+      "learning_rate": 1.478932263845435e-06,
+      "loss": 0.3056,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6850094876660342,
+      "grad_norm": 0.5742418315877563,
+      "learning_rate": 1.4682733198987353e-06,
+      "loss": 0.3248,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6861772004086995,
+      "grad_norm": 0.5764549523138892,
+      "learning_rate": 1.457649881407417e-06,
+      "loss": 0.3199,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6873449131513647,
+      "grad_norm": 0.5731105483502383,
+      "learning_rate": 1.4470619925817654e-06,
+      "loss": 0.3206,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6885126258940302,
+      "grad_norm": 0.564687801716726,
+      "learning_rate": 1.436509697484111e-06,
+      "loss": 0.3266,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6896803386366954,
+      "grad_norm": 0.6085667829920601,
+      "learning_rate": 1.425993040028667e-06,
+      "loss": 0.3378,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6908480513793607,
+      "grad_norm": 0.5500349861170349,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.3173,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6920157641220261,
+      "grad_norm": 0.5887371195537995,
+      "learning_rate": 1.4050668129595346e-06,
+      "loss": 0.3001,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6931834768646912,
+      "grad_norm": 0.578763069487608,
+      "learning_rate": 1.39465733043199e-06,
+      "loss": 0.3024,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6943511896073566,
+      "grad_norm": 0.5998988113414421,
+      "learning_rate": 1.3842836597185916e-06,
+      "loss": 0.3418,
+      "step": 1451
+    },
+    {
+      "epoch": 1.695518902350022,
+      "grad_norm": 0.8092393209943611,
+      "learning_rate": 1.373945843990192e-06,
+      "loss": 0.3206,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6966866150926871,
+      "grad_norm": 0.5867601944355602,
+      "learning_rate": 1.3636439262684299e-06,
+      "loss": 0.3154,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6978543278353526,
+      "grad_norm": 0.5766120684552886,
+      "learning_rate": 1.3533779494255483e-06,
+      "loss": 0.313,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6990220405780176,
+      "grad_norm": 0.5927525508550977,
+      "learning_rate": 1.343147956184221e-06,
+      "loss": 0.3373,
+      "step": 1455
+    },
+    {
+      "epoch": 1.7001897533206831,
+      "grad_norm": 0.5705292633232889,
+      "learning_rate": 1.332953989117377e-06,
+      "loss": 0.3092,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7013574660633484,
+      "grad_norm": 0.5701358672886555,
+      "learning_rate": 1.322796090648013e-06,
+      "loss": 0.3075,
+      "step": 1457
+    },
+    {
+      "epoch": 1.7025251788060136,
+      "grad_norm": 0.5693601691251622,
+      "learning_rate": 1.3126743030490307e-06,
+      "loss": 0.318,
+      "step": 1458
+    },
+    {
+      "epoch": 1.703692891548679,
+      "grad_norm": 0.6007584006641404,
+      "learning_rate": 1.3025886684430467e-06,
+      "loss": 0.3042,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7048606042913443,
+      "grad_norm": 0.5835117312351015,
+      "learning_rate": 1.2925392288022299e-06,
+      "loss": 0.3075,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7060283170340096,
+      "grad_norm": 0.5135182337659171,
+      "learning_rate": 1.2825260259481143e-06,
+      "loss": 0.2773,
+      "step": 1461
+    },
+    {
+      "epoch": 1.707196029776675,
+      "grad_norm": 0.5605833248282587,
+      "learning_rate": 1.272549101551438e-06,
+      "loss": 0.3324,
+      "step": 1462
+    },
+    {
+      "epoch": 1.70836374251934,
+      "grad_norm": 0.5831006937045596,
+      "learning_rate": 1.2626084971319641e-06,
+      "loss": 0.3048,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7095314552620056,
+      "grad_norm": 0.6090873824177094,
+      "learning_rate": 1.2527042540583e-06,
+      "loss": 0.3604,
+      "step": 1464
+    },
+    {
+      "epoch": 1.7106991680046708,
+      "grad_norm": 0.6345039325258167,
+      "learning_rate": 1.2428364135477356e-06,
+      "loss": 0.3273,
+      "step": 1465
+    },
+    {
+      "epoch": 1.711866880747336,
+      "grad_norm": 0.578713513952325,
+      "learning_rate": 1.2330050166660712e-06,
+      "loss": 0.3181,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7130345934900015,
+      "grad_norm": 0.5600573063550451,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.2916,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7142023062326668,
+      "grad_norm": 0.5516828869804629,
+      "learning_rate": 1.2134517172941563e-06,
+      "loss": 0.3097,
+      "step": 1468
+    },
+    {
+      "epoch": 1.715370018975332,
+      "grad_norm": 0.5791208974834355,
+      "learning_rate": 1.2037298961765065e-06,
+      "loss": 0.3371,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7165377317179975,
+      "grad_norm": 0.5926062063190835,
+      "learning_rate": 1.19404468143262e-06,
+      "loss": 0.3226,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7177054444606625,
+      "grad_norm": 0.6100855894147033,
+      "learning_rate": 1.184396113368288e-06,
+      "loss": 0.3008,
+      "step": 1471
+    },
+    {
+      "epoch": 1.718873157203328,
+      "grad_norm": 0.5719996654303868,
+      "learning_rate": 1.1747842321367886e-06,
+      "loss": 0.3185,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7200408699459933,
+      "grad_norm": 0.6766536512704038,
+      "learning_rate": 1.1652090777387248e-06,
+      "loss": 0.3327,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7212085826886585,
+      "grad_norm": 0.5665770135178536,
+      "learning_rate": 1.1556706900218572e-06,
+      "loss": 0.318,
+      "step": 1474
+    },
+    {
+      "epoch": 1.722376295431324,
+      "grad_norm": 0.6041065516107177,
+      "learning_rate": 1.1461691086809434e-06,
+      "loss": 0.3379,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7235440081739892,
+      "grad_norm": 0.6016737371801228,
