diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,16 +1,16 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.22680412371134,
+  "epoch": 0.02577319587628866,
   "eval_steps": 97,
-  "global_step": 820,
+  "global_step": 5,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.005154639175257732,
-      "grad_norm": 0.8343322277069092,
+      "grad_norm": 0.8365415930747986,
       "learning_rate": 2.0000000000000003e-06,
       "loss": 0.9672,
       "step": 1
@@ -18,5807 +18,38 @@
     {
       "epoch": 0.005154639175257732,
       "eval_loss": 0.986353874206543,
-      "eval_runtime": 22.8548,
-      "eval_samples_per_second": 7.176,
-      "eval_steps_per_second": 1.794,
+      "eval_runtime": 22.9621,
+      "eval_samples_per_second": 7.142,
+      "eval_steps_per_second": 1.786,
       "step": 1
     },
     {
       "epoch": 0.010309278350515464,
-      "grad_norm": 0.7854290008544922,
+      "grad_norm": 0.7841001749038696,
       "learning_rate": 4.000000000000001e-06,
       "loss": 1.0214,
       "step": 2
     },
     {
       "epoch": 0.015463917525773196,
-      "grad_norm": 0.7100067734718323,
+      "grad_norm": 0.6986056566238403,
       "learning_rate": 6e-06,
-      "loss": 0.8938,
+      "loss": 0.8946,
       "step": 3
     },
     {
       "epoch": 0.020618556701030927,
-      "grad_norm": 0.80195152759552,
+      "grad_norm": 0.8076366782188416,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.901,
+      "loss": 0.904,
       "step": 4
     },
     {
       "epoch": 0.02577319587628866,
-      "grad_norm": 1.1374644041061401,
+      "grad_norm": 1.1367769241333008,
       "learning_rate": 1e-05,
-      "loss": 1.0694,
+      "loss": 1.0636,
       "step": 5
-    },
-    {
-      "epoch": 0.030927835051546393,
-      "grad_norm": 1.244362473487854,
-      "learning_rate": 1.2e-05,
-      "loss": 1.1299,
-      "step": 6
-    },
-    {
-      "epoch": 0.03608247422680412,
-      "grad_norm": 0.7648528218269348,
-      "learning_rate": 1.4000000000000001e-05,
-      "loss": 0.9204,
-      "step": 7
-    },
-    {
-      "epoch": 0.041237113402061855,
-      "grad_norm": 0.6700325608253479,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 0.6981,
-      "step": 8
-    },
-    {
-      "epoch": 0.04639175257731959,
-      "grad_norm": 0.9419430494308472,
-      "learning_rate": 1.8e-05,
-      "loss": 0.8936,
-      "step": 9
-    },
-    {
-      "epoch": 0.05154639175257732,
-      "grad_norm": 0.8192828893661499,
-      "learning_rate": 2e-05,
-      "loss": 0.9367,
-      "step": 10
-    },
-    {
-      "epoch": 0.05670103092783505,
-      "grad_norm": 1.0335086584091187,
-      "learning_rate": 2.2000000000000003e-05,
-      "loss": 0.9871,
-      "step": 11
-    },
-    {
-      "epoch": 0.061855670103092786,
-      "grad_norm": 0.7089415788650513,
-      "learning_rate": 2.4e-05,
-      "loss": 1.061,
-      "step": 12
-    },
-    {
-      "epoch": 0.06701030927835051,
-      "grad_norm": 0.6549484133720398,
-      "learning_rate": 2.6000000000000002e-05,
-      "loss": 0.8456,
-      "step": 13
-    },
-    {
-      "epoch": 0.07216494845360824,
-      "grad_norm": 0.7343222498893738,
-      "learning_rate": 2.8000000000000003e-05,
-      "loss": 0.8595,
-      "step": 14
-    },
-    {
-      "epoch": 0.07731958762886598,
-      "grad_norm": 0.7335950136184692,
-      "learning_rate": 3e-05,
-      "loss": 0.9044,
-      "step": 15
-    },
-    {
-      "epoch": 0.08247422680412371,
-      "grad_norm": 0.6129910945892334,
-      "learning_rate": 3.2000000000000005e-05,
-      "loss": 0.9111,
-      "step": 16
-    },
-    {
-      "epoch": 0.08762886597938144,
-      "grad_norm": 0.6522359848022461,
-      "learning_rate": 3.4000000000000007e-05,
-      "loss": 0.9435,
-      "step": 17
-    },
-    {
-      "epoch": 0.09278350515463918,
-      "grad_norm": 0.4705953299999237,
-      "learning_rate": 3.6e-05,
-      "loss": 0.8463,
-      "step": 18
-    },
-    {
-      "epoch": 0.0979381443298969,
-      "grad_norm": 0.42669880390167236,
-      "learning_rate": 3.8e-05,
-      "loss": 0.8329,
-      "step": 19
-    },
-    {
-      "epoch": 0.10309278350515463,
-      "grad_norm": 0.6905720233917236,
-      "learning_rate": 4e-05,
-      "loss": 1.1096,
-      "step": 20
-    },
-    {
-      "epoch": 0.10824742268041238,
-      "grad_norm": 0.3870203495025635,
-      "learning_rate": 4.2e-05,
-      "loss": 0.8613,
-      "step": 21
-    },
-    {
-      "epoch": 0.1134020618556701,
-      "grad_norm": 0.5222039818763733,
-      "learning_rate": 4.4000000000000006e-05,
-      "loss": 0.8606,
-      "step": 22
-    },
-    {
-      "epoch": 0.11855670103092783,
-      "grad_norm": 0.3936106264591217,
-      "learning_rate": 4.600000000000001e-05,
-      "loss": 0.9014,
-      "step": 23
-    },
-    {
-      "epoch": 0.12371134020618557,
-      "grad_norm": 0.3917650878429413,
-      "learning_rate": 4.8e-05,
-      "loss": 0.9613,
-      "step": 24
-    },
-    {
-      "epoch": 0.12886597938144329,
-      "grad_norm": 0.371128112077713,
-      "learning_rate": 5e-05,
-      "loss": 0.8298,
-      "step": 25
-    },
-    {
-      "epoch": 0.13402061855670103,
-      "grad_norm": 0.44909828901290894,
-      "learning_rate": 5.2000000000000004e-05,
-      "loss": 0.8887,
-      "step": 26
-    },
-    {
-      "epoch": 0.13917525773195877,
-      "grad_norm": 0.5001224875450134,
-      "learning_rate": 5.4000000000000005e-05,
-      "loss": 0.8943,
-      "step": 27
-    },
-    {
-      "epoch": 0.14432989690721648,
-      "grad_norm": 0.4336227774620056,
-      "learning_rate": 5.6000000000000006e-05,
-      "loss": 0.8709,
-      "step": 28
-    },
-    {
-      "epoch": 0.14948453608247422,
-      "grad_norm": 0.36210742592811584,
-      "learning_rate": 5.8e-05,
-      "loss": 0.7933,
-      "step": 29
-    },
-    {
-      "epoch": 0.15463917525773196,
-      "grad_norm": 0.560114324092865,
-      "learning_rate": 6e-05,
-      "loss": 0.9966,
-      "step": 30
-    },
-    {
-      "epoch": 0.15979381443298968,
-      "grad_norm": 0.3248458802700043,
-      "learning_rate": 6.2e-05,
-      "loss": 0.6494,
-      "step": 31
-    },
-    {
-      "epoch": 0.16494845360824742,
-      "grad_norm": 0.51533442735672,
-      "learning_rate": 6.400000000000001e-05,
-      "loss": 0.9284,
-      "step": 32
-    },
-    {
-      "epoch": 0.17010309278350516,
-      "grad_norm": 0.4753219187259674,
-      "learning_rate": 6.6e-05,
-      "loss": 0.722,
-      "step": 33
-    },
-    {
-      "epoch": 0.17525773195876287,
-      "grad_norm": 0.42082515358924866,
-      "learning_rate": 6.800000000000001e-05,
-      "loss": 0.8662,
-      "step": 34
-    },
-    {
-      "epoch": 0.18041237113402062,
-      "grad_norm": 0.5010572075843811,
-      "learning_rate": 7e-05,
-      "loss": 0.8844,
-      "step": 35
-    },
-    {
-      "epoch": 0.18556701030927836,
-      "grad_norm": 0.3900461792945862,
-      "learning_rate": 7.2e-05,
-      "loss": 0.7872,
-      "step": 36
-    },
-    {
-      "epoch": 0.19072164948453607,
-      "grad_norm": 0.3964938223361969,
-      "learning_rate": 7.4e-05,
-      "loss": 0.7122,
-      "step": 37
-    },
-    {
-      "epoch": 0.1958762886597938,
-      "grad_norm": 0.44170987606048584,
-      "learning_rate": 7.6e-05,
-      "loss": 0.7552,
-      "step": 38
-    },
-    {
-      "epoch": 0.20103092783505155,
-      "grad_norm": 0.42387211322784424,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 0.9254,
-      "step": 39
-    },
-    {
-      "epoch": 0.20618556701030927,
-      "grad_norm": 0.4199470281600952,
-      "learning_rate": 8e-05,
-      "loss": 0.7172,
-      "step": 40
-    },
-    {
-      "epoch": 0.211340206185567,
-      "grad_norm": 0.5539634227752686,
-      "learning_rate": 8.2e-05,
-      "loss": 0.8196,
-      "step": 41
-    },
-    {
-      "epoch": 0.21649484536082475,
-      "grad_norm": 0.422987699508667,
-      "learning_rate": 8.4e-05,
-      "loss": 0.8814,
-      "step": 42
-    },
-    {
-      "epoch": 0.22164948453608246,
-      "grad_norm": 0.431035578250885,
-      "learning_rate": 8.6e-05,
-      "loss": 0.9498,
-      "step": 43
-    },
-    {
-      "epoch": 0.2268041237113402,
-      "grad_norm": 0.36294519901275635,
-      "learning_rate": 8.800000000000001e-05,
-      "loss": 0.6622,
-      "step": 44
-    },
-    {
-      "epoch": 0.23195876288659795,
-      "grad_norm": 0.43537455797195435,
-      "learning_rate": 9e-05,
-      "loss": 0.7567,
-      "step": 45
-    },
-    {
-      "epoch": 0.23711340206185566,
-      "grad_norm": 0.513528048992157,
-      "learning_rate": 9.200000000000001e-05,
-      "loss": 1.104,
-      "step": 46
-    },
-    {
-      "epoch": 0.2422680412371134,
-      "grad_norm": 0.5338684916496277,
-      "learning_rate": 9.4e-05,
-      "loss": 0.8215,
-      "step": 47
-    },
-    {
-      "epoch": 0.24742268041237114,
-      "grad_norm": 0.4168243408203125,
-      "learning_rate": 9.6e-05,
-      "loss": 0.8727,
-      "step": 48
-    },
-    {
-      "epoch": 0.25257731958762886,
-      "grad_norm": 0.4411843419075012,
-      "learning_rate": 9.8e-05,
-      "loss": 0.9479,
-      "step": 49
-    },
-    {
-      "epoch": 0.25773195876288657,
-      "grad_norm": 0.43782272934913635,
-      "learning_rate": 0.0001,
-      "loss": 0.9748,
-      "step": 50
-    },
-    {
-      "epoch": 0.26288659793814434,
-      "grad_norm": 0.37000253796577454,
-      "learning_rate": 9.999970848314005e-05,
-      "loss": 0.6958,
-      "step": 51
-    },
-    {
-      "epoch": 0.26804123711340205,
-      "grad_norm": 0.4658275544643402,
-      "learning_rate": 9.999883393595947e-05,
-      "loss": 0.8385,
-      "step": 52
-    },
-    {
-      "epoch": 0.27319587628865977,
-      "grad_norm": 0.37418484687805176,
-      "learning_rate": 9.999737636865609e-05,
-      "loss": 0.7309,
-      "step": 53
-    },
-    {
-      "epoch": 0.27835051546391754,
-      "grad_norm": 0.42213866114616394,
-      "learning_rate": 9.99953357982261e-05,
-      "loss": 0.7723,
-      "step": 54
-    },
-    {
-      "epoch": 0.28350515463917525,
-      "grad_norm": 0.45320671796798706,
-      "learning_rate": 9.999271224846396e-05,
-      "loss": 0.8041,
-      "step": 55
-    },
-    {
-      "epoch": 0.28865979381443296,
-      "grad_norm": 0.34378930926322937,
-      "learning_rate": 9.998950574996199e-05,
-      "loss": 0.6543,
-      "step": 56
-    },
-    {
-      "epoch": 0.29381443298969073,
-      "grad_norm": 0.5220825672149658,
-      "learning_rate": 9.998571634011015e-05,
-      "loss": 0.931,
-      "step": 57
-    },
-    {
-      "epoch": 0.29896907216494845,
-      "grad_norm": 0.3490442931652069,
-      "learning_rate": 9.998134406309554e-05,
-      "loss": 0.748,
-      "step": 58
-    },
-    {
-      "epoch": 0.30412371134020616,
-      "grad_norm": 0.35038453340530396,
-      "learning_rate": 9.99763889699018e-05,
-      "loss": 0.6783,
-      "step": 59
-    },
-    {
-      "epoch": 0.30927835051546393,
-      "grad_norm": 0.4343213140964508,
-      "learning_rate": 9.99708511183087e-05,
-      "loss": 0.8773,
-      "step": 60
-    },
-    {
-      "epoch": 0.31443298969072164,
-      "grad_norm": 0.3930794596672058,
-      "learning_rate": 9.996473057289132e-05,
-      "loss": 0.8495,
-      "step": 61
-    },
-    {
-      "epoch": 0.31958762886597936,
-      "grad_norm": 0.41221436858177185,
-      "learning_rate": 9.995802740501933e-05,
-      "loss": 0.8531,
-      "step": 62
-    },
-    {
-      "epoch": 0.3247422680412371,
-      "grad_norm": 0.4284820854663849,
-      "learning_rate": 9.99507416928562e-05,
-      "loss": 0.7711,
-      "step": 63
-    },
-    {
-      "epoch": 0.32989690721649484,
-      "grad_norm": 0.3693944215774536,
-      "learning_rate": 9.994287352135825e-05,
-      "loss": 0.8896,
-      "step": 64
-    },