+      "learning_rate": 1.1367043732575666e-06,
+      "loss": 0.3059,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7247117209166545,
+      "grad_norm": 0.6051130594578749,
+      "learning_rate": 1.1272765231399684e-06,
+      "loss": 0.3173,
+      "step": 1477
+    },
+    {
+      "epoch": 1.72587943365932,
+      "grad_norm": 0.5733460824309699,
+      "learning_rate": 1.1178855975628967e-06,
+      "loss": 0.32,
+      "step": 1478
+    },
+    {
+      "epoch": 1.727047146401985,
+      "grad_norm": 0.5562114116710137,
+      "learning_rate": 1.1085316356074272e-06,
+      "loss": 0.3013,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7282148591446505,
+      "grad_norm": 0.6104439707638637,
+      "learning_rate": 1.099214676200816e-06,
+      "loss": 0.3088,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7293825718873157,
+      "grad_norm": 0.5935918981001682,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.3174,
+      "step": 1481
+    },
+    {
+      "epoch": 1.730550284629981,
+      "grad_norm": 0.6063945505472416,
+      "learning_rate": 1.0806919199730614e-06,
+      "loss": 0.3099,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7317179973726464,
+      "grad_norm": 0.58550309687165,
+      "learning_rate": 1.0714862002358385e-06,
+      "loss": 0.3176,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7328857101153117,
+      "grad_norm": 0.5903961465847414,
+      "learning_rate": 1.0623176372149802e-06,
+      "loss": 0.3269,
+      "step": 1484
+    },
+    {
+      "epoch": 1.734053422857977,
+      "grad_norm": 0.6027356247406607,
+      "learning_rate": 1.0531862690661865e-06,
+      "loss": 0.32,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7352211356006424,
+      "grad_norm": 0.5600197671542984,
+      "learning_rate": 1.0440921337903699e-06,
+      "loss": 0.3046,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7363888483433074,
+      "grad_norm": 0.5689057746629985,
+      "learning_rate": 1.035035269233493e-06,
+      "loss": 0.3122,
+      "step": 1487
+    },
+    {
+      "epoch": 1.737556561085973,
+      "grad_norm": 0.5332475228301878,
+      "learning_rate": 1.0260157130864178e-06,
+      "loss": 0.2804,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7387242738286381,
+      "grad_norm": 0.6120614399610733,
+      "learning_rate": 1.0170335028847345e-06,
+      "loss": 0.3157,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7398919865713034,
+      "grad_norm": 0.5691041401545055,
+      "learning_rate": 1.0080886760086227e-06,
+      "loss": 0.3076,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7410596993139689,
+      "grad_norm": 0.5929420776779124,
+      "learning_rate": 9.991812696826887e-07,
+      "loss": 0.3403,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7422274120566341,
+      "grad_norm": 0.6159937867825703,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.3255,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7433951247992994,
+      "grad_norm": 0.5976780572145182,
+      "learning_rate": 9.814788668009766e-07,
+      "loss": 0.3023,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7445628375419648,
+      "grad_norm": 0.5547772280466156,
+      "learning_rate": 9.726839439151447e-07,
+      "loss": 0.3065,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7457305502846299,
+      "grad_norm": 0.5783566017853665,
+      "learning_rate": 9.63926588919083e-07,
+      "loss": 0.3173,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7468982630272953,
+      "grad_norm": 0.5937435085245287,
+      "learning_rate": 9.55206838257219e-07,
+      "loss": 0.3366,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7480659757699606,
+      "grad_norm": 0.5599877854994373,
+      "learning_rate": 9.465247282174806e-07,
+      "loss": 0.2887,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7492336885126258,
+      "grad_norm": 0.562338171569034,
+      "learning_rate": 9.378802949311583e-07,
+      "loss": 0.2982,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7504014012552913,
+      "grad_norm": 0.6218570308435145,
+      "learning_rate": 9.292735743727432e-07,
+      "loss": 0.3078,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7515691139979563,
+      "grad_norm": 0.5693599515940374,
+      "learning_rate": 9.207046023597866e-07,
+      "loss": 0.317,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7527368267406218,
+      "grad_norm": 0.6103574167170646,
+      "learning_rate": 9.121734145527394e-07,
+      "loss": 0.3051,
+      "step": 1501
+    },
+    {
+      "epoch": 1.753904539483287,
+      "grad_norm": 0.5823385949401512,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.3248,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7550722522259523,
+      "grad_norm": 0.561264319397526,
+      "learning_rate": 8.952245334118415e-07,
+      "loss": 0.2876,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7562399649686178,
+      "grad_norm": 0.5388951701570748,
+      "learning_rate": 8.868069106121002e-07,
+      "loss": 0.3008,
+      "step": 1504
+    },
+    {
+      "epoch": 1.757407677711283,
+      "grad_norm": 0.6179269600470059,
+      "learning_rate": 8.784272130861937e-07,
+      "loss": 0.3387,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7585753904539483,
+      "grad_norm": 0.6002001304135094,
+      "learning_rate": 8.700854757068988e-07,
+      "loss": 0.3317,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7597431031966138,
+      "grad_norm": 0.5642033268260335,
+      "learning_rate": 8.617817331890155e-07,
+      "loss": 0.3134,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7609108159392788,
+      "grad_norm": 0.551396798840725,
+      "learning_rate": 8.535160200892234e-07,
+      "loss": 0.315,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7620785286819443,
+      "grad_norm": 0.5556702079632481,
+      "learning_rate": 8.4528837080594e-07,
+      "loss": 0.3132,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7632462414246095,
+      "grad_norm": 0.5749432495541398,
+      "learning_rate": 8.370988195791807e-07,
+      "loss": 0.3205,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7644139541672748,
+      "grad_norm": 0.5781169921082643,
+      "learning_rate": 8.289474004904108e-07,
+      "loss": 0.308,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7655816669099402,
+      "grad_norm": 0.5755593871597986,
+      "learning_rate": 8.208341474624071e-07,
+      "loss": 0.3159,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7667493796526055,
+      "grad_norm": 0.5663269724481796,
+      "learning_rate": 8.127590942591112e-07,
+      "loss": 0.2915,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7679170923952707,
+      "grad_norm": 0.5508331856712115,
+      "learning_rate": 8.047222744854943e-07,
+      "loss": 0.3023,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7690848051379362,
+      "grad_norm": 0.6157365758887196,
+      "learning_rate": 7.96723721587419e-07,
+      "loss": 0.3173,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7702525178806012,
+      "grad_norm": 0.561213849870954,
+      "learning_rate": 7.887634688515e-07,
+      "loss": 0.3333,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7714202306232667,
+      "grad_norm": 0.5810707548752171,
+      "learning_rate": 7.808415494049515e-07,
+      "loss": 0.3268,
+      "step": 1517
+    },
+    {
+      "epoch": 1.772587943365932,
+      "grad_norm": 0.5672602265457064,
+      "learning_rate": 7.729579962154743e-07,
+      "loss": 0.2974,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7737556561085972,
+      "grad_norm": 0.5707079957076364,
+      "learning_rate": 7.651128420910925e-07,
+      "loss": 0.3072,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7749233688512627,
+      "grad_norm": 0.596894240612541,
+      "learning_rate": 7.573061196800414e-07,
+      "loss": 0.3295,
+      "step": 1520
+    },
+    {
+      "epoch": 1.776091081593928,
+      "grad_norm": 0.5928957941943206,
+      "learning_rate": 7.495378614706095e-07,
+      "loss": 0.343,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7772587943365932,
+      "grad_norm": 0.5481090189884966,
+      "learning_rate": 7.418080997910204e-07,
+      "loss": 0.3015,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7784265070792586,
+      "grad_norm": 0.5755912143062839,
+      "learning_rate": 7.341168668092857e-07,
+      "loss": 0.3254,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7795942198219237,
+      "grad_norm": 0.5814468494794668,
+      "learning_rate": 7.26464194533083e-07,
+      "loss": 0.3244,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7807619325645891,
+      "grad_norm": 0.577328035886048,
+      "learning_rate": 7.188501148096117e-07,
+      "loss": 0.304,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7819296453072544,
+      "grad_norm": 0.5682338236615159,
+      "learning_rate": 7.11274659325465e-07,
+      "loss": 0.3031,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7830973580499196,
+      "grad_norm": 0.5864504260644838,
+      "learning_rate": 7.037378596065047e-07,
+      "loss": 0.3185,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7842650707925851,
+      "grad_norm": 0.5901677988866156,
+      "learning_rate": 6.962397470177162e-07,
+      "loss": 0.3255,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7854327835352504,
+      "grad_norm": 0.5641337284172518,
+      "learning_rate": 6.887803527630865e-07,
+      "loss": 0.3192,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7866004962779156,
+      "grad_norm": 0.5909155549613954,
+      "learning_rate": 6.813597078854772e-07,
+      "loss": 0.3119,
+      "step": 1530
+    },
+    {
+      "epoch": 1.787768209020581,
+      "grad_norm": 0.5874984380175977,
+      "learning_rate": 6.739778432664867e-07,
+      "loss": 0.2982,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7889359217632461,
+      "grad_norm": 0.58517210777628,
+      "learning_rate": 6.666347896263326e-07,
+      "loss": 0.3016,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7901036345059116,
+      "grad_norm": 0.5588020637810773,
+      "learning_rate": 6.593305775237102e-07,
+      "loss": 0.3284,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7912713472485768,
+      "grad_norm": 0.5895059702191431,
+      "learning_rate": 6.520652373556746e-07,
+      "loss": 0.315,
+      "step": 1534
+    },
+    {
+      "epoch": 1.792439059991242,
+      "grad_norm": 0.5976588739720979,
+      "learning_rate": 6.448387993575155e-07,
+      "loss": 0.3436,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7936067727339076,
+      "grad_norm": 0.5787289022345898,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.303,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7947744854765728,
+      "grad_norm": 0.6053711480232326,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.3159,
+      "step": 1537
+    },
+    {
+      "epoch": 1.795942198219238,
+      "grad_norm": 0.5970738745618172,
+      "learning_rate": 6.233931983060104e-07,
+      "loss": 0.3534,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7971099109619035,
+      "grad_norm": 0.5575224504974947,
+      "learning_rate": 6.16322668100472e-07,
+      "loss": 0.3093,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7982776237045686,
+      "grad_norm": 0.5664747282281758,
+      "learning_rate": 6.092911888103404e-07,
+      "loss": 0.298,
+      "step": 1540
+    },
+    {
+      "epoch": 1.799445336447234,
+      "grad_norm": 0.5762925895193795,
+      "learning_rate": 6.022987896976706e-07,
+      "loss": 0.3215,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8006130491898993,
+      "grad_norm": 0.5941325284719488,
+      "learning_rate": 5.953454998618857e-07,
+      "loss": 0.3141,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8017807619325645,
+      "grad_norm": 0.5791831639150871,
+      "learning_rate": 5.884313482396497e-07,
+      "loss": 0.3293,
+      "step": 1543
+    },
+    {
+      "epoch": 1.80294847467523,
+      "grad_norm": 0.626855858859899,
+      "learning_rate": 5.815563636047539e-07,