-    {
-      "epoch": 0.33505154639175255,
-      "grad_norm": 0.3671146631240845,
-      "learning_rate": 9.993442298227365e-05,
-      "loss": 0.8124,
-      "step": 65
-    },
-    {
-      "epoch": 0.3402061855670103,
-      "grad_norm": 0.4267442524433136,
-      "learning_rate": 9.99253901741414e-05,
-      "loss": 0.7468,
-      "step": 66
-    },
-    {
-      "epoch": 0.34536082474226804,
-      "grad_norm": 0.392914742231369,
-      "learning_rate": 9.991577520229014e-05,
-      "loss": 0.9241,
-      "step": 67
-    },
-    {
-      "epoch": 0.35051546391752575,
-      "grad_norm": 0.3600424528121948,
-      "learning_rate": 9.99055781788369e-05,
-      "loss": 0.7838,
-      "step": 68
-    },
-    {
-      "epoch": 0.3556701030927835,
-      "grad_norm": 0.38290566205978394,
-      "learning_rate": 9.989479922268588e-05,
-      "loss": 0.6526,
-      "step": 69
-    },
-    {
-      "epoch": 0.36082474226804123,
-      "grad_norm": 0.38870134949684143,
-      "learning_rate": 9.988343845952697e-05,
-      "loss": 0.8359,
-      "step": 70
-    },
-    {
-      "epoch": 0.36597938144329895,
-      "grad_norm": 0.4147457480430603,
-      "learning_rate": 9.98714960218343e-05,
-      "loss": 0.82,
-      "step": 71
-    },
-    {
-      "epoch": 0.3711340206185567,
-      "grad_norm": 0.3855450451374054,
-      "learning_rate": 9.985897204886481e-05,
-      "loss": 0.797,
-      "step": 72
-    },
-    {
-      "epoch": 0.37628865979381443,
-      "grad_norm": 0.4507073163986206,
-      "learning_rate": 9.98458666866564e-05,
-      "loss": 0.8844,
-      "step": 73
-    },
-    {
-      "epoch": 0.38144329896907214,
-      "grad_norm": 0.44775328040122986,
-      "learning_rate": 9.983218008802648e-05,
-      "loss": 0.8082,
-      "step": 74
-    },
-    {
-      "epoch": 0.3865979381443299,
-      "grad_norm": 0.41650065779685974,
-      "learning_rate": 9.981791241257e-05,
-      "loss": 0.8983,
-      "step": 75
-    },
-    {
-      "epoch": 0.3917525773195876,
-      "grad_norm": 0.3322182595729828,
-      "learning_rate": 9.98030638266577e-05,
-      "loss": 0.6726,
-      "step": 76
-    },
-    {
-      "epoch": 0.39690721649484534,
-      "grad_norm": 0.4027014374732971,
-      "learning_rate": 9.978763450343407e-05,
-      "loss": 0.9871,
-      "step": 77
-    },
-    {
-      "epoch": 0.4020618556701031,
-      "grad_norm": 0.45693254470825195,
-      "learning_rate": 9.977162462281544e-05,
-      "loss": 0.8502,
-      "step": 78
-    },
-    {
-      "epoch": 0.4072164948453608,
-      "grad_norm": 0.3622555136680603,
-      "learning_rate": 9.975503437148783e-05,
-      "loss": 0.7157,
-      "step": 79
-    },
-    {
-      "epoch": 0.41237113402061853,
-      "grad_norm": 0.4588169455528259,
-      "learning_rate": 9.973786394290474e-05,
-      "loss": 0.898,
-      "step": 80
-    },
-    {
-      "epoch": 0.4175257731958763,
-      "grad_norm": 0.3862941265106201,
-      "learning_rate": 9.972011353728496e-05,
-      "loss": 0.6313,
-      "step": 81
-    },
-    {
-      "epoch": 0.422680412371134,
-      "grad_norm": 0.421906441450119,
-      "learning_rate": 9.970178336161018e-05,
-      "loss": 0.8687,
-      "step": 82
-    },
-    {
-      "epoch": 0.42783505154639173,
-      "grad_norm": 0.43780192732810974,
-      "learning_rate": 9.968287362962264e-05,
-      "loss": 0.8284,
-      "step": 83
-    },
-    {
-      "epoch": 0.4329896907216495,
-      "grad_norm": 0.3814310133457184,
-      "learning_rate": 9.96633845618225e-05,
-      "loss": 0.6429,
-      "step": 84
-    },
-    {
-      "epoch": 0.4381443298969072,
-      "grad_norm": 0.38887372612953186,
-      "learning_rate": 9.96433163854655e-05,
-      "loss": 0.8692,
-      "step": 85
-    },
-    {
-      "epoch": 0.44329896907216493,
-      "grad_norm": 0.42212334275245667,
-      "learning_rate": 9.962266933456008e-05,
-      "loss": 0.8881,
-      "step": 86
-    },
-    {
-      "epoch": 0.4484536082474227,
-      "grad_norm": 0.4047574996948242,
-      "learning_rate": 9.96014436498648e-05,
-      "loss": 0.7942,
-      "step": 87
-    },
-    {
-      "epoch": 0.4536082474226804,
-      "grad_norm": 0.3563002347946167,
-      "learning_rate": 9.957963957888542e-05,
-      "loss": 0.702,
-      "step": 88
-    },
-    {
-      "epoch": 0.4587628865979381,
-      "grad_norm": 0.383470743894577,
-      "learning_rate": 9.955725737587214e-05,
-      "loss": 0.7593,
-      "step": 89
-    },
-    {
-      "epoch": 0.4639175257731959,
-      "grad_norm": 0.4369049668312073,
-      "learning_rate": 9.953429730181653e-05,
-      "loss": 0.8449,
-      "step": 90
-    },
-    {
-      "epoch": 0.4690721649484536,
-      "grad_norm": 0.5086058378219604,
-      "learning_rate": 9.951075962444856e-05,
-      "loss": 0.9816,
-      "step": 91
-    },
-    {
-      "epoch": 0.4742268041237113,
-      "grad_norm": 0.4870317578315735,
-      "learning_rate": 9.94866446182334e-05,
-      "loss": 0.9004,
-      "step": 92
-    },
-    {
-      "epoch": 0.4793814432989691,
-      "grad_norm": 0.4313347339630127,
-      "learning_rate": 9.94619525643683e-05,
-      "loss": 0.9499,
-      "step": 93
-    },
-    {
-      "epoch": 0.4845360824742268,
-      "grad_norm": 0.3745776116847992,
-      "learning_rate": 9.943668375077925e-05,
-      "loss": 0.819,
-      "step": 94
-    },
-    {
-      "epoch": 0.4896907216494845,
-      "grad_norm": 0.3991529643535614,
-      "learning_rate": 9.941083847211765e-05,
-      "loss": 0.7799,
-      "step": 95
-    },
-    {
-      "epoch": 0.4948453608247423,
-      "grad_norm": 0.4022909998893738,
-      "learning_rate": 9.938441702975689e-05,
-      "loss": 0.8125,
-      "step": 96
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 0.3985065221786499,
-      "learning_rate": 9.93574197317888e-05,
-      "loss": 0.7959,
-      "step": 97
-    },
-    {
-      "epoch": 0.5,
-      "eval_loss": 0.802886426448822,
-      "eval_runtime": 23.123,
-      "eval_samples_per_second": 7.092,
-      "eval_steps_per_second": 1.773,
-      "step": 97
-    },
-    {
-      "epoch": 0.5051546391752577,
-      "grad_norm": 0.5668719410896301,
-      "learning_rate": 9.93298468930201e-05,
-      "loss": 0.9617,
-      "step": 98
-    },
-    {
-      "epoch": 0.5103092783505154,
-      "grad_norm": 0.3678329885005951,
-      "learning_rate": 9.930169883496867e-05,
-      "loss": 0.7494,
-      "step": 99
-    },
-    {
-      "epoch": 0.5154639175257731,
-      "grad_norm": 0.4211786985397339,
-      "learning_rate": 9.927297588585984e-05,
-      "loss": 0.6733,
-      "step": 100
-    },
-    {
-      "epoch": 0.520618556701031,
-      "grad_norm": 0.4105909764766693,
-      "learning_rate": 9.924367838062259e-05,
-      "loss": 0.6879,
-      "step": 101
-    },
-    {
-      "epoch": 0.5257731958762887,
-      "grad_norm": 0.4837935268878937,
-      "learning_rate": 9.921380666088558e-05,
-      "loss": 0.8099,
-      "step": 102
-    },
-    {
-      "epoch": 0.5309278350515464,
-      "grad_norm": 0.3713263273239136,
-      "learning_rate": 9.91833610749732e-05,
-      "loss": 0.8607,
-      "step": 103
-    },
-    {
-      "epoch": 0.5360824742268041,
-      "grad_norm": 0.3924923241138458,
-      "learning_rate": 9.915234197790152e-05,
-      "loss": 0.7541,
-      "step": 104
-    },
-    {
-      "epoch": 0.5412371134020618,
-      "grad_norm": 0.4768580198287964,
-      "learning_rate": 9.912074973137412e-05,
-      "loss": 0.9051,
-      "step": 105
-    },
-    {
-      "epoch": 0.5463917525773195,
-      "grad_norm": 0.42937901616096497,
-      "learning_rate": 9.908858470377793e-05,
-      "loss": 0.7494,
-      "step": 106
-    },
-    {
-      "epoch": 0.5515463917525774,
-      "grad_norm": 0.391781747341156,
-      "learning_rate": 9.905584727017884e-05,
-      "loss": 0.8718,
-      "step": 107
-    },
-    {
-      "epoch": 0.5567010309278351,
-      "grad_norm": 0.35565879940986633,
-      "learning_rate": 9.90225378123174e-05,
-      "loss": 0.7339,
-      "step": 108
-    },
-    {
-      "epoch": 0.5618556701030928,
-      "grad_norm": 0.33125853538513184,
-      "learning_rate": 9.898865671860438e-05,
-      "loss": 0.7042,
-      "step": 109
-    },
-    {
-      "epoch": 0.5670103092783505,
-      "grad_norm": 0.511025607585907,
-      "learning_rate": 9.895420438411616e-05,
-      "loss": 0.7971,
-      "step": 110
-    },
-    {
-      "epoch": 0.5721649484536082,
-      "grad_norm": 0.41539090871810913,
-      "learning_rate": 9.891918121059019e-05,
-      "loss": 0.8643,
-      "step": 111
-    },
-    {
-      "epoch": 0.5773195876288659,
-      "grad_norm": 0.371518611907959,
-      "learning_rate": 9.888358760642029e-05,
-      "loss": 0.6514,
-      "step": 112
-    },
-    {
-      "epoch": 0.5824742268041238,
-      "grad_norm": 0.5038712620735168,
-      "learning_rate": 9.884742398665191e-05,
-      "loss": 0.8113,
-      "step": 113
-    },
-    {
-      "epoch": 0.5876288659793815,
-      "grad_norm": 0.3781215250492096,
-      "learning_rate": 9.881069077297723e-05,
-      "loss": 0.702,
-      "step": 114
-    },
-    {
-      "epoch": 0.5927835051546392,
-      "grad_norm": 0.4231729805469513,
-      "learning_rate": 9.877338839373032e-05,
-      "loss": 0.7659,
-      "step": 115
-    },
-    {
-      "epoch": 0.5979381443298969,
-      "grad_norm": 0.4518745541572571,
-      "learning_rate": 9.873551728388203e-05,
-      "loss": 0.7877,
-      "step": 116
-    },
-    {
-      "epoch": 0.6030927835051546,
-      "grad_norm": 0.4570085406303406,
-      "learning_rate": 9.869707788503508e-05,
-      "loss": 0.8543,
-      "step": 117
-    },
-    {
-      "epoch": 0.6082474226804123,
-      "grad_norm": 0.38847261667251587,
-      "learning_rate": 9.865807064541877e-05,
-      "loss": 0.787,
-      "step": 118
-    },
-    {
-      "epoch": 0.6134020618556701,
-      "grad_norm": 0.3872878849506378,
-      "learning_rate": 9.861849601988383e-05,
-      "loss": 0.7859,
-      "step": 119
-    },
-    {
-      "epoch": 0.6185567010309279,
-      "grad_norm": 0.4313697814941406,
-      "learning_rate": 9.857835446989707e-05,
-      "loss": 0.7365,
-      "step": 120
-    },
-    {
-      "epoch": 0.6237113402061856,
-      "grad_norm": 0.4301742911338806,
-      "learning_rate": 9.853764646353605e-05,
-      "loss": 0.9206,
-      "step": 121
-    },
-    {
-      "epoch": 0.6288659793814433,
-      "grad_norm": 0.3453035354614258,
-      "learning_rate": 9.849637247548356e-05,
-      "loss": 0.6391,
-      "step": 122
-    },
-    {
-      "epoch": 0.634020618556701,
-      "grad_norm": 0.49422284960746765,
-      "learning_rate": 9.845453298702216e-05,
-      "loss": 0.8156,
-      "step": 123
-    },
-    {
-      "epoch": 0.6391752577319587,
-      "grad_norm": 0.37356287240982056,
-      "learning_rate": 9.841212848602846e-05,
-      "loss": 0.688,
-      "step": 124
-    },
-    {
-      "epoch": 0.6443298969072165,
-      "grad_norm": 0.47690337896347046,
-      "learning_rate": 9.836915946696759e-05,
-      "loss": 0.741,
-      "step": 125
-    },
-    {
-      "epoch": 0.6494845360824743,
-      "grad_norm": 0.5058567523956299,
-      "learning_rate": 9.832562643088724e-05,
-      "loss": 0.8239,
-      "step": 126
-    },
-    {
-      "epoch": 0.654639175257732,
-      "grad_norm": 0.36732959747314453,
-      "learning_rate": 9.828152988541201e-05,
-      "loss": 0.6655,
-      "step": 127
-    },
-    {
-      "epoch": 0.6597938144329897,
-      "grad_norm": 0.3805343508720398,
-      "learning_rate": 9.823687034473735e-05,
-      "loss": 0.7894,
-      "step": 128
-    },
-    {
-      "epoch": 0.6649484536082474,
-      "grad_norm": 0.36144956946372986,
-      "learning_rate": 9.81916483296236e-05,
-      "loss": 0.553,
-      "step": 129
-    },
-    {
-      "epoch": 0.6701030927835051,
-      "grad_norm": 0.38662460446357727,