+      "loss": 0.3178,
+      "step": 1544
+    },
+    {
+      "epoch": 1.804116187417895,
+      "grad_norm": 0.5734873372243158,
+      "learning_rate": 5.747205745679885e-07,
+      "loss": 0.2852,
+      "step": 1545
+    },
+    {
+      "epoch": 1.8052839001605605,
+      "grad_norm": 0.5915841723634604,
+      "learning_rate": 5.6792400957703e-07,
+      "loss": 0.3138,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8064516129032258,
+      "grad_norm": 0.5612611047256977,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.3106,
+      "step": 1547
+    },
+    {
+      "epoch": 1.807619325645891,
+      "grad_norm": 0.6437875201075615,
+      "learning_rate": 5.544486647069614e-07,
+      "loss": 0.3384,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8087870383885565,
+      "grad_norm": 0.5862655519662925,
+      "learning_rate": 5.477699409065607e-07,
+      "loss": 0.3164,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8099547511312217,
+      "grad_norm": 0.5550516030308618,
+      "learning_rate": 5.411305533091604e-07,
+      "loss": 0.3046,
+      "step": 1550
+    },
+    {
+      "epoch": 1.811122463873887,
+      "grad_norm": 0.5713870015500606,
+      "learning_rate": 5.345305295450997e-07,
+      "loss": 0.309,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8122901766165525,
+      "grad_norm": 0.5966077278731945,
+      "learning_rate": 5.279698970809011e-07,
+      "loss": 0.3298,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8134578893592175,
+      "grad_norm": 0.5584365011903022,
+      "learning_rate": 5.21448683219149e-07,
+      "loss": 0.2948,
+      "step": 1553
+    },
+    {
+      "epoch": 1.814625602101883,
+      "grad_norm": 0.6057743555453048,
+      "learning_rate": 5.149669150983938e-07,
+      "loss": 0.3368,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8157933148445482,
+      "grad_norm": 0.5597193417726661,
+      "learning_rate": 5.085246196930271e-07,
+      "loss": 0.3078,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8169610275872135,
+      "grad_norm": 0.5715416568929371,
+      "learning_rate": 5.02121823813172e-07,
+      "loss": 0.321,
+      "step": 1556
+    },
+    {
+      "epoch": 1.818128740329879,
+      "grad_norm": 0.59358464518882,
+      "learning_rate": 4.957585541045684e-07,
+      "loss": 0.3297,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8192964530725442,
+      "grad_norm": 0.5441316901391953,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.3055,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8204641658152094,
+      "grad_norm": 0.5651641370197638,
+      "learning_rate": 4.831506989615109e-07,
+      "loss": 0.3288,
+      "step": 1559
+    },
+    {
+      "epoch": 1.821631878557875,
+      "grad_norm": 0.5653876018725464,
+      "learning_rate": 4.769061659956464e-07,
+      "loss": 0.3021,
+      "step": 1560
+    },
+    {
+      "epoch": 1.82279959130054,
+      "grad_norm": 0.567623200788668,
+      "learning_rate": 4.707012641379871e-07,
+      "loss": 0.303,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8239673040432054,
+      "grad_norm": 0.5827965227358617,
+      "learning_rate": 4.64536019210724e-07,
+      "loss": 0.3031,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8251350167858706,
+      "grad_norm": 0.5789782877642979,
+      "learning_rate": 4.5841045687101017e-07,
+      "loss": 0.3394,
+      "step": 1563
+    },
+    {
+      "epoch": 1.826302729528536,
+      "grad_norm": 0.5546155910449154,
+      "learning_rate": 4.5232460261085966e-07,
+      "loss": 0.2749,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8274704422712014,
+      "grad_norm": 0.5709501694572496,
+      "learning_rate": 4.4627848175703315e-07,
+      "loss": 0.3119,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8286381550138666,
+      "grad_norm": 0.5899647368758275,
+      "learning_rate": 4.402721194709436e-07,
+      "loss": 0.3377,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8298058677565319,
+      "grad_norm": 0.6111776020058227,
+      "learning_rate": 4.34305540748543e-07,
+      "loss": 0.3337,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8309735804991973,
+      "grad_norm": 0.5827269838234809,
+      "learning_rate": 4.2837877042021915e-07,
+      "loss": 0.322,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8321412932418624,
+      "grad_norm": 0.5728327182844689,
+      "learning_rate": 4.224918331506955e-07,
+      "loss": 0.3043,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8333090059845278,
+      "grad_norm": 0.5293581669665233,
+      "learning_rate": 4.1664475343892727e-07,
+      "loss": 0.3001,
+      "step": 1570
+    },
+    {
+      "epoch": 1.834476718727193,
+      "grad_norm": 0.5782051028825217,
+      "learning_rate": 4.1083755561799997e-07,
+      "loss": 0.3236,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8356444314698583,
+      "grad_norm": 0.5654415131846452,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.3053,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8368121442125238,
+      "grad_norm": 0.5588799742817488,
+      "learning_rate": 3.9934290215104754e-07,
+      "loss": 0.3193,
+      "step": 1573
+    },
+    {
+      "epoch": 1.837979856955189,
+      "grad_norm": 0.5884697890517858,
+      "learning_rate": 3.9365549434092987e-07,
+      "loss": 0.3047,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8391475696978543,
+      "grad_norm": 0.5357653665258247,
+      "learning_rate": 3.8800806409327263e-07,
+      "loss": 0.3013,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8403152824405198,
+      "grad_norm": 0.5730827708681207,
+      "learning_rate": 3.82400634910306e-07,
+      "loss": 0.3154,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8414829951831848,
+      "grad_norm": 0.5865437799008645,
+      "learning_rate": 3.768332301277866e-07,
+      "loss": 0.3203,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8426507079258503,
+      "grad_norm": 0.5835561223853021,
+      "learning_rate": 3.7130587291490995e-07,
+      "loss": 0.3254,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8438184206685155,
+      "grad_norm": 0.6015242841104644,
+      "learning_rate": 3.658185862742103e-07,
+      "loss": 0.324,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8449861334111808,
+      "grad_norm": 0.5890506425544774,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.3172,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.628372828185003,
+      "learning_rate": 3.549643158856042e-07,
+      "loss": 0.3621,
+      "step": 1581
+    },
+    {
+      "epoch": 1.8473215588965115,
+      "grad_norm": 0.577508703569783,
+      "learning_rate": 3.4959737730860144e-07,
+      "loss": 0.2965,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8484892716391768,
+      "grad_norm": 0.5266849035718948,
+      "learning_rate": 3.4427059964539654e-07,
+      "loss": 0.296,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8496569843818422,
+      "grad_norm": 0.5421014272470868,
+      "learning_rate": 3.3898400506379937e-07,
+      "loss": 0.2922,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8508246971245073,
+      "grad_norm": 0.5739931169234624,
+      "learning_rate": 3.337376155643879e-07,
+      "loss": 0.3486,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8519924098671727,
+      "grad_norm": 0.5485713369037525,
+      "learning_rate": 3.2853145298042954e-07,
+      "loss": 0.3047,
+      "step": 1586
+    },
+    {
+      "epoch": 1.853160122609838,
+      "grad_norm": 0.5809070178775718,
+      "learning_rate": 3.2336553897778014e-07,
+      "loss": 0.298,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8543278353525032,
+      "grad_norm": 0.6030718931909618,
+      "learning_rate": 3.1823989505479934e-07,
+      "loss": 0.3052,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8554955480951687,
+      "grad_norm": 0.5939794259344422,
+      "learning_rate": 3.1315454254226105e-07,
+      "loss": 0.306,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8566632608378337,
+      "grad_norm": 0.59403088250487,
+      "learning_rate": 3.081095026032599e-07,
+      "loss": 0.3185,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8578309735804992,
+      "grad_norm": 0.6152984102403254,
+      "learning_rate": 3.0310479623313125e-07,
+      "loss": 0.3211,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8589986863231645,
+      "grad_norm": 0.5936908736701824,
+      "learning_rate": 2.9814044425935605e-07,
+      "loss": 0.3177,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8601663990658297,
+      "grad_norm": 0.5820913071966881,
+      "learning_rate": 2.93216467341475e-07,
+      "loss": 0.3188,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8613341118084952,
+      "grad_norm": 0.5410293149310855,
+      "learning_rate": 2.8833288597100996e-07,
+      "loss": 0.286,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8625018245511604,
+      "grad_norm": 0.5463317051609053,
+      "learning_rate": 2.8348972047137155e-07,
+      "loss": 0.3313,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8636695372938257,
+      "grad_norm": 0.5886116556526418,
+      "learning_rate": 2.7868699099777295e-07,
+      "loss": 0.2937,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8648372500364911,
+      "grad_norm": 0.5834953915581415,
+      "learning_rate": 2.7392471753715623e-07,
+      "loss": 0.336,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8660049627791562,
+      "grad_norm": 0.5778769478905829,
+      "learning_rate": 2.69202919908097e-07,
+      "loss": 0.3252,
+      "step": 1598
+    },
+    {
+      "epoch": 1.8671726755218216,
+      "grad_norm": 0.5562916083723717,
+      "learning_rate": 2.645216177607313e-07,
+      "loss": 0.2958,
+      "step": 1599
+    },
+    {
+      "epoch": 1.868340388264487,
+      "grad_norm": 0.5717791200566672,
+      "learning_rate": 2.5988083057666534e-07,
+      "loss": 0.3152,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8695081010071521,
+      "grad_norm": 0.5683081779192658,
+      "learning_rate": 2.552805776689049e-07,
+      "loss": 0.2866,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8706758137498176,
+      "grad_norm": 0.5971011618737242,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.3364,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8718435264924829,
+      "grad_norm": 0.6019958045969178,
+      "learning_rate": 2.462017510907955e-07,
+      "loss": 0.3087,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8730112392351481,
+      "grad_norm": 0.5465559629198596,
+      "learning_rate": 2.417232152027016e-07,
+      "loss": 0.2882,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8741789519778136,
+      "grad_norm": 0.6123102510763562,
+      "learning_rate": 2.3728528915526462e-07,
+      "loss": 0.3353,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8753466647204786,
+      "grad_norm": 0.5856566324252752,
+      "learning_rate": 2.3288799141726548e-07,
+      "loss": 0.3247,
+      "step": 1606
+    },
+    {
+      "epoch": 1.876514377463144,
+      "grad_norm": 0.6574972755297316,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.327,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8776820902058093,
+      "grad_norm": 0.5627565390623674,
+      "learning_rate": 2.242153538992331e-07,
+      "loss": 0.3081,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8788498029484746,
+      "grad_norm": 0.585365695507248,
+      "learning_rate": 2.1994005021106713e-07,
+      "loss": 0.3356,
+      "step": 1609
+    },
+    {
+      "epoch": 1.88001751569114,
+      "grad_norm": 0.5729201144590117,
+      "learning_rate": 2.1570544701592254e-07,
+      "loss": 0.3103,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8811852284338053,
+      "grad_norm": 0.5479012582163604,
+      "learning_rate": 2.1151156193643496e-07,
+      "loss": 0.3295,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.5655106077657962,
+      "learning_rate": 2.0735841242578992e-07,
+      "loss": 0.3271,
+      "step": 1612
+    },
+    {
+      "epoch": 1.883520653919136,
+      "grad_norm": 0.6766687673872185,