-      "learning_rate": 9.814586436738998e-05,
-      "loss": 0.8187,
-      "step": 130
-    },
-    {
-      "epoch": 0.6752577319587629,
-      "grad_norm": 0.4534616470336914,
-      "learning_rate": 9.809951899190835e-05,
-      "loss": 0.9291,
-      "step": 131
-    },
-    {
-      "epoch": 0.6804123711340206,
-      "grad_norm": 0.42224985361099243,
-      "learning_rate": 9.805261274359705e-05,
-      "loss": 0.8359,
-      "step": 132
-    },
-    {
-      "epoch": 0.6855670103092784,
-      "grad_norm": 0.3735229969024658,
-      "learning_rate": 9.800514616941457e-05,
-      "loss": 0.8374,
-      "step": 133
-    },
-    {
-      "epoch": 0.6907216494845361,
-      "grad_norm": 0.3265180289745331,
-      "learning_rate": 9.795711982285316e-05,
-      "loss": 0.5942,
-      "step": 134
-    },
-    {
-      "epoch": 0.6958762886597938,
-      "grad_norm": 0.41515135765075684,
-      "learning_rate": 9.790853426393245e-05,
-      "loss": 0.7797,
-      "step": 135
-    },
-    {
-      "epoch": 0.7010309278350515,
-      "grad_norm": 0.41013866662979126,
-      "learning_rate": 9.785939005919278e-05,
-      "loss": 0.7919,
-      "step": 136
-    },
-    {
-      "epoch": 0.7061855670103093,
-      "grad_norm": 0.3023189902305603,
-      "learning_rate": 9.780968778168874e-05,
-      "loss": 0.6237,
-      "step": 137
-    },
-    {
-      "epoch": 0.711340206185567,
-      "grad_norm": 0.5121697187423706,
-      "learning_rate": 9.77594280109824e-05,
-      "loss": 0.834,
-      "step": 138
-    },
-    {
-      "epoch": 0.7164948453608248,
-      "grad_norm": 0.3422345519065857,
-      "learning_rate": 9.77086113331366e-05,
-      "loss": 0.7347,
-      "step": 139
-    },
-    {
-      "epoch": 0.7216494845360825,
-      "grad_norm": 0.4234163165092468,
-      "learning_rate": 9.765723834070804e-05,
-      "loss": 0.6087,
-      "step": 140
-    },
-    {
-      "epoch": 0.7268041237113402,
-      "grad_norm": 0.4280671775341034,
-      "learning_rate": 9.760530963274048e-05,
-      "loss": 0.6604,
-      "step": 141
-    },
-    {
-      "epoch": 0.7319587628865979,
-      "grad_norm": 0.4124312698841095,
-      "learning_rate": 9.755282581475769e-05,
-      "loss": 0.7191,
-      "step": 142
-    },
-    {
-      "epoch": 0.7371134020618557,
-      "grad_norm": 0.42981845140457153,
-      "learning_rate": 9.749978749875635e-05,
-      "loss": 0.7464,
-      "step": 143
-    },
-    {
-      "epoch": 0.7422680412371134,
-      "grad_norm": 0.44360974431037903,
-      "learning_rate": 9.744619530319899e-05,
-      "loss": 0.7628,
-      "step": 144
-    },
-    {
-      "epoch": 0.7474226804123711,
-      "grad_norm": 0.43546900153160095,
-      "learning_rate": 9.739204985300679e-05,
-      "loss": 0.8613,
-      "step": 145
-    },
-    {
-      "epoch": 0.7525773195876289,
-      "grad_norm": 0.3809047043323517,
-      "learning_rate": 9.733735177955219e-05,
-      "loss": 0.7123,
-      "step": 146
-    },
-    {
-      "epoch": 0.7577319587628866,
-      "grad_norm": 0.48975661396980286,
-      "learning_rate": 9.728210172065162e-05,
-      "loss": 0.9697,
-      "step": 147
-    },
-    {
-      "epoch": 0.7628865979381443,
-      "grad_norm": 0.4534354507923126,
-      "learning_rate": 9.722630032055803e-05,
-      "loss": 0.8023,
-      "step": 148
-    },
-    {
-      "epoch": 0.7680412371134021,
-      "grad_norm": 0.45175039768218994,
-      "learning_rate": 9.716994822995338e-05,
-      "loss": 0.9383,
-      "step": 149
-    },
-    {
-      "epoch": 0.7731958762886598,
-      "grad_norm": 0.35632631182670593,
-      "learning_rate": 9.711304610594104e-05,
-      "loss": 0.6622,
-      "step": 150
-    },
-    {
-      "epoch": 0.7783505154639175,
-      "grad_norm": 0.41617223620414734,
-      "learning_rate": 9.705559461203815e-05,
-      "loss": 0.8882,
-      "step": 151
-    },
-    {
-      "epoch": 0.7835051546391752,
-      "grad_norm": 0.4077165126800537,
-      "learning_rate": 9.699759441816787e-05,
-      "loss": 0.7558,
-      "step": 152
-    },
-    {
-      "epoch": 0.788659793814433,
-      "grad_norm": 0.42368316650390625,
-      "learning_rate": 9.69390462006516e-05,
-      "loss": 0.7681,
-      "step": 153
-    },
-    {
-      "epoch": 0.7938144329896907,
-      "grad_norm": 0.3329184055328369,
-      "learning_rate": 9.687995064220102e-05,
-      "loss": 0.7748,
-      "step": 154
-    },
-    {
-      "epoch": 0.7989690721649485,
-      "grad_norm": 0.45332592725753784,
-      "learning_rate": 9.682030843191022e-05,
-      "loss": 0.8737,
-      "step": 155
-    },
-    {
-      "epoch": 0.8041237113402062,
-      "grad_norm": 0.4276467561721802,
-      "learning_rate": 9.676012026524755e-05,
-      "loss": 0.7728,
-      "step": 156
-    },
-    {
-      "epoch": 0.8092783505154639,
-      "grad_norm": 0.4873369038105011,
-      "learning_rate": 9.669938684404766e-05,
-      "loss": 0.965,
-      "step": 157
-    },
-    {
-      "epoch": 0.8144329896907216,
-      "grad_norm": 0.33252307772636414,
-      "learning_rate": 9.663810887650318e-05,
-      "loss": 0.6058,
-      "step": 158
-    },
-    {
-      "epoch": 0.8195876288659794,
-      "grad_norm": 0.397871196269989,
-      "learning_rate": 9.657628707715655e-05,
-      "loss": 0.6472,
-      "step": 159
-    },
-    {
-      "epoch": 0.8247422680412371,
-      "grad_norm": 0.46122172474861145,
-      "learning_rate": 9.651392216689165e-05,
-      "loss": 0.9086,
-      "step": 160
-    },
-    {
-      "epoch": 0.8298969072164949,
-      "grad_norm": 0.39958837628364563,
-      "learning_rate": 9.645101487292539e-05,
-      "loss": 0.7856,
-      "step": 161
-    },
-    {
-      "epoch": 0.8350515463917526,
-      "grad_norm": 0.4407312273979187,
-      "learning_rate": 9.638756592879922e-05,
-      "loss": 0.7794,
-      "step": 162
-    },
-    {
-      "epoch": 0.8402061855670103,
-      "grad_norm": 0.388991117477417,
-      "learning_rate": 9.632357607437065e-05,
-      "loss": 0.7394,
-      "step": 163
-    },
-    {
-      "epoch": 0.845360824742268,
-      "grad_norm": 0.3550252616405487,
-      "learning_rate": 9.625904605580452e-05,
-      "loss": 0.6471,
-      "step": 164
-    },
-    {
-      "epoch": 0.8505154639175257,
-      "grad_norm": 0.37362176179885864,
-      "learning_rate": 9.619397662556435e-05,
-      "loss": 0.7905,
-      "step": 165
-    },
-    {
-      "epoch": 0.8556701030927835,
-      "grad_norm": 0.4304026961326599,
-      "learning_rate": 9.612836854240358e-05,
-      "loss": 0.8113,
-      "step": 166
-    },
-    {
-      "epoch": 0.8608247422680413,
-      "grad_norm": 0.43434932827949524,
-      "learning_rate": 9.606222257135675e-05,
-      "loss": 0.7418,
-      "step": 167
-    },
-    {
-      "epoch": 0.865979381443299,
-      "grad_norm": 0.4242919683456421,
-      "learning_rate": 9.599553948373045e-05,
-      "loss": 0.5827,
-      "step": 168
-    },
-    {
-      "epoch": 0.8711340206185567,
-      "grad_norm": 0.4583244025707245,
-      "learning_rate": 9.592832005709448e-05,
-      "loss": 0.816,
-      "step": 169
-    },
-    {
-      "epoch": 0.8762886597938144,
-      "grad_norm": 0.42892760038375854,
-      "learning_rate": 9.586056507527266e-05,
-      "loss": 0.8334,
-      "step": 170
-    },
-    {
-      "epoch": 0.8814432989690721,
-      "grad_norm": 0.4265868663787842,
-      "learning_rate": 9.579227532833377e-05,
-      "loss": 0.8701,
-      "step": 171
-    },
-    {
-      "epoch": 0.8865979381443299,
-      "grad_norm": 0.49999022483825684,
-      "learning_rate": 9.572345161258235e-05,
-      "loss": 0.8467,
-      "step": 172
-    },
-    {
-      "epoch": 0.8917525773195877,
-      "grad_norm": 0.3935216963291168,
-      "learning_rate": 9.565409473054932e-05,
-      "loss": 0.6476,
-      "step": 173
-    },
-    {
-      "epoch": 0.8969072164948454,
-      "grad_norm": 0.3507629930973053,
-      "learning_rate": 9.558420549098268e-05,
-      "loss": 0.883,
-      "step": 174
-    },
-    {
-      "epoch": 0.9020618556701031,
-      "grad_norm": 0.453337162733078,
-      "learning_rate": 9.551378470883812e-05,
-      "loss": 0.753,
-      "step": 175
-    },
-    {
-      "epoch": 0.9072164948453608,
-      "grad_norm": 0.438959002494812,
-      "learning_rate": 9.544283320526943e-05,
-      "loss": 0.8533,
-      "step": 176
-    },
-    {
-      "epoch": 0.9123711340206185,
-      "grad_norm": 0.43038681149482727,
-      "learning_rate": 9.537135180761903e-05,
-      "loss": 0.8053,
-      "step": 177
-    },
-    {
-      "epoch": 0.9175257731958762,
-      "grad_norm": 0.5258646011352539,
-      "learning_rate": 9.52993413494082e-05,
-      "loss": 0.9231,
-      "step": 178
-    },
-    {
-      "epoch": 0.9226804123711341,
-      "grad_norm": 0.3199097216129303,
-      "learning_rate": 9.522680267032742e-05,
-      "loss": 0.6158,
-      "step": 179
-    },
-    {
-      "epoch": 0.9278350515463918,
-      "grad_norm": 0.4209578335285187,
-      "learning_rate": 9.515373661622664e-05,
-      "loss": 0.8393,
-      "step": 180
-    },
-    {
-      "epoch": 0.9329896907216495,
-      "grad_norm": 0.47767141461372375,
-      "learning_rate": 9.508014403910533e-05,
-      "loss": 0.8761,
-      "step": 181
-    },
-    {
-      "epoch": 0.9381443298969072,
-      "grad_norm": 0.386308878660202,
-      "learning_rate": 9.500602579710256e-05,
-      "loss": 0.8358,
-      "step": 182
-    },
-    {
-      "epoch": 0.9432989690721649,
-      "grad_norm": 0.41348996758461,
-      "learning_rate": 9.4931382754487e-05,
-      "loss": 0.7701,
-      "step": 183
-    },
-    {
-      "epoch": 0.9484536082474226,
-      "grad_norm": 0.37899458408355713,
-      "learning_rate": 9.485621578164689e-05,
-      "loss": 0.6727,
-      "step": 184
-    },
-    {
-      "epoch": 0.9536082474226805,
-      "grad_norm": 0.45421937108039856,
-      "learning_rate": 9.478052575507982e-05,
-      "loss": 0.7223,
-      "step": 185
-    },
-    {
-      "epoch": 0.9587628865979382,
-      "grad_norm": 0.3540864884853363,
-      "learning_rate": 9.470431355738257e-05,
-      "loss": 0.7068,
-      "step": 186
-    },
-    {
-      "epoch": 0.9639175257731959,
-      "grad_norm": 0.3731318712234497,
-      "learning_rate": 9.46275800772407e-05,
-      "loss": 0.8276,
-      "step": 187
-    },
-    {
-      "epoch": 0.9690721649484536,
-      "grad_norm": 0.3973620533943176,
-      "learning_rate": 9.45503262094184e-05,
-      "loss": 0.7225,
-      "step": 188
-    },
-    {
-      "epoch": 0.9742268041237113,
-      "grad_norm": 0.4032728672027588,
-      "learning_rate": 9.447255285474783e-05,
-      "loss": 0.7378,
-      "step": 189
-    },
-    {
-      "epoch": 0.979381443298969,
-      "grad_norm": 0.4338615834712982,
-      "learning_rate": 9.439426092011875e-05,
-      "loss": 0.9921,
-      "step": 190
-    },
-    {
-      "epoch": 0.9845360824742269,
-      "grad_norm": 0.38775742053985596,
-      "learning_rate": 9.431545131846797e-05,
-      "loss": 0.8594,
-      "step": 191
-    },
-    {
-      "epoch": 0.9896907216494846,
-      "grad_norm": 0.5204732418060303,
-      "learning_rate": 9.423612496876855e-05,
-      "loss": 0.7099,
-      "step": 192
-    },
-    {
-      "epoch": 0.9948453608247423,
-      "grad_norm": 0.3858681917190552,
-      "learning_rate": 9.415628279601923e-05,
-      "loss": 0.7499,
-      "step": 193
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.5360407829284668,
-      "learning_rate": 9.407592573123358e-05,
-      "loss": 0.747,
-      "step": 194
-    },
-    {
-      "epoch": 1.0,
-      "eval_loss": 0.7822192311286926,
-      "eval_runtime": 23.1976,
-      "eval_samples_per_second": 7.07,
-      "eval_steps_per_second": 1.767,
-      "step": 194
-    },
-    {
-      "epoch": 1.0051546391752577,
-      "grad_norm": 0.36987513303756714,
-      "learning_rate": 9.39950547114292e-05,