+      "learning_rate": 2.0324601576764525e-07,
+      "loss": 0.3032,
+      "step": 1613
+    },
+    {
+      "epoch": 1.884688366661801,
+      "grad_norm": 0.5741397792266839,
+      "learning_rate": 1.9917438907606556e-07,
+      "loss": 0.3105,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8858560794044665,
+      "grad_norm": 0.5760279859650697,
+      "learning_rate": 1.951435492954501e-07,
+      "loss": 0.3191,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8870237921471318,
+      "grad_norm": 0.5448591070588273,
+      "learning_rate": 1.9115351320045495e-07,
+      "loss": 0.3203,
+      "step": 1616
+    },
+    {
+      "epoch": 1.888191504889797,
+      "grad_norm": 0.5997646624339391,
+      "learning_rate": 1.8720429739592983e-07,
+      "loss": 0.3614,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8893592176324625,
+      "grad_norm": 0.5854894720098576,
+      "learning_rate": 1.8329591831685146e-07,
+      "loss": 0.2828,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8905269303751278,
+      "grad_norm": 0.5783778292665679,
+      "learning_rate": 1.7942839222824804e-07,
+      "loss": 0.3379,
+      "step": 1619
+    },
+    {
+      "epoch": 1.891694643117793,
+      "grad_norm": 0.5710351049566196,
+      "learning_rate": 1.7560173522513268e-07,
+      "loss": 0.3169,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8928623558604585,
+      "grad_norm": 0.613220428251377,
+      "learning_rate": 1.7181596323244453e-07,
+      "loss": 0.322,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8940300686031235,
+      "grad_norm": 0.6076188199019776,
+      "learning_rate": 1.6807109200496995e-07,
+      "loss": 0.2969,
+      "step": 1622
+    },
+    {
+      "epoch": 1.895197781345789,
+      "grad_norm": 0.5588169384747448,
+      "learning_rate": 1.6436713712729036e-07,
+      "loss": 0.2976,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8963654940884542,
+      "grad_norm": 0.6022141869950728,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.3018,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8975332068311195,
+      "grad_norm": 0.5913215641048575,
+      "learning_rate": 1.5708203790817166e-07,
+      "loss": 0.3088,
+      "step": 1625
+    },
+    {
+      "epoch": 1.898700919573785,
+      "grad_norm": 0.5541774936556441,
+      "learning_rate": 1.5350092388425108e-07,
+      "loss": 0.303,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8998686323164502,
+      "grad_norm": 0.5906427477532417,
+      "learning_rate": 1.4996078684503145e-07,
+      "loss": 0.3249,
+      "step": 1627
+    },
+    {
+      "epoch": 1.9010363450591155,
+      "grad_norm": 0.5605785265862366,
+      "learning_rate": 1.464616415230702e-07,
+      "loss": 0.2995,
+      "step": 1628
+    },
+    {
+      "epoch": 1.902204057801781,
+      "grad_norm": 0.6076539677533489,
+      "learning_rate": 1.4300350248034002e-07,
+      "loss": 0.3273,
+      "step": 1629
+    },
+    {
+      "epoch": 1.903371770544446,
+      "grad_norm": 0.5623484602443884,
+      "learning_rate": 1.3958638410815905e-07,
+      "loss": 0.2845,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9045394832871114,
+      "grad_norm": 0.5722985549178372,
+      "learning_rate": 1.3621030062713759e-07,
+      "loss": 0.2983,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9057071960297767,
+      "grad_norm": 0.5564661280896065,
+      "learning_rate": 1.3287526608711132e-07,
+      "loss": 0.3234,
+      "step": 1632
+    },
+    {
+      "epoch": 1.906874908772442,
+      "grad_norm": 0.5999293725782077,
+      "learning_rate": 1.2958129436709154e-07,
+      "loss": 0.3202,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9080426215151074,
+      "grad_norm": 0.5402180629120761,
+      "learning_rate": 1.2632839917520179e-07,
+      "loss": 0.3033,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9092103342577724,
+      "grad_norm": 0.5960117593490167,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.3202,
+      "step": 1635
+    },
+    {
+      "epoch": 1.910378047000438,
+      "grad_norm": 0.5602220371818334,
+      "learning_rate": 1.1994589235353682e-07,
+      "loss": 0.3148,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9115457597431031,
+      "grad_norm": 0.5450879406404854,
+      "learning_rate": 1.1681630728506699e-07,
+      "loss": 0.3028,
+      "step": 1637
+    },
+    {
+      "epoch": 1.9127134724857684,
+      "grad_norm": 0.5808031736241751,
+      "learning_rate": 1.1372785186723135e-07,
+      "loss": 0.3123,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9138811852284339,
+      "grad_norm": 0.5657787115207192,
+      "learning_rate": 1.1068053895288089e-07,
+      "loss": 0.3409,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9150488979710991,
+      "grad_norm": 0.5447975880936112,
+      "learning_rate": 1.0767438122364914e-07,
+      "loss": 0.2807,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9162166107137644,
+      "grad_norm": 0.5754145503055412,
+      "learning_rate": 1.0470939118989997e-07,
+      "loss": 0.3148,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9173843234564298,
+      "grad_norm": 0.5561889406195449,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2991,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9185520361990949,
+      "grad_norm": 0.5947344486643699,
+      "learning_rate": 9.890296339363448e-08,
+      "loss": 0.3263,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9197197489417603,
+      "grad_norm": 0.6113390242831112,
+      "learning_rate": 9.606154979502791e-08,
+      "loss": 0.3425,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9208874616844256,
+      "grad_norm": 0.5825714582029453,
+      "learning_rate": 9.326135221961907e-08,
+      "loss": 0.3189,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9220551744270908,
+      "grad_norm": 0.5461470272349032,
+      "learning_rate": 9.0502382320653e-08,
+      "loss": 0.3016,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9232228871697563,