-      "loss": 0.8177,
-      "step": 195
-    },
-    {
-      "epoch": 1.0103092783505154,
-      "grad_norm": 0.38242754340171814,
-      "learning_rate": 9.39136706796167e-05,
-      "loss": 0.6535,
-      "step": 196
-    },
-    {
-      "epoch": 1.0154639175257731,
-      "grad_norm": 0.39804112911224365,
-      "learning_rate": 9.383177458478878e-05,
-      "loss": 0.7294,
-      "step": 197
-    },
-    {
-      "epoch": 1.0206185567010309,
-      "grad_norm": 0.4012157917022705,
-      "learning_rate": 9.374936738190914e-05,
-      "loss": 0.7149,
-      "step": 198
-    },
-    {
-      "epoch": 1.0257731958762886,
-      "grad_norm": 0.45552578568458557,
-      "learning_rate": 9.366645003190132e-05,
-      "loss": 0.7603,
-      "step": 199
-    },
-    {
-      "epoch": 1.0309278350515463,
-      "grad_norm": 0.4522833824157715,
-      "learning_rate": 9.358302350163757e-05,
-      "loss": 0.9213,
-      "step": 200
-    },
-    {
-      "epoch": 1.0360824742268042,
-      "grad_norm": 0.4032951295375824,
-      "learning_rate": 9.349908876392748e-05,
-      "loss": 0.6416,
-      "step": 201
-    },
-    {
-      "epoch": 1.041237113402062,
-      "grad_norm": 0.48034724593162537,
-      "learning_rate": 9.341464679750669e-05,
-      "loss": 0.7987,
-      "step": 202
-    },
-    {
-      "epoch": 1.0463917525773196,
-      "grad_norm": 0.4653926193714142,
-      "learning_rate": 9.33296985870255e-05,
-      "loss": 0.7531,
-      "step": 203
-    },
-    {
-      "epoch": 1.0515463917525774,
-      "grad_norm": 0.3766481280326843,
-      "learning_rate": 9.32442451230373e-05,
-      "loss": 0.6282,
-      "step": 204
-    },
-    {
-      "epoch": 1.056701030927835,
-      "grad_norm": 0.4457147717475891,
-      "learning_rate": 9.315828740198714e-05,
-      "loss": 0.8084,
-      "step": 205
-    },
-    {
-      "epoch": 1.0618556701030928,
-      "grad_norm": 0.3701726496219635,
-      "learning_rate": 9.30718264262e-05,
-      "loss": 0.8328,
-      "step": 206
-    },
-    {
-      "epoch": 1.0670103092783505,
-      "grad_norm": 0.3876953721046448,
-      "learning_rate": 9.298486320386919e-05,
-      "loss": 0.6424,
-      "step": 207
-    },
-    {
-      "epoch": 1.0721649484536082,
-      "grad_norm": 0.4125080704689026,
-      "learning_rate": 9.289739874904449e-05,
-      "loss": 0.6097,
-      "step": 208
-    },
-    {
-      "epoch": 1.077319587628866,
-      "grad_norm": 0.38787776231765747,
-      "learning_rate": 9.280943408162046e-05,
-      "loss": 0.6664,
-      "step": 209
-    },
-    {
-      "epoch": 1.0824742268041236,
-      "grad_norm": 0.44620412588119507,
-      "learning_rate": 9.272097022732443e-05,
-      "loss": 0.7523,
-      "step": 210
-    },
-    {
-      "epoch": 1.0876288659793814,
-      "grad_norm": 0.4375419616699219,
-      "learning_rate": 9.263200821770461e-05,
-      "loss": 0.7148,
-      "step": 211
-    },
-    {
-      "epoch": 1.0927835051546393,
-      "grad_norm": 0.5171458721160889,
-      "learning_rate": 9.254254909011804e-05,
-      "loss": 0.8076,
-      "step": 212
-    },
-    {
-      "epoch": 1.097938144329897,
-      "grad_norm": 0.4271126091480255,
-      "learning_rate": 9.245259388771845e-05,
-      "loss": 0.7198,
-      "step": 213
-    },
-    {
-      "epoch": 1.1030927835051547,
-      "grad_norm": 0.4279671311378479,
-      "learning_rate": 9.236214365944418e-05,
-      "loss": 0.7592,
-      "step": 214
-    },
-    {
-      "epoch": 1.1082474226804124,
-      "grad_norm": 0.3776322901248932,
-      "learning_rate": 9.22711994600059e-05,
-      "loss": 0.6371,
-      "step": 215
-    },
-    {
-      "epoch": 1.1134020618556701,
-      "grad_norm": 0.40393704175949097,
-      "learning_rate": 9.217976234987428e-05,
-      "loss": 0.7749,
-      "step": 216
-    },
-    {
-      "epoch": 1.1185567010309279,
-      "grad_norm": 0.3984552025794983,
-      "learning_rate": 9.208783339526773e-05,
-      "loss": 0.5648,
-      "step": 217
-    },
-    {
-      "epoch": 1.1237113402061856,
-      "grad_norm": 0.37290528416633606,
-      "learning_rate": 9.199541366813982e-05,
-      "loss": 0.7259,
-      "step": 218
-    },
-    {
-      "epoch": 1.1288659793814433,
-      "grad_norm": 0.4922865629196167,
-      "learning_rate": 9.190250424616693e-05,
-      "loss": 0.7142,
-      "step": 219
-    },
-    {
-      "epoch": 1.134020618556701,
-      "grad_norm": 0.45908308029174805,
-      "learning_rate": 9.180910621273555e-05,
-      "loss": 0.6767,
-      "step": 220
-    },
-    {
-      "epoch": 1.1391752577319587,
-      "grad_norm": 0.4036223888397217,
-      "learning_rate": 9.171522065692975e-05,
-      "loss": 0.7196,
-      "step": 221
-    },
-    {
-      "epoch": 1.1443298969072164,
-      "grad_norm": 0.4769727885723114,
-      "learning_rate": 9.162084867351842e-05,
-      "loss": 0.6734,
-      "step": 222
-    },
-    {
-      "epoch": 1.1494845360824741,
-      "grad_norm": 0.35914334654808044,
-      "learning_rate": 9.152599136294253e-05,
-      "loss": 0.5459,
-      "step": 223
-    },
-    {
-      "epoch": 1.1546391752577319,
-      "grad_norm": 0.49013835191726685,
-      "learning_rate": 9.14306498313023e-05,
-      "loss": 0.7468,
-      "step": 224
-    },
-    {
-      "epoch": 1.1597938144329896,
-      "grad_norm": 0.47017601132392883,
-      "learning_rate": 9.133482519034428e-05,
-      "loss": 0.6986,
-      "step": 225
-    },
-    {
-      "epoch": 1.1649484536082475,
-      "grad_norm": 0.43976277112960815,
-      "learning_rate": 9.123851855744843e-05,
-      "loss": 0.7512,
-      "step": 226
-    },
-    {
-      "epoch": 1.1701030927835052,
-      "grad_norm": 0.42843174934387207,
-      "learning_rate": 9.114173105561501e-05,
-      "loss": 0.7984,
-      "step": 227
-    },
-    {
-      "epoch": 1.175257731958763,
-      "grad_norm": 0.3842880129814148,
-      "learning_rate": 9.104446381345159e-05,
-      "loss": 0.7076,
-      "step": 228
-    },
-    {
-      "epoch": 1.1804123711340206,
-      "grad_norm": 0.39397862553596497,
-      "learning_rate": 9.094671796515978e-05,
-      "loss": 0.6256,
-      "step": 229
-    },
-    {
-      "epoch": 1.1855670103092784,
-      "grad_norm": 0.5267696976661682,
-      "learning_rate": 9.08484946505221e-05,
-      "loss": 0.7114,
-      "step": 230
-    },
-    {
-      "epoch": 1.190721649484536,
-      "grad_norm": 0.41206100583076477,
-      "learning_rate": 9.074979501488867e-05,
-      "loss": 0.5898,
-      "step": 231
-    },
-    {
-      "epoch": 1.1958762886597938,
-      "grad_norm": 0.4033416211605072,
-      "learning_rate": 9.065062020916377e-05,
-      "loss": 0.604,
-      "step": 232
-    },
-    {
-      "epoch": 1.2010309278350515,
-      "grad_norm": 0.43909236788749695,
-      "learning_rate": 9.055097138979252e-05,
-      "loss": 0.7177,
-      "step": 233
-    },
-    {
-      "epoch": 1.2061855670103092,
-      "grad_norm": 0.4308321475982666,
-      "learning_rate": 9.045084971874738e-05,
-      "loss": 0.6634,
-      "step": 234
-    },
-    {
-      "epoch": 1.211340206185567,
-      "grad_norm": 0.5138210654258728,
-      "learning_rate": 9.035025636351452e-05,
-      "loss": 0.6342,
-      "step": 235
-    },
-    {
-      "epoch": 1.2164948453608249,
-      "grad_norm": 0.44698837399482727,
-      "learning_rate": 9.024919249708035e-05,
-      "loss": 0.792,
-      "step": 236
-    },
-    {
-      "epoch": 1.2216494845360826,
-      "grad_norm": 0.5599440932273865,
-      "learning_rate": 9.014765929791768e-05,
-      "loss": 0.7513,
-      "step": 237
-    },
-    {
-      "epoch": 1.2268041237113403,
-      "grad_norm": 0.4447702467441559,
-      "learning_rate": 9.004565794997209e-05,
-      "loss": 0.6762,
-      "step": 238
-    },
-    {
-      "epoch": 1.231958762886598,
-      "grad_norm": 0.40888068079948425,
-      "learning_rate": 8.994318964264809e-05,
-      "loss": 0.6168,
-      "step": 239
-    },
-    {
-      "epoch": 1.2371134020618557,
-      "grad_norm": 0.48757925629615784,
-      "learning_rate": 8.984025557079523e-05,
-      "loss": 0.7511,
-      "step": 240
-    },
-    {
-      "epoch": 1.2422680412371134,
-      "grad_norm": 0.522798478603363,
-      "learning_rate": 8.973685693469423e-05,
-      "loss": 0.7837,
-      "step": 241
-    },
-    {
-      "epoch": 1.2474226804123711,
-      "grad_norm": 0.46510541439056396,
-      "learning_rate": 8.963299494004291e-05,
-      "loss": 0.7429,
-      "step": 242
-    },
-    {
-      "epoch": 1.2525773195876289,
-      "grad_norm": 0.3879891037940979,
-      "learning_rate": 8.952867079794218e-05,
-      "loss": 0.5545,
-      "step": 243
-    },
-    {
-      "epoch": 1.2577319587628866,
-      "grad_norm": 0.4209011197090149,
-      "learning_rate": 8.942388572488187e-05,
-      "loss": 0.9004,
-      "step": 244
-    },
-    {
-      "epoch": 1.2628865979381443,
-      "grad_norm": 0.462494820356369,
-      "learning_rate": 8.931864094272663e-05,
-      "loss": 0.5892,
-      "step": 245
-    },
-    {
-      "epoch": 1.268041237113402,
-      "grad_norm": 0.5652675032615662,
-      "learning_rate": 8.921293767870157e-05,
-      "loss": 0.7347,
-      "step": 246
-    },
-    {
-      "epoch": 1.2731958762886597,
-      "grad_norm": 0.45603305101394653,
-      "learning_rate": 8.910677716537806e-05,
-      "loss": 0.8051,
-      "step": 247
-    },
-    {
-      "epoch": 1.2783505154639174,
-      "grad_norm": 0.4945577085018158,
-      "learning_rate": 8.900016064065923e-05,
-      "loss": 0.7267,
-      "step": 248
-    },
-    {
-      "epoch": 1.2835051546391751,
-      "grad_norm": 0.5541755557060242,
-      "learning_rate": 8.889308934776572e-05,
-      "loss": 0.8397,
-      "step": 249
-    },
-    {
-      "epoch": 1.2886597938144329,
-      "grad_norm": 0.5028125643730164,
-      "learning_rate": 8.8785564535221e-05,
-      "loss": 0.8193,
-      "step": 250
-    },
-    {
-      "epoch": 1.2938144329896908,
-      "grad_norm": 0.48722466826438904,
-      "learning_rate": 8.867758745683687e-05,
-      "loss": 0.8063,
-      "step": 251
-    },
-    {
-      "epoch": 1.2989690721649485,
-      "grad_norm": 0.5241202116012573,
-      "learning_rate": 8.85691593716989e-05,
-      "loss": 0.8016,
-      "step": 252
-    },
-    {
-      "epoch": 1.3041237113402062,
-      "grad_norm": 0.5034804344177246,
-      "learning_rate": 8.84602815441517e-05,
-      "loss": 0.7144,
-      "step": 253
-    },
-    {
-      "epoch": 1.309278350515464,
-      "grad_norm": 0.4778522849082947,
-      "learning_rate": 8.835095524378414e-05,
-      "loss": 0.8033,
-      "step": 254
-    },
-    {
-      "epoch": 1.3144329896907216,
-      "grad_norm": 0.4429955780506134,
-      "learning_rate": 8.824118174541464e-05,
-      "loss": 0.6884,
-      "step": 255
-    },
-    {
-      "epoch": 1.3195876288659794,
-      "grad_norm": 0.5443170666694641,
-      "learning_rate": 8.81309623290762e-05,
-      "loss": 0.8507,
-      "step": 256
-    },
-    {
-      "epoch": 1.324742268041237,
-      "grad_norm": 0.47870996594429016,
-      "learning_rate": 8.802029828000156e-05,
-      "loss": 0.8128,
-      "step": 257
-    },
-    {
-      "epoch": 1.3298969072164948,
-      "grad_norm": 0.48450103402137756,
-      "learning_rate": 8.790919088860814e-05,
-      "loss": 0.7608,
-      "step": 258
-    },
-    {
-      "epoch": 1.3350515463917525,
-      "grad_norm": 0.5307578444480896,
-      "learning_rate": 8.779764145048308e-05,
-      "loss": 0.7209,
-      "step": 259
-    },
-    {
-      "epoch": 1.3402061855670104,
-      "grad_norm": 0.530427098274231,
-      "learning_rate": 8.768565126636806e-05,
-      "loss": 0.712,
-      "step": 260
-    },
-    {
-      "epoch": 1.3453608247422681,
-      "grad_norm": 0.48835986852645874,
-      "learning_rate": 8.757322164214413e-05,
-      "loss": 0.6929,
-      "step": 261
-    },
-    {
-      "epoch": 1.3505154639175259,
-      "grad_norm": 0.52711021900177,
-      "learning_rate": 8.746035388881655e-05,
-      "loss": 0.8256,
-      "step": 262
-    },
-    {
-      "epoch": 1.3556701030927836,
-      "grad_norm": 0.48932960629463196,
-      "learning_rate": 8.734704932249944e-05,
-      "loss": 0.9334,
-      "step": 263
-    },
-    {
-      "epoch": 1.3608247422680413,
-      "grad_norm": 0.5296959280967712,
-      "learning_rate": 8.723330926440045e-05,
-      "loss": 0.8442,
-      "step": 264
-    },
-    {
-      "epoch": 1.365979381443299,
-      "grad_norm": 0.4768982529640198,
-      "learning_rate": 8.711913504080534e-05,
-      "loss": 0.6466,
-      "step": 265
-    },
-    {
-      "epoch": 1.3711340206185567,
-      "grad_norm": 0.4151005744934082,
-      "learning_rate": 8.70045279830626e-05,
-      "loss": 0.6588,
-      "step": 266
-    },
-    {
-      "epoch": 1.3762886597938144,
-      "grad_norm": 0.4385485053062439,
-      "learning_rate": 8.688948942756778e-05,
-      "loss": 0.6547,
-      "step": 267
-    },
-    {
-      "epoch": 1.3814432989690721,
-      "grad_norm": 0.5252705812454224,
-      "learning_rate": 8.677402071574805e-05,
-      "loss": 0.6968,
-      "step": 268
-    },
-    {
-      "epoch": 1.3865979381443299,
-      "grad_norm": 0.5336004495620728,
-      "learning_rate": 8.665812319404643e-05,
-      "loss": 0.6702,
-      "step": 269
-    },
-    {
-      "epoch": 1.3917525773195876,
-      "grad_norm": 0.5621787905693054,
-      "learning_rate": 8.654179821390621e-05,
-      "loss": 0.8236,
-      "step": 270
-    },
-    {
-      "epoch": 1.3969072164948453,
-      "grad_norm": 0.3868958353996277,
-      "learning_rate": 8.642504713175508e-05,
-      "loss": 0.7168,
-      "step": 271
-    },
-    {
-      "epoch": 1.402061855670103,
-      "grad_norm": 0.5103661417961121,
-      "learning_rate": 8.630787130898943e-05,
-      "loss": 0.8211,
-      "step": 272
-    },
-    {
-      "epoch": 1.4072164948453607,
-      "grad_norm": 0.46545475721359253,
-      "learning_rate": 8.619027211195836e-05,
-      "loss": 0.7592,
-      "step": 273
-    },
-    {
-      "epoch": 1.4123711340206184,
-      "grad_norm": 0.4790486991405487,
-      "learning_rate": 8.607225091194779e-05,
-      "loss": 0.6835,
-      "step": 274
-    },
-    {
-      "epoch": 1.4175257731958764,
-      "grad_norm": 0.4155111610889435,
-      "learning_rate": 8.595380908516454e-05,
-      "loss": 0.5178,
-      "step": 275
-    },
-    {
-      "epoch": 1.422680412371134,
-      "grad_norm": 0.49269962310791016,
-      "learning_rate": 8.583494801272018e-05,
-      "loss": 0.7367,
-      "step": 276
-    },
-    {
-      "epoch": 1.4278350515463918,
-      "grad_norm": 0.4321769177913666,
-      "learning_rate": 8.571566908061497e-05,
-      "loss": 0.6714,
-      "step": 277
-    },
-    {
-      "epoch": 1.4329896907216495,
-      "grad_norm": 0.5094105005264282,
-      "learning_rate": 8.559597367972168e-05,
-      "loss": 0.7646,
-      "step": 278
-    },
-    {
-      "epoch": 1.4381443298969072,
-      "grad_norm": 0.5336481928825378,
-      "learning_rate": 8.547586320576945e-05,
-      "loss": 0.7603,
-      "step": 279
-    },
-    {
-      "epoch": 1.443298969072165,
-      "grad_norm": 0.46947407722473145,
-      "learning_rate": 8.535533905932738e-05,
-      "loss": 0.6531,
-      "step": 280
-    },
-    {
-      "epoch": 1.4484536082474226,
-      "grad_norm": 0.5053247213363647,
-      "learning_rate": 8.52344026457883e-05,
-      "loss": 0.753,
-      "step": 281
-    },
-    {
-      "epoch": 1.4536082474226804,
-      "grad_norm": 0.5084853172302246,
-      "learning_rate": 8.511305537535237e-05,
-      "loss": 0.6503,
-      "step": 282
-    },
-    {
-      "epoch": 1.458762886597938,
-      "grad_norm": 0.49093201756477356,
-      "learning_rate": 8.499129866301057e-05,
-      "loss": 0.689,
-      "step": 283
-    },
-    {
-      "epoch": 1.463917525773196,
-      "grad_norm": 0.4668569564819336,
-      "learning_rate": 8.48691339285283e-05,
-      "loss": 0.726,
-      "step": 284
-    },
-    {
-      "epoch": 1.4690721649484537,
-      "grad_norm": 0.5379082560539246,
-      "learning_rate": 8.474656259642873e-05,
-      "loss": 0.7205,
-      "step": 285
-    },
-    {
-      "epoch": 1.4742268041237114,
-      "grad_norm": 0.6103803515434265,
-      "learning_rate": 8.46235860959763e-05,
-      "loss": 0.8544,
-      "step": 286
-    },
-    {
-      "epoch": 1.4793814432989691,
-      "grad_norm": 0.5668134093284607,
-      "learning_rate": 8.450020586115987e-05,
-      "loss": 0.7428,
-      "step": 287
-    },
-    {
-      "epoch": 1.4845360824742269,
-      "grad_norm": 0.4947795271873474,
-      "learning_rate": 8.437642333067625e-05,
-      "loss": 0.7288,
-      "step": 288
-    },
-    {
-      "epoch": 1.4896907216494846,
-      "grad_norm": 0.683354377746582,
-      "learning_rate": 8.42522399479132e-05,
-      "loss": 0.7857,
-      "step": 289
-    },
-    {
-      "epoch": 1.4948453608247423,
-      "grad_norm": 0.42545896768569946,
-      "learning_rate": 8.412765716093272e-05,
-      "loss": 0.6486,
-      "step": 290
-    },
-    {
-      "epoch": 1.5,
-      "grad_norm": 0.5766370296478271,
-      "learning_rate": 8.40026764224541e-05,
-      "loss": 0.8488,
-      "step": 291
-    },
-    {
-      "epoch": 1.5,
-      "eval_loss": 0.7756565809249878,
-      "eval_runtime": 23.1361,
-      "eval_samples_per_second": 7.088,
-      "eval_steps_per_second": 1.772,
-      "step": 291
-    },
-    {
-      "epoch": 1.5051546391752577,
-      "grad_norm": 0.47217854857444763,
-      "learning_rate": 8.387729918983706e-05,
-      "loss": 0.5576,
-      "step": 292
-    },
-    {
-      "epoch": 1.5103092783505154,
-      "grad_norm": 0.5849595665931702,
-      "learning_rate": 8.375152692506468e-05,
-      "loss": 0.9562,
-      "step": 293
-    },
-    {
-      "epoch": 1.5154639175257731,
-      "grad_norm": 0.5576279163360596,
-      "learning_rate": 8.362536109472636e-05,
-      "loss": 0.8353,
-      "step": 294
-    },
-    {
-      "epoch": 1.5206185567010309,
-      "grad_norm": 0.5278394222259521,
-      "learning_rate": 8.349880317000082e-05,
-      "loss": 0.6668,
-      "step": 295
-    },
-    {
-      "epoch": 1.5257731958762886,
-      "grad_norm": 0.47538498044013977,
-      "learning_rate": 8.337185462663878e-05,
-      "loss": 0.6829,
-      "step": 296
-    },
-    {
-      "epoch": 1.5309278350515463,
-      "grad_norm": 0.5417651534080505,
-      "learning_rate": 8.32445169449459e-05,
-      "loss": 0.9411,
-      "step": 297
-    },
-    {
-      "epoch": 1.536082474226804,
-      "grad_norm": 0.516247034072876,
-      "learning_rate": 8.311679160976539e-05,
-      "loss": 0.8907,
-      "step": 298
-    },
-    {
-      "epoch": 1.5412371134020617,
-      "grad_norm": 0.5147285461425781,
-      "learning_rate": 8.29886801104608e-05,
-      "loss": 0.7747,
-      "step": 299
-    },
-    {
-      "epoch": 1.5463917525773194,
-      "grad_norm": 0.4982938766479492,
-      "learning_rate": 8.286018394089863e-05,
-      "loss": 0.7649,
-      "step": 300
-    },
-    {
-      "epoch": 1.5515463917525774,
-      "grad_norm": 0.5095208883285522,
-      "learning_rate": 8.273130459943086e-05,
-      "loss": 0.7414,
-      "step": 301
-    },
-    {
-      "epoch": 1.556701030927835,
-      "grad_norm": 0.6313266754150391,
-      "learning_rate": 8.260204358887754e-05,
-      "loss": 0.9327,
-      "step": 302
-    },
-    {
-      "epoch": 1.5618556701030928,
-      "grad_norm": 0.46112972497940063,
-      "learning_rate": 8.247240241650918e-05,
-      "loss": 0.6451,
-      "step": 303
-    },
-    {
-      "epoch": 1.5670103092783505,
-      "grad_norm": 0.5244841575622559,
-      "learning_rate": 8.234238259402935e-05,
-      "loss": 0.7223,
-      "step": 304
-    },
-    {
-      "epoch": 1.5721649484536082,
-      "grad_norm": 0.4885084927082062,
-      "learning_rate": 8.221198563755682e-05,
-      "loss": 0.6624,
-      "step": 305
-    },
-    {
-      "epoch": 1.577319587628866,
-      "grad_norm": 0.47980961203575134,
-      "learning_rate": 8.208121306760805e-05,
-      "loss": 0.7423,
-      "step": 306
-    },
-    {
-      "epoch": 1.5824742268041239,
-      "grad_norm": 0.4056549072265625,
-      "learning_rate": 8.195006640907942e-05,
-      "loss": 0.588,
-      "step": 307
-    },
-    {
-      "epoch": 1.5876288659793816,
-      "grad_norm": 0.5350937843322754,
-      "learning_rate": 8.181854719122939e-05,
-      "loss": 0.742,
-      "step": 308
-    },
-    {
-      "epoch": 1.5927835051546393,
-      "grad_norm": 0.555401086807251,
-      "learning_rate": 8.168665694766073e-05,
-      "loss": 0.8262,
-      "step": 309
-    },
-    {
-      "epoch": 1.597938144329897,
-      "grad_norm": 0.4927338659763336,
-      "learning_rate": 8.155439721630264e-05,
-      "loss": 0.7753,
-      "step": 310
-    },
-    {
-      "epoch": 1.6030927835051547,
-      "grad_norm": 0.4449557363986969,
-      "learning_rate": 8.142176953939279e-05,
-      "loss": 0.6069,
-      "step": 311
-    },
-    {
-      "epoch": 1.6082474226804124,
-      "grad_norm": 0.4355365037918091,
-      "learning_rate": 8.128877546345933e-05,
-      "loss": 0.5786,
-      "step": 312
-    },
-    {
-      "epoch": 1.6134020618556701,
-      "grad_norm": 0.5652406215667725,
-      "learning_rate": 8.115541653930286e-05,
-      "loss": 0.648,
-      "step": 313
-    },
-    {
-      "epoch": 1.6185567010309279,
-      "grad_norm": 0.4682961702346802,
-      "learning_rate": 8.102169432197842e-05,
-      "loss": 0.6993,
-      "step": 314
-    },
-    {
-      "epoch": 1.6237113402061856,
-      "grad_norm": 0.4309576451778412,
-      "learning_rate": 8.088761037077718e-05,
-      "loss": 0.843,
-      "step": 315
-    },
-    {
-      "epoch": 1.6288659793814433,
-      "grad_norm": 0.4121072292327881,
-      "learning_rate": 8.075316624920848e-05,
-      "loss": 0.569,
-      "step": 316
-    },
-    {
-      "epoch": 1.634020618556701,
-      "grad_norm": 0.5206684470176697,
-      "learning_rate": 8.061836352498145e-05,
-      "loss": 0.7025,
-      "step": 317
-    },
-    {
-      "epoch": 1.6391752577319587,
-      "grad_norm": 0.5165379643440247,
-      "learning_rate": 8.048320376998673e-05,
-      "loss": 0.6356,
-      "step": 318
-    },
-    {
-      "epoch": 1.6443298969072164,
-      "grad_norm": 0.5222402811050415,
-      "learning_rate": 8.034768856027826e-05,
-      "loss": 0.8059,
-      "step": 319
-    },
-    {
-      "epoch": 1.6494845360824741,
-      "grad_norm": 0.6080288887023926,
-      "learning_rate": 8.021181947605473e-05,
-      "loss": 0.7947,
-      "step": 320
-    },
-    {
-      "epoch": 1.6546391752577319,
-      "grad_norm": 0.47382211685180664,
-      "learning_rate": 8.007559810164133e-05,
-      "loss": 0.7139,
-      "step": 321
-    },
-    {
-      "epoch": 1.6597938144329896,
-      "grad_norm": 0.568452775478363,
-      "learning_rate": 7.993902602547113e-05,
-      "loss": 0.6434,
-      "step": 322
-    },
-    {
-      "epoch": 1.6649484536082473,
-      "grad_norm": 0.5808526277542114,
-      "learning_rate": 7.980210484006666e-05,
-      "loss": 0.7782,
-      "step": 323
-    },
-    {
-      "epoch": 1.670103092783505,
-      "grad_norm": 0.4880453050136566,
-      "learning_rate": 7.966483614202128e-05,
-      "loss": 0.7838,
-      "step": 324
-    },
-    {
-      "epoch": 1.675257731958763,
-      "grad_norm": 0.4492298662662506,
-      "learning_rate": 7.952722153198054e-05,
-      "loss": 0.5719,
-      "step": 325
-    },
-    {
-      "epoch": 1.6804123711340206,
-      "grad_norm": 0.5690252184867859,
-      "learning_rate": 7.938926261462366e-05,
-      "loss": 0.7023,
-      "step": 326
-    },
-    {
-      "epoch": 1.6855670103092784,