+      "grad_norm": 0.6174274816540521,
+      "learning_rate": 8.778465157979976e-08,
+      "loss": 0.3001,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9243905999124216,
+      "grad_norm": 0.5548027377219115,
+      "learning_rate": 8.510817130711224e-08,
+      "loss": 0.3127,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9255583126550868,
+      "grad_norm": 0.5832172033204495,
+      "learning_rate": 8.247295264097288e-08,
+      "loss": 0.2963,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9267260253977523,
+      "grad_norm": 0.5924137992423935,
+      "learning_rate": 7.987900654805481e-08,
+      "loss": 0.3193,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9278937381404173,
+      "grad_norm": 0.581674164473265,
+      "learning_rate": 7.732634382326965e-08,
+      "loss": 0.3137,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9290614508830828,
+      "grad_norm": 0.5706021711230395,
+      "learning_rate": 7.481497508972313e-08,
+      "loss": 0.3262,
+      "step": 1652
+    },
+    {
+      "epoch": 1.930229163625748,
+      "grad_norm": 0.5613187175028672,
+      "learning_rate": 7.234491079867623e-08,
+      "loss": 0.3138,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9313968763684133,
+      "grad_norm": 0.5937998787826614,
+      "learning_rate": 6.991616122949629e-08,
+      "loss": 0.3272,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9325645891110788,
+      "grad_norm": 0.5657398856868785,
+      "learning_rate": 6.752873648961821e-08,
+      "loss": 0.3167,
+      "step": 1655
+    },
+    {
+      "epoch": 1.933732301853744,
+      "grad_norm": 0.6145539718076212,
+      "learning_rate": 6.51826465144978e-08,
+      "loss": 0.341,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9349000145964093,
+      "grad_norm": 0.5472619384230469,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.3034,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9360677273390747,
+      "grad_norm": 0.5770904874349805,
+      "learning_rate": 6.061450974022775e-08,
+      "loss": 0.3212,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9372354400817398,
+      "grad_norm": 0.5329147106833191,
+      "learning_rate": 5.839248195174008e-08,
+      "loss": 0.3092,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9384031528244052,
+      "grad_norm": 0.5700598742456262,
+      "learning_rate": 5.621182694925731e-08,
+      "loss": 0.3168,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9395708655670705,
+      "grad_norm": 0.548007743351208,
+      "learning_rate": 5.407255380774579e-08,
+      "loss": 0.3029,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9407385783097357,
+      "grad_norm": 0.5893534871878833,
+      "learning_rate": 5.19746714299596e-08,
+      "loss": 0.3189,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9419062910524012,
+      "grad_norm": 0.5474173021952493,
+      "learning_rate": 4.991818854640396e-08,
+      "loss": 0.3128,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9430740037950665,
+      "grad_norm": 0.5529864303172826,
+      "learning_rate": 4.79031137152941e-08,
+      "loss": 0.3203,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9442417165377317,
+      "grad_norm": 0.5127438928849724,
+      "learning_rate": 4.592945532252424e-08,
+      "loss": 0.3001,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9454094292803972,
+      "grad_norm": 0.5937990468604879,
+      "learning_rate": 4.399722158162867e-08,
+      "loss": 0.2963,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9465771420230622,
+      "grad_norm": 0.5516208240302628,
+      "learning_rate": 4.210642053375069e-08,
+      "loss": 0.3006,
+      "step": 1667
+    },
+    {
+      "epoch": 1.9477448547657277,
+      "grad_norm": 0.5821955928588767,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.2993,
+      "step": 1668
+    },
+    {
+      "epoch": 1.948912567508393,
+      "grad_norm": 0.5836409139433892,
+      "learning_rate": 3.844914781946485e-08,
+      "loss": 0.2897,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9500802802510582,
+      "grad_norm": 0.548074486984568,
+      "learning_rate": 3.668269137308666e-08,
+      "loss": 0.2925,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9512479929937236,
+      "grad_norm": 0.5568578991164567,
+      "learning_rate": 3.495769805972215e-08,
+      "loss": 0.3223,
+      "step": 1671
+    },
+    {
+      "epoch": 1.952415705736389,
+      "grad_norm": 0.573457695080691,
+      "learning_rate": 3.327417505806785e-08,
+      "loss": 0.3066,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9535834184790541,
+      "grad_norm": 0.5670478605517256,
+      "learning_rate": 3.163212937423832e-08,
+      "loss": 0.319,
+      "step": 1673
+    },
+    {
+      "epoch": 1.9547511312217196,
+      "grad_norm": 0.5574977933150648,
+      "learning_rate": 3.003156784173511e-08,
+      "loss": 0.303,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9559188439643846,
+      "grad_norm": 0.6047331183538928,
+      "learning_rate": 2.8472497121424524e-08,
+      "loss": 0.3153,
+      "step": 1675
+    },
+    {
+      "epoch": 1.9570865567070501,
+      "grad_norm": 0.5907333203967309,
+      "learning_rate": 2.695492370149988e-08,
+      "loss": 0.3067,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9582542694497154,
+      "grad_norm": 0.5369816474044258,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.3007,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9594219821923806,
+      "grad_norm": 0.5478868876712718,
+      "learning_rate": 2.4044293852099055e-08,
+      "loss": 0.3172,
+      "step": 1678
+    },
+    {
+      "epoch": 1.960589694935046,
+      "grad_norm": 0.5704982876125324,
+      "learning_rate": 2.265124953543918e-08,
+      "loss": 0.3277,
+      "step": 1679
+    },
+    {
+      "epoch": 1.9617574076777111,
+      "grad_norm": 0.5693559213289971,
+      "learning_rate": 2.1299726744747896e-08,
+      "loss": 0.3138,
+      "step": 1680
+    },
+    {
+      "epoch": 1.9629251204203766,
+      "grad_norm": 0.6306631714372946,