-      "grad_norm": 0.4502432644367218,
-      "learning_rate": 7.925096099864464e-05,
-      "loss": 0.5499,
-      "step": 327
-    },
-    {
-      "epoch": 1.690721649484536,
-      "grad_norm": 0.5307354927062988,
-      "learning_rate": 7.911231829673356e-05,
-      "loss": 0.7422,
-      "step": 328
-    },
-    {
-      "epoch": 1.6958762886597938,
-      "grad_norm": 0.45356300473213196,
-      "learning_rate": 7.897333612555785e-05,
-      "loss": 0.5429,
-      "step": 329
-    },
-    {
-      "epoch": 1.7010309278350515,
-      "grad_norm": 0.5033542513847351,
-      "learning_rate": 7.883401610574336e-05,
-      "loss": 0.7098,
-      "step": 330
-    },
-    {
-      "epoch": 1.7061855670103094,
-      "grad_norm": 0.5916339159011841,
-      "learning_rate": 7.869435986185547e-05,
-      "loss": 0.8832,
-      "step": 331
-    },
-    {
-      "epoch": 1.7113402061855671,
-      "grad_norm": 0.49946147203445435,
-      "learning_rate": 7.855436902238017e-05,
-      "loss": 0.8333,
-      "step": 332
-    },
-    {
-      "epoch": 1.7164948453608249,
-      "grad_norm": 0.4719221293926239,
-      "learning_rate": 7.841404521970505e-05,
-      "loss": 0.56,
-      "step": 333
-    },
-    {
-      "epoch": 1.7216494845360826,
-      "grad_norm": 0.4418396055698395,
-      "learning_rate": 7.82733900901003e-05,
-      "loss": 0.4716,
-      "step": 334
-    },
-    {
-      "epoch": 1.7268041237113403,
-      "grad_norm": 0.45411449670791626,
-      "learning_rate": 7.813240527369959e-05,
-      "loss": 0.6581,
-      "step": 335
-    },
-    {
-      "epoch": 1.731958762886598,
-      "grad_norm": 0.48833197355270386,
-      "learning_rate": 7.799109241448091e-05,
-      "loss": 0.652,
-      "step": 336
-    },
-    {
-      "epoch": 1.7371134020618557,
-      "grad_norm": 0.5444779396057129,
-      "learning_rate": 7.784945316024756e-05,
-      "loss": 0.7405,
-      "step": 337
-    },
-    {
-      "epoch": 1.7422680412371134,
-      "grad_norm": 0.4810863733291626,
-      "learning_rate": 7.770748916260875e-05,
-      "loss": 0.7171,
-      "step": 338
-    },
-    {
-      "epoch": 1.7474226804123711,
-      "grad_norm": 0.49916332960128784,
-      "learning_rate": 7.756520207696041e-05,
-      "loss": 0.7158,
-      "step": 339
-    },
-    {
-      "epoch": 1.7525773195876289,
-      "grad_norm": 0.46805036067962646,
-      "learning_rate": 7.742259356246593e-05,
-      "loss": 0.6331,
-      "step": 340
-    },
-    {
-      "epoch": 1.7577319587628866,
-      "grad_norm": 0.48011356592178345,
-      "learning_rate": 7.727966528203678e-05,
-      "loss": 0.591,
-      "step": 341
-    },
-    {
-      "epoch": 1.7628865979381443,
-      "grad_norm": 0.4629339277744293,
-      "learning_rate": 7.71364189023131e-05,
-      "loss": 0.7329,
-      "step": 342
-    },
-    {
-      "epoch": 1.768041237113402,
-      "grad_norm": 0.5280289649963379,
-      "learning_rate": 7.699285609364424e-05,
-      "loss": 0.711,
-      "step": 343
-    },
-    {
-      "epoch": 1.7731958762886597,
-      "grad_norm": 0.6236256957054138,
-      "learning_rate": 7.68489785300694e-05,
-      "loss": 0.826,
-      "step": 344
-    },
-    {
-      "epoch": 1.7783505154639174,
-      "grad_norm": 0.49681684374809265,
-      "learning_rate": 7.670478788929802e-05,
-      "loss": 0.8561,
-      "step": 345
-    },
-    {
-      "epoch": 1.7835051546391751,
-      "grad_norm": 0.4826050400733948,
-      "learning_rate": 7.656028585269018e-05,
-      "loss": 0.6096,
-      "step": 346
-    },
-    {
-      "epoch": 1.7886597938144329,
-      "grad_norm": 0.513830840587616,
-      "learning_rate": 7.641547410523709e-05,
-      "loss": 0.6318,
-      "step": 347
-    },
-    {
-      "epoch": 1.7938144329896906,
-      "grad_norm": 0.4502808153629303,
-      "learning_rate": 7.627035433554138e-05,
-      "loss": 0.696,
-      "step": 348
-    },
-    {
-      "epoch": 1.7989690721649485,
-      "grad_norm": 0.5793465375900269,
-      "learning_rate": 7.612492823579745e-05,
-      "loss": 0.6974,
-      "step": 349
-    },
-    {
-      "epoch": 1.8041237113402062,
-      "grad_norm": 0.4899560809135437,
-      "learning_rate": 7.597919750177168e-05,
-      "loss": 0.6863,
-      "step": 350
-    },
-    {
-      "epoch": 1.809278350515464,
-      "grad_norm": 0.527481198310852,
-      "learning_rate": 7.583316383278273e-05,
-      "loss": 0.6467,
-      "step": 351
-    },
-    {
-      "epoch": 1.8144329896907216,
-      "grad_norm": 0.5642676949501038,
-      "learning_rate": 7.568682893168164e-05,
-      "loss": 0.71,
-      "step": 352
-    },
-    {
-      "epoch": 1.8195876288659794,
-      "grad_norm": 0.46447527408599854,
-      "learning_rate": 7.554019450483208e-05,
-      "loss": 0.791,
-      "step": 353
-    },
-    {
-      "epoch": 1.824742268041237,
-      "grad_norm": 0.5443849563598633,
-      "learning_rate": 7.539326226209031e-05,
-      "loss": 0.6473,
-      "step": 354
-    },
-    {
-      "epoch": 1.829896907216495,
-      "grad_norm": 0.4533563554286957,
-      "learning_rate": 7.524603391678541e-05,
-      "loss": 0.8136,
-      "step": 355
-    },
-    {
-      "epoch": 1.8350515463917527,
-      "grad_norm": 0.5911179780960083,
-      "learning_rate": 7.509851118569915e-05,
-      "loss": 0.6774,
-      "step": 356
-    },
-    {
-      "epoch": 1.8402061855670104,
-      "grad_norm": 0.38583889603614807,
-      "learning_rate": 7.495069578904608e-05,
-      "loss": 0.5882,
-      "step": 357
-    },
-    {
-      "epoch": 1.8453608247422681,
-      "grad_norm": 0.492844820022583,
-      "learning_rate": 7.48025894504534e-05,
-      "loss": 0.6018,
-      "step": 358
-    },
-    {
-      "epoch": 1.8505154639175259,
-      "grad_norm": 0.5465859770774841,
-      "learning_rate": 7.465419389694092e-05,
-      "loss": 0.7754,
-      "step": 359
-    },
-    {
-      "epoch": 1.8556701030927836,
-      "grad_norm": 0.5601467490196228,
-      "learning_rate": 7.450551085890087e-05,
-      "loss": 0.8233,
-      "step": 360
-    },
-    {
-      "epoch": 1.8608247422680413,
-      "grad_norm": 0.4624144434928894,
-      "learning_rate": 7.435654207007773e-05,
-      "loss": 0.6372,
-      "step": 361
-    },
-    {
-      "epoch": 1.865979381443299,
-      "grad_norm": 0.5944560766220093,
-      "learning_rate": 7.420728926754803e-05,
-      "loss": 0.8245,
-      "step": 362
-    },
-    {
-      "epoch": 1.8711340206185567,
-      "grad_norm": 0.44903764128685,
-      "learning_rate": 7.405775419170014e-05,
-      "loss": 0.5653,
-      "step": 363
-    },
-    {
-      "epoch": 1.8762886597938144,
-      "grad_norm": 0.401335746049881,
-      "learning_rate": 7.390793858621386e-05,
-      "loss": 0.658,
-      "step": 364
-    },
-    {
-      "epoch": 1.8814432989690721,
-      "grad_norm": 0.5450597405433655,
-      "learning_rate": 7.375784419804019e-05,
-      "loss": 0.6493,
-      "step": 365
-    },
-    {
-      "epoch": 1.8865979381443299,
-      "grad_norm": 0.4412764608860016,
-      "learning_rate": 7.360747277738094e-05,
-      "loss": 0.6466,
-      "step": 366
-    },
-    {
-      "epoch": 1.8917525773195876,
-      "grad_norm": 0.503523588180542,
-      "learning_rate": 7.345682607766826e-05,
-      "loss": 0.8579,
-      "step": 367
-    },
-    {
-      "epoch": 1.8969072164948453,
-      "grad_norm": 0.48517486453056335,
-      "learning_rate": 7.330590585554428e-05,
-      "loss": 0.7231,
-      "step": 368
-    },
-    {
-      "epoch": 1.902061855670103,
-      "grad_norm": 0.39123615622520447,
-      "learning_rate": 7.315471387084056e-05,
-      "loss": 0.5414,
-      "step": 369
-    },
-    {
-      "epoch": 1.9072164948453607,
-      "grad_norm": 0.5755649209022522,
-      "learning_rate": 7.300325188655761e-05,
-      "loss": 0.8224,
-      "step": 370
-    },
-    {
-      "epoch": 1.9123711340206184,
-      "grad_norm": 0.5162664651870728,
-      "learning_rate": 7.285152166884432e-05,
-      "loss": 0.7323,
-      "step": 371
-    },
-    {
-      "epoch": 1.9175257731958761,
-      "grad_norm": 0.47386038303375244,
-      "learning_rate": 7.269952498697734e-05,
-      "loss": 0.7753,
-      "step": 372
-    },
-    {
-      "epoch": 1.922680412371134,
-      "grad_norm": 0.5395263433456421,
-      "learning_rate": 7.25472636133405e-05,
-      "loss": 0.623,
-      "step": 373
-    },
-    {
-      "epoch": 1.9278350515463918,
-      "grad_norm": 0.5538620352745056,
-      "learning_rate": 7.23947393234041e-05,
-      "loss": 0.8091,
-      "step": 374
-    },
-    {
-      "epoch": 1.9329896907216495,
-      "grad_norm": 0.615933358669281,
-      "learning_rate": 7.224195389570422e-05,
-      "loss": 0.7262,
-      "step": 375
-    },
-    {
-      "epoch": 1.9381443298969072,
-      "grad_norm": 0.6319516897201538,
-      "learning_rate": 7.208890911182197e-05,
-      "loss": 0.9344,
-      "step": 376
-    },
-    {
-      "epoch": 1.943298969072165,
-      "grad_norm": 0.5680528879165649,
-      "learning_rate": 7.193560675636277e-05,
-      "loss": 0.8644,
-      "step": 377
-    },
-    {
-      "epoch": 1.9484536082474226,
-      "grad_norm": 0.5286463499069214,
-      "learning_rate": 7.178204861693545e-05,
-      "loss": 0.7695,
-      "step": 378
-    },
-    {
-      "epoch": 1.9536082474226806,
-      "grad_norm": 0.43873798847198486,
-      "learning_rate": 7.162823648413151e-05,
-      "loss": 0.6283,
-      "step": 379
-    },
-    {
-      "epoch": 1.9587628865979383,
-      "grad_norm": 0.4513707458972931,
-      "learning_rate": 7.14741721515041e-05,
-      "loss": 0.6292,
-      "step": 380
-    },
-    {
-      "epoch": 1.963917525773196,
-      "grad_norm": 0.4606168568134308,
-      "learning_rate": 7.131985741554728e-05,
-      "loss": 0.6355,
-      "step": 381
-    },
-    {
-      "epoch": 1.9690721649484537,
-      "grad_norm": 0.5343955159187317,
-      "learning_rate": 7.116529407567489e-05,
-      "loss": 0.7579,
-      "step": 382
-    },
-    {
-      "epoch": 1.9742268041237114,
-      "grad_norm": 0.5064263343811035,
-      "learning_rate": 7.101048393419977e-05,
-      "loss": 0.6727,
-      "step": 383
-    },
-    {
-      "epoch": 1.9793814432989691,
-      "grad_norm": 0.4762507677078247,
-      "learning_rate": 7.085542879631253e-05,
-      "loss": 0.7425,
-      "step": 384
-    },
-    {
-      "epoch": 1.9845360824742269,
-      "grad_norm": 0.44575661420822144,
-      "learning_rate": 7.070013047006068e-05,
-      "loss": 0.6514,
-      "step": 385
-    },
-    {
-      "epoch": 1.9896907216494846,
-      "grad_norm": 0.44127196073532104,
-      "learning_rate": 7.054459076632743e-05,
-      "loss": 0.7233,
-      "step": 386
-    },
-    {
-      "epoch": 1.9948453608247423,
-      "grad_norm": 0.44311100244522095,
-      "learning_rate": 7.038881149881058e-05,
-      "loss": 0.672,
-      "step": 387
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 0.6681380867958069,
-      "learning_rate": 7.02327944840015e-05,
-      "loss": 0.7942,
-      "step": 388
-    },
-    {
-      "epoch": 2.0,
-      "eval_loss": 0.7649276256561279,
-      "eval_runtime": 23.0997,
-      "eval_samples_per_second": 7.1,
-      "eval_steps_per_second": 1.775,
-      "step": 388
-    },
-    {
-      "epoch": 2.0051546391752577,
-      "grad_norm": 0.4572388827800751,
-      "learning_rate": 7.007654154116377e-05,
-      "loss": 0.6377,
-      "step": 389
-    },
-    {
-      "epoch": 2.0103092783505154,
-      "grad_norm": 0.42942386865615845,
-      "learning_rate": 6.992005449231208e-05,
-      "loss": 0.6381,
-      "step": 390
-    },
-    {
-      "epoch": 2.015463917525773,
-      "grad_norm": 0.5012478828430176,
-      "learning_rate": 6.976333516219096e-05,
-      "loss": 0.6658,
-      "step": 391
-    },
-    {
-      "epoch": 2.020618556701031,
-      "grad_norm": 0.5113167762756348,