+      "learning_rate": 1.998973110449498e-08,
+      "loss": 0.3036,
+      "step": 1681
+    },
+    {
+      "epoch": 1.9640928331630418,
+      "grad_norm": 0.5514448472228228,
+      "learning_rate": 1.872126806633068e-08,
+      "loss": 0.323,
+      "step": 1682
+    },
+    {
+      "epoch": 1.965260545905707,
+      "grad_norm": 0.5877525995893049,
+      "learning_rate": 1.749434290906571e-08,
+      "loss": 0.2928,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9664282586483726,
+      "grad_norm": 0.6007990288362013,
+      "learning_rate": 1.630896073864352e-08,
+      "loss": 0.3262,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9675959713910378,
+      "grad_norm": 0.5609966204355938,
+      "learning_rate": 1.5165126488126957e-08,
+      "loss": 0.3132,
+      "step": 1685
+    },
+    {
+      "epoch": 1.968763684133703,
+      "grad_norm": 0.5738593672629273,
+      "learning_rate": 1.4062844917672735e-08,
+      "loss": 0.3309,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9699313968763685,
+      "grad_norm": 0.5407080980930443,
+      "learning_rate": 1.300212061451367e-08,
+      "loss": 0.3007,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9710991096190336,
+      "grad_norm": 0.6356771830823165,
+      "learning_rate": 1.1982957992936472e-08,
+      "loss": 0.3178,
+      "step": 1688
+    },
+    {
+      "epoch": 1.972266822361699,
+      "grad_norm": 0.6205844287799247,
+      "learning_rate": 1.1005361294268435e-08,
+      "loss": 0.347,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9734345351043643,
+      "grad_norm": 0.5629008202330193,
+      "learning_rate": 1.0069334586854106e-08,
+      "loss": 0.3176,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9746022478470295,
+      "grad_norm": 0.5710179242590563,
+      "learning_rate": 9.174881766043086e-09,
+      "loss": 0.3131,
+      "step": 1691
+    },
+    {
+      "epoch": 1.975769960589695,
+      "grad_norm": 0.5979055562436246,
+      "learning_rate": 8.322006554171147e-09,
+      "loss": 0.3222,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9769376733323603,
+      "grad_norm": 0.5896594529025474,
+      "learning_rate": 7.510712500546913e-09,
+      "loss": 0.319,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9781053860750255,
+      "grad_norm": 0.5854713920350555,
+      "learning_rate": 6.741002981435207e-09,
+      "loss": 0.3251,
+      "step": 1694
+    },
+    {
+      "epoch": 1.979273098817691,
+      "grad_norm": 0.5836473802473287,
+      "learning_rate": 6.0128812000415096e-09,
+      "loss": 0.3337,
+      "step": 1695
+    },
+    {
+      "epoch": 1.980440811560356,
+      "grad_norm": 0.5818896865508579,
+      "learning_rate": 5.326350186503071e-09,
+      "loss": 0.336,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9816085243030215,
+      "grad_norm": 0.5759857182472697,
+      "learning_rate": 4.681412797872265e-09,
+      "loss": 0.3239,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9827762370456867,
+      "grad_norm": 0.5971422431217652,
+      "learning_rate": 4.0780717181077015e-09,
+      "loss": 0.3661,
+      "step": 1698
+    },
+    {
+      "epoch": 1.983943949788352,
+      "grad_norm": 0.5432739703561003,
+      "learning_rate": 3.5163294580609076e-09,
+      "loss": 0.2994,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9851116625310175,
+      "grad_norm": 0.5960430671196062,
+      "learning_rate": 2.9961883554674443e-09,
+      "loss": 0.3186,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9862793752736827,
+      "grad_norm": 0.58369696055803,
+      "learning_rate": 2.5176505749346937e-09,
+      "loss": 0.3103,
+      "step": 1701
+    },
+    {
+      "epoch": 1.987447088016348,
+      "grad_norm": 0.565553496180529,
+      "learning_rate": 2.080718107935198e-09,
+      "loss": 0.2999,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9886148007590134,
+      "grad_norm": 0.6042259898055473,
+      "learning_rate": 1.6853927727988883e-09,
+      "loss": 0.306,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9897825135016785,
+      "grad_norm": 0.5595029869127405,
+      "learning_rate": 1.3316762147030925e-09,
+      "loss": 0.3428,
+      "step": 1704
+    },
+    {
+      "epoch": 1.990950226244344,
+      "grad_norm": 0.592613206105393,
+      "learning_rate": 1.019569905666984e-09,
+      "loss": 0.3432,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9921179389870092,
+      "grad_norm": 0.5835029302825108,
+      "learning_rate": 7.490751445449196e-10,
+      "loss": 0.3302,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9932856517296744,
+      "grad_norm": 0.5657750133088484,
+      "learning_rate": 5.201930570242209e-10,
+      "loss": 0.3123,
+      "step": 1707
+    },
+    {
+      "epoch": 1.99445336447234,
+      "grad_norm": 0.5774197497896288,
+      "learning_rate": 3.3292459561518055e-10,
+      "loss": 0.3149,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9956210772150051,
+      "grad_norm": 0.5608768646582435,
+      "learning_rate": 1.872705396488428e-10,
+      "loss": 0.3212,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9967887899576704,
+      "grad_norm": 0.551873237668583,
+      "learning_rate": 8.323149527811325e-11,
+      "loss": 0.3138,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9979565027003359,
+      "grad_norm": 0.5751895955442534,
+      "learning_rate": 2.080789546776707e-11,
+      "loss": 0.3273,
+      "step": 1711
+    },
+    {
+      "epoch": 1.999124215443001,
+      "grad_norm": 0.5383716638167484,
+      "learning_rate": 0.0,
+      "loss": 0.3178,
+      "step": 1712
+    },
+    {
+      "epoch": 1.999124215443001,
+      "step": 1712,
+      "total_flos": 312322909716480.0,
+      "train_loss": 0.39118012141819314,
+      "train_runtime": 6498.1994,
+      "train_samples_per_second": 33.733,
+      "train_steps_per_second": 0.263
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1712,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 312322909716480.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}