-      "learning_rate": 6.960638537825352e-05,
-      "loss": 0.5872,
-      "step": 392
-    },
-    {
-      "epoch": 2.0257731958762886,
-      "grad_norm": 0.5024546980857849,
-      "learning_rate": 6.944920697064004e-05,
-      "loss": 0.7016,
-      "step": 393
-    },
-    {
-      "epoch": 2.0309278350515463,
-      "grad_norm": 0.5500175952911377,
-      "learning_rate": 6.929180177215678e-05,
-      "loss": 0.6198,
-      "step": 394
-    },
-    {
-      "epoch": 2.036082474226804,
-      "grad_norm": 0.468348890542984,
-      "learning_rate": 6.91341716182545e-05,
-      "loss": 0.6204,
-      "step": 395
-    },
-    {
-      "epoch": 2.0412371134020617,
-      "grad_norm": 0.5198677778244019,
-      "learning_rate": 6.897631834700709e-05,
-      "loss": 0.5496,
-      "step": 396
-    },
-    {
-      "epoch": 2.0463917525773194,
-      "grad_norm": 0.5778740644454956,
-      "learning_rate": 6.881824379909017e-05,
-      "loss": 0.6795,
-      "step": 397
-    },
-    {
-      "epoch": 2.051546391752577,
-      "grad_norm": 0.44100233912467957,
-      "learning_rate": 6.865994981775957e-05,
-      "loss": 0.5305,
-      "step": 398
-    },
-    {
-      "epoch": 2.056701030927835,
-      "grad_norm": 0.5090631246566772,
-      "learning_rate": 6.850143824882986e-05,
-      "loss": 0.5958,
-      "step": 399
-    },
-    {
-      "epoch": 2.0618556701030926,
-      "grad_norm": 0.577616810798645,
-      "learning_rate": 6.834271094065283e-05,
-      "loss": 0.7075,
-      "step": 400
-    },
-    {
-      "epoch": 2.0670103092783507,
-      "grad_norm": 0.5909703373908997,
-      "learning_rate": 6.818376974409593e-05,
-      "loss": 0.629,
-      "step": 401
-    },
-    {
-      "epoch": 2.0721649484536084,
-      "grad_norm": 0.5738269686698914,
-      "learning_rate": 6.802461651252073e-05,
-      "loss": 0.7301,
-      "step": 402
-    },
-    {
-      "epoch": 2.077319587628866,
-      "grad_norm": 0.49154719710350037,
-      "learning_rate": 6.786525310176123e-05,
-      "loss": 0.6386,
-      "step": 403
-    },
-    {
-      "epoch": 2.082474226804124,
-      "grad_norm": 0.5145403742790222,
-      "learning_rate": 6.770568137010226e-05,
-      "loss": 0.6848,
-      "step": 404
-    },
-    {
-      "epoch": 2.0876288659793816,
-      "grad_norm": 0.496865838766098,
-      "learning_rate": 6.754590317825785e-05,
-      "loss": 0.605,
-      "step": 405
-    },
-    {
-      "epoch": 2.0927835051546393,
-      "grad_norm": 0.6585299372673035,
-      "learning_rate": 6.738592038934946e-05,
-      "loss": 0.7737,
-      "step": 406
-    },
-    {
-      "epoch": 2.097938144329897,
-      "grad_norm": 0.5088334679603577,
-      "learning_rate": 6.722573486888427e-05,
-      "loss": 0.6252,
-      "step": 407
-    },
-    {
-      "epoch": 2.1030927835051547,
-      "grad_norm": 0.5195786356925964,
-      "learning_rate": 6.706534848473352e-05,
-      "loss": 0.78,
-      "step": 408
-    },
-    {
-      "epoch": 2.1082474226804124,
-      "grad_norm": 0.6322928071022034,
-      "learning_rate": 6.69047631071106e-05,
-      "loss": 0.7107,
-      "step": 409
-    },
-    {
-      "epoch": 2.11340206185567,
-      "grad_norm": 0.5255240797996521,
-      "learning_rate": 6.674398060854931e-05,
-      "loss": 0.7664,
-      "step": 410
-    },
-    {
-      "epoch": 2.118556701030928,
-      "grad_norm": 0.5459422469139099,
-      "learning_rate": 6.658300286388203e-05,
-      "loss": 0.6449,
-      "step": 411
-    },
-    {
-      "epoch": 2.1237113402061856,
-      "grad_norm": 0.3948241174221039,
-      "learning_rate": 6.642183175021779e-05,
-      "loss": 0.5072,
-      "step": 412
-    },
-    {
-      "epoch": 2.1288659793814433,
-      "grad_norm": 0.6196532845497131,
-      "learning_rate": 6.62604691469205e-05,
-      "loss": 0.7733,
-      "step": 413
-    },
-    {
-      "epoch": 2.134020618556701,
-      "grad_norm": 0.6511134505271912,
-      "learning_rate": 6.609891693558692e-05,
-      "loss": 0.5305,
-      "step": 414
-    },
-    {
-      "epoch": 2.1391752577319587,
-      "grad_norm": 0.6522036194801331,
-      "learning_rate": 6.59371770000248e-05,
-      "loss": 0.7147,
-      "step": 415
-    },
-    {
-      "epoch": 2.1443298969072164,
-      "grad_norm": 0.532319188117981,
-      "learning_rate": 6.577525122623084e-05,
-      "loss": 0.6214,
-      "step": 416
-    },
-    {
-      "epoch": 2.149484536082474,
-      "grad_norm": 0.5139403343200684,
-      "learning_rate": 6.561314150236882e-05,
-      "loss": 0.551,
-      "step": 417
-    },
-    {
-      "epoch": 2.154639175257732,
-      "grad_norm": 0.6814868450164795,
-      "learning_rate": 6.545084971874738e-05,
-      "loss": 0.6843,
-      "step": 418
-    },
-    {
-      "epoch": 2.1597938144329896,
-      "grad_norm": 0.6147984862327576,
-      "learning_rate": 6.528837776779819e-05,
-      "loss": 0.5977,
-      "step": 419
-    },
-    {
-      "epoch": 2.1649484536082473,
-      "grad_norm": 0.5623170137405396,
-      "learning_rate": 6.51257275440538e-05,
-      "loss": 0.6759,
-      "step": 420
-    },
-    {
-      "epoch": 2.170103092783505,
-      "grad_norm": 0.491099089384079,
-      "learning_rate": 6.496290094412546e-05,
-      "loss": 0.5957,
-      "step": 421
-    },
-    {
-      "epoch": 2.1752577319587627,
-      "grad_norm": 0.6292416453361511,
-      "learning_rate": 6.479989986668118e-05,
-      "loss": 0.6852,
-      "step": 422
-    },
-    {
-      "epoch": 2.1804123711340204,
-      "grad_norm": 0.5607630610466003,
-      "learning_rate": 6.463672621242342e-05,
-      "loss": 0.5422,
-      "step": 423
-    },
-    {
-      "epoch": 2.1855670103092786,
-      "grad_norm": 0.5499788522720337,
-      "learning_rate": 6.447338188406704e-05,
-      "loss": 0.7426,
-      "step": 424
-    },
-    {
-      "epoch": 2.1907216494845363,
-      "grad_norm": 0.5566446781158447,
-      "learning_rate": 6.430986878631707e-05,
-      "loss": 0.518,
-      "step": 425
-    },
-    {
-      "epoch": 2.195876288659794,
-      "grad_norm": 0.5721368789672852,
-      "learning_rate": 6.41461888258465e-05,
-      "loss": 0.5814,
-      "step": 426
-    },
-    {
-      "epoch": 2.2010309278350517,
-      "grad_norm": 0.6252114176750183,
-      "learning_rate": 6.398234391127406e-05,
-      "loss": 0.8248,
-      "step": 427
-    },
-    {
-      "epoch": 2.2061855670103094,
-      "grad_norm": 0.5612536668777466,
-      "learning_rate": 6.381833595314195e-05,
-      "loss": 0.6407,
-      "step": 428
-    },
-    {
-      "epoch": 2.211340206185567,
-      "grad_norm": 0.6003265380859375,
-      "learning_rate": 6.365416686389358e-05,
-      "loss": 0.7016,
-      "step": 429
-    },
-    {
-      "epoch": 2.216494845360825,
-      "grad_norm": 0.4541587829589844,
-      "learning_rate": 6.348983855785121e-05,
-      "loss": 0.5283,
-      "step": 430
-    },
-    {
-      "epoch": 2.2216494845360826,
-      "grad_norm": 0.5990555882453918,
-      "learning_rate": 6.332535295119377e-05,
-      "loss": 0.6052,
-      "step": 431
-    },
-    {
-      "epoch": 2.2268041237113403,
-      "grad_norm": 0.5104555487632751,
-      "learning_rate": 6.31607119619343e-05,
-      "loss": 0.551,
-      "step": 432
-    },
-    {
-      "epoch": 2.231958762886598,
-      "grad_norm": 0.5985530018806458,
-      "learning_rate": 6.299591750989779e-05,
-      "loss": 0.6921,
-      "step": 433
-    },
-    {
-      "epoch": 2.2371134020618557,
-      "grad_norm": 0.5501850843429565,
-      "learning_rate": 6.283097151669869e-05,
-      "loss": 0.6221,
-      "step": 434
-    },
-    {
-      "epoch": 2.2422680412371134,
-      "grad_norm": 0.6366755366325378,
-      "learning_rate": 6.266587590571852e-05,
-      "loss": 0.8114,
-      "step": 435
-    },
-    {
-      "epoch": 2.247422680412371,
-      "grad_norm": 0.6541351675987244,
-      "learning_rate": 6.250063260208346e-05,
-      "loss": 0.7157,
-      "step": 436
-    },
-    {
-      "epoch": 2.252577319587629,
-      "grad_norm": 0.4696384370326996,
-      "learning_rate": 6.233524353264187e-05,
-      "loss": 0.5158,
-      "step": 437
-    },
-    {
-      "epoch": 2.2577319587628866,
-      "grad_norm": 0.5908606648445129,
-      "learning_rate": 6.216971062594179e-05,
-      "loss": 0.5988,
-      "step": 438
-    },
-    {
-      "epoch": 2.2628865979381443,
-      "grad_norm": 0.6051414608955383,
-      "learning_rate": 6.200403581220861e-05,
-      "loss": 0.5305,
-      "step": 439
-    },
-    {
-      "epoch": 2.268041237113402,
-      "grad_norm": 0.544257402420044,
-      "learning_rate": 6.183822102332234e-05,
-      "loss": 0.5912,
-      "step": 440
-    },
-    {
-      "epoch": 2.2731958762886597,
-      "grad_norm": 0.6186553239822388,
-      "learning_rate": 6.167226819279528e-05,
-      "loss": 0.7164,
-      "step": 441
-    },
-    {
-      "epoch": 2.2783505154639174,
-      "grad_norm": 0.6269963979721069,
-      "learning_rate": 6.150617925574933e-05,
-      "loss": 0.7621,
-      "step": 442
-    },
-    {
-      "epoch": 2.283505154639175,
-      "grad_norm": 0.5552319288253784,
-      "learning_rate": 6.13399561488935e-05,
-      "loss": 0.6343,
-      "step": 443
-    },
-    {
-      "epoch": 2.288659793814433,
-      "grad_norm": 0.5796776413917542,
-      "learning_rate": 6.117360081050136e-05,
-      "loss": 0.6439,
-      "step": 444
-    },
-    {
-      "epoch": 2.2938144329896906,
-      "grad_norm": 0.7522923946380615,
-      "learning_rate": 6.1007115180388285e-05,
-      "loss": 0.716,
-      "step": 445
-    },
-    {
-      "epoch": 2.2989690721649483,
-      "grad_norm": 0.6246132850646973,
-      "learning_rate": 6.0840501199889046e-05,
-      "loss": 0.5752,
-      "step": 446
-    },
-    {
-      "epoch": 2.304123711340206,
-      "grad_norm": 0.6404234766960144,
-      "learning_rate": 6.067376081183499e-05,
-      "loss": 0.6072,
-      "step": 447
-    },
-    {
-      "epoch": 2.3092783505154637,
-      "grad_norm": 0.5800334811210632,
-      "learning_rate": 6.050689596053151e-05,
-      "loss": 0.6807,
-      "step": 448
-    },
-    {
-      "epoch": 2.3144329896907214,
-      "grad_norm": 0.5738640427589417,
-      "learning_rate": 6.0339908591735296e-05,
-      "loss": 0.6723,
-      "step": 449
-    },
-    {
-      "epoch": 2.319587628865979,
-      "grad_norm": 0.5659841895103455,
-      "learning_rate": 6.01728006526317e-05,
-      "loss": 0.5924,
-      "step": 450
-    },
-    {
-      "epoch": 2.3247422680412373,
-      "grad_norm": 0.6528278589248657,
-      "learning_rate": 6.0005574091811964e-05,
-      "loss": 0.5736,
-      "step": 451
-    },
-    {
-      "epoch": 2.329896907216495,
-      "grad_norm": 0.5550704002380371,
-      "learning_rate": 5.9838230859250586e-05,
-      "loss": 0.5671,
-      "step": 452
-    },
-    {
-      "epoch": 2.3350515463917527,
-      "grad_norm": 0.5788085460662842,
-      "learning_rate": 5.967077290628249e-05,
-      "loss": 0.6304,
-      "step": 453
-    },
-    {
-      "epoch": 2.3402061855670104,
-      "grad_norm": 0.5700084567070007,
-      "learning_rate": 5.950320218558037e-05,
-      "loss": 0.6385,
-      "step": 454
-    },
-    {
-      "epoch": 2.345360824742268,
-      "grad_norm": 0.7324642539024353,
-      "learning_rate": 5.9335520651131814e-05,
-      "loss": 0.6061,
-      "step": 455
-    },
-    {
-      "epoch": 2.350515463917526,
-      "grad_norm": 0.5059091448783875,
-      "learning_rate": 5.9167730258216627e-05,
-      "loss": 0.5316,
-      "step": 456
-    },
-    {
-      "epoch": 2.3556701030927836,
-      "grad_norm": 0.678951621055603,
-      "learning_rate": 5.899983296338392e-05,
-      "loss": 0.6013,
-      "step": 457
-    },
-    {
-      "epoch": 2.3608247422680413,
-      "grad_norm": 0.6194437742233276,
-      "learning_rate": 5.8831830724429384e-05,
-      "loss": 0.5817,
-      "step": 458
-    },
-    {
-      "epoch": 2.365979381443299,
-      "grad_norm": 0.6131486892700195,
-      "learning_rate": 5.866372550037242e-05,
-      "loss": 0.5523,
-      "step": 459
-    },
-    {
-      "epoch": 2.3711340206185567,
-      "grad_norm": 0.5392377376556396,
-      "learning_rate": 5.849551925143334e-05,
-      "loss": 0.6512,
-      "step": 460
-    },
-    {
-      "epoch": 2.3762886597938144,
-      "grad_norm": 0.6425219774246216,
-      "learning_rate": 5.8327213939010414e-05,
-      "loss": 0.6352,
-      "step": 461
-    },
-    {
-      "epoch": 2.381443298969072,
-      "grad_norm": 0.5376563668251038,
-      "learning_rate": 5.815881152565712e-05,
-      "loss": 0.6266,
-      "step": 462
-    },
-    {
-      "epoch": 2.38659793814433,
-      "grad_norm": 0.6099573969841003,
-      "learning_rate": 5.799031397505913e-05,
-      "loss": 0.6225,
-      "step": 463
-    },
-    {
-      "epoch": 2.3917525773195876,
-      "grad_norm": 0.5236539244651794,
-      "learning_rate": 5.782172325201155e-05,
-      "loss": 0.6142,
-      "step": 464
-    },
-    {
-      "epoch": 2.3969072164948453,
-      "grad_norm": 0.5796985626220703,
-      "learning_rate": 5.7653041322395895e-05,
-      "loss": 0.5774,
-      "step": 465
-    },
-    {
-      "epoch": 2.402061855670103,
-      "grad_norm": 0.6779174208641052,
-      "learning_rate": 5.748427015315722e-05,
-      "loss": 0.7775,
-      "step": 466
-    },
-    {
-      "epoch": 2.4072164948453607,
-      "grad_norm": 0.5431979894638062,
-      "learning_rate": 5.7315411712281186e-05,
-      "loss": 0.5797,
-      "step": 467
-    },
-    {
-      "epoch": 2.4123711340206184,
-      "grad_norm": 0.5350040197372437,
-      "learning_rate": 5.714646796877108e-05,
-      "loss": 0.5288,
-      "step": 468
-    },
-    {
-      "epoch": 2.417525773195876,
-      "grad_norm": 0.5466572642326355,
-      "learning_rate": 5.697744089262491e-05,
-      "loss": 0.5502,
-      "step": 469
-    },
-    {
-      "epoch": 2.422680412371134,
-      "grad_norm": 0.6810032725334167,
-      "learning_rate": 5.680833245481234e-05,
-      "loss": 0.581,
-      "step": 470
-    },
-    {
-      "epoch": 2.4278350515463916,
-      "grad_norm": 0.6316937208175659,
-      "learning_rate": 5.6639144627251816e-05,
-      "loss": 0.7774,
-      "step": 471
-    },
-    {
-      "epoch": 2.4329896907216497,
-      "grad_norm": 0.5792400240898132,
-      "learning_rate": 5.646987938278753e-05,
-      "loss": 0.5606,
-      "step": 472
-    },
-    {
-      "epoch": 2.4381443298969074,
-      "grad_norm": 0.5614747405052185,
-      "learning_rate": 5.630053869516635e-05,
-      "loss": 0.5707,
-      "step": 473
-    },
-    {
-      "epoch": 2.443298969072165,
-      "grad_norm": 0.6040056347846985,
-      "learning_rate": 5.6131124539014926e-05,
-      "loss": 0.5709,
-      "step": 474
-    },
-    {
-      "epoch": 2.448453608247423,
-      "grad_norm": 0.5701522827148438,
-      "learning_rate": 5.596163888981656e-05,
-      "loss": 0.6838,
-      "step": 475
-    },
-    {
-      "epoch": 2.4536082474226806,
-      "grad_norm": 0.6044667363166809,
-      "learning_rate": 5.5792083723888225e-05,
-      "loss": 0.6885,
-      "step": 476
-    },
-    {
-      "epoch": 2.4587628865979383,
-      "grad_norm": 0.5631812214851379,
-      "learning_rate": 5.5622461018357486e-05,
-      "loss": 0.6584,
-      "step": 477
-    },
-    {
-      "epoch": 2.463917525773196,
-      "grad_norm": 0.5131292939186096,
-      "learning_rate": 5.5452772751139496e-05,
-      "loss": 0.4903,
-      "step": 478
-    },
-    {
-      "epoch": 2.4690721649484537,
-      "grad_norm": 0.7719168066978455,
-      "learning_rate": 5.5283020900913886e-05,
-      "loss": 0.673,
-      "step": 479
-    },
-    {
-      "epoch": 2.4742268041237114,
-      "grad_norm": 0.7306579947471619,
-      "learning_rate": 5.511320744710171e-05,
-      "loss": 0.9067,
-      "step": 480
-    },
-    {
-      "epoch": 2.479381443298969,
-      "grad_norm": 0.7012883424758911,
-      "learning_rate": 5.494333436984238e-05,
-      "loss": 0.6512,
-      "step": 481
-    },
-    {
-      "epoch": 2.484536082474227,
-      "grad_norm": 0.5695425868034363,
-      "learning_rate": 5.477340364997051e-05,
-      "loss": 0.6639,
-      "step": 482
-    },
-    {
-      "epoch": 2.4896907216494846,
-      "grad_norm": 0.561549723148346,
-      "learning_rate": 5.460341726899291e-05,
-      "loss": 0.6025,
-      "step": 483
-    },
-    {
-      "epoch": 2.4948453608247423,
-      "grad_norm": 0.6385191679000854,
-      "learning_rate": 5.4433377209065414e-05,
-      "loss": 0.6205,
-      "step": 484
-    },
-    {
-      "epoch": 2.5,
-      "grad_norm": 0.5434442758560181,
-      "learning_rate": 5.4263285452969806e-05,
-      "loss": 0.5898,
-      "step": 485
-    },
-    {
-      "epoch": 2.5,
-      "eval_loss": 0.7724881768226624,
-      "eval_runtime": 23.1178,
-      "eval_samples_per_second": 7.094,
-      "eval_steps_per_second": 1.774,
-      "step": 485
-    },
-    {
-      "epoch": 2.5051546391752577,
-      "grad_norm": 0.6231464147567749,
-      "learning_rate": 5.409314398409067e-05,
-      "loss": 0.5591,
-      "step": 486
-    },
-    {
-      "epoch": 2.5103092783505154,
-      "grad_norm": 0.5185215473175049,
-      "learning_rate": 5.392295478639225e-05,
-      "loss": 0.6953,
-      "step": 487
-    },
-    {
-      "epoch": 2.515463917525773,
-      "grad_norm": 0.6416254639625549,
-      "learning_rate": 5.3752719844395405e-05,
-      "loss": 0.7254,
-      "step": 488
-    },
-    {
-      "epoch": 2.520618556701031,
-      "grad_norm": 0.5942103266716003,
-      "learning_rate": 5.358244114315434e-05,
-      "loss": 0.629,
-      "step": 489
-    },
-    {
-      "epoch": 2.5257731958762886,
-      "grad_norm": 0.5755306482315063,
-      "learning_rate": 5.341212066823355e-05,
-      "loss": 0.6762,
-      "step": 490
-    },
-    {
-      "epoch": 2.5309278350515463,
-      "grad_norm": 0.6905494332313538,
-      "learning_rate": 5.324176040568465e-05,
-      "loss": 0.6028,
-      "step": 491
-    },
-    {
-      "epoch": 2.536082474226804,
-      "grad_norm": 0.6803380846977234,
-      "learning_rate": 5.307136234202318e-05,
-      "loss": 0.6943,
-      "step": 492
-    },
-    {
-      "epoch": 2.5412371134020617,
-      "grad_norm": 0.5133246779441833,
-      "learning_rate": 5.290092846420548e-05,
-      "loss": 0.6591,
-      "step": 493
-    },
-    {
-      "epoch": 2.5463917525773194,
-      "grad_norm": 0.5719724893569946,
-      "learning_rate": 5.27304607596055e-05,
-      "loss": 0.7672,
-      "step": 494
-    },
-    {
-      "epoch": 2.551546391752577,
-      "grad_norm": 0.5757763981819153,
-      "learning_rate": 5.255996121599167e-05,
-      "loss": 0.683,
-      "step": 495
-    },
-    {
-      "epoch": 2.556701030927835,
-      "grad_norm": 0.6311424374580383,
-      "learning_rate": 5.2389431821503606e-05,
-      "loss": 0.6142,
-      "step": 496
-    },
-    {
-      "epoch": 2.5618556701030926,
-      "grad_norm": 0.653990626335144,
-      "learning_rate": 5.221887456462907e-05,
-      "loss": 0.7109,
-      "step": 497
-    },
-    {
-      "epoch": 2.5670103092783503,
-      "grad_norm": 0.6488648653030396,
-      "learning_rate": 5.2048291434180716e-05,
-      "loss": 0.6484,
-      "step": 498
-    },
-    {
-      "epoch": 2.572164948453608,
-      "grad_norm": 0.6430789828300476,
-      "learning_rate": 5.1877684419272875e-05,
-      "loss": 0.6893,
-      "step": 499
-    },
-    {
-      "epoch": 2.5773195876288657,
-      "grad_norm": 0.701564610004425,
-      "learning_rate": 5.1707055509298396e-05,
-      "loss": 0.6976,
-      "step": 500
-    },
-    {
-      "epoch": 2.582474226804124,
-      "grad_norm": 0.6223208904266357,
-      "learning_rate": 5.153640669390546e-05,
-      "loss": 0.7153,
-      "step": 501
-    },
-    {
-      "epoch": 2.5876288659793816,
-      "grad_norm": 0.7090323567390442,
-      "learning_rate": 5.1365739962974304e-05,
-      "loss": 0.6964,
-      "step": 502
-    },
-    {
-      "epoch": 2.5927835051546393,
-      "grad_norm": 0.6134834885597229,
-      "learning_rate": 5.119505730659413e-05,
-      "loss": 0.5904,
-      "step": 503
-    },
-    {
-      "epoch": 2.597938144329897,
-      "grad_norm": 0.6040003895759583,
-      "learning_rate": 5.102436071503982e-05,
-      "loss": 0.5459,
-      "step": 504
-    },
-    {
-      "epoch": 2.6030927835051547,
-      "grad_norm": 0.6588661670684814,
-      "learning_rate": 5.0853652178748746e-05,
-      "loss": 0.5639,
-      "step": 505
-    },
-    {
-      "epoch": 2.6082474226804124,
-      "grad_norm": 0.6262742280960083,
-      "learning_rate": 5.068293368829755e-05,
-      "loss": 0.6609,
-      "step": 506
-    },
-    {
-      "epoch": 2.61340206185567,
-      "grad_norm": 0.6228057742118835,
-      "learning_rate": 5.0512207234379004e-05,
-      "loss": 0.5804,
-      "step": 507
-    },
-    {
-      "epoch": 2.618556701030928,
-      "grad_norm": 0.6566773056983948,
-      "learning_rate": 5.0341474807778663e-05,
-      "loss": 0.7437,
-      "step": 508
-    },
-    {
-      "epoch": 2.6237113402061856,
-      "grad_norm": 0.5632613301277161,
-      "learning_rate": 5.017073839935178e-05,
-      "loss": 0.6294,
-      "step": 509
-    },
-    {
-      "epoch": 2.6288659793814433,
-      "grad_norm": 0.7143430113792419,
-      "learning_rate": 5e-05,
-      "loss": 0.6383,
-      "step": 510
-    },
-    {
-      "epoch": 2.634020618556701,
-      "grad_norm": 0.5768882036209106,
-      "learning_rate": 4.982926160064823e-05,
-      "loss": 0.569,
-      "step": 511
-    },
-    {
-      "epoch": 2.6391752577319587,
-      "grad_norm": 0.6064475178718567,
-      "learning_rate": 4.965852519222134e-05,
-      "loss": 0.5433,
-      "step": 512
-    },
-    {
-      "epoch": 2.6443298969072164,
-      "grad_norm": 0.6445603966712952,
-      "learning_rate": 4.948779276562101e-05,
-      "loss": 0.623,
-      "step": 513
-    },
-    {
-      "epoch": 2.649484536082474,
-      "grad_norm": 0.6420986652374268,
-      "learning_rate": 4.9317066311702456e-05,
-      "loss": 0.7876,
-      "step": 514
-    },
-    {
-      "epoch": 2.654639175257732,
-      "grad_norm": 0.7606147527694702,
-      "learning_rate": 4.9146347821251266e-05,
-      "loss": 0.6335,
-      "step": 515
-    },
-    {
-      "epoch": 2.6597938144329896,
-      "grad_norm": 0.6954572200775146,
-      "learning_rate": 4.89756392849602e-05,
-      "loss": 0.7617,
-      "step": 516
-    },
-    {
-      "epoch": 2.6649484536082473,
-      "grad_norm": 0.7447741031646729,
-      "learning_rate": 4.880494269340588e-05,
-      "loss": 0.6967,
-      "step": 517
-    },
-    {
-      "epoch": 2.670103092783505,
-      "grad_norm": 0.553658127784729,
-      "learning_rate": 4.863426003702572e-05,
-      "loss": 0.6838,
-      "step": 518
-    },
-    {
-      "epoch": 2.675257731958763,
-      "grad_norm": 0.5830461978912354,
-      "learning_rate": 4.8463593306094555e-05,
-      "loss": 0.5642,
-      "step": 519
-    },
-    {
-      "epoch": 2.680412371134021,
-      "grad_norm": 0.4478597342967987,
-      "learning_rate": 4.829294449070161e-05,
-      "loss": 0.5929,
-      "step": 520
-    },
-    {
-      "epoch": 2.6855670103092786,
-      "grad_norm": 0.5738862752914429,
-      "learning_rate": 4.8122315580727136e-05,
-      "loss": 0.5355,
-      "step": 521
-    },
-    {
-      "epoch": 2.6907216494845363,
-      "grad_norm": 0.6429448127746582,
-      "learning_rate": 4.795170856581929e-05,
-      "loss": 0.7282,
-      "step": 522
-    },
-    {
-      "epoch": 2.695876288659794,
-      "grad_norm": 0.6432612538337708,
-      "learning_rate": 4.778112543537094e-05,
-      "loss": 0.7107,
-      "step": 523
-    },
-    {
-      "epoch": 2.7010309278350517,
-      "grad_norm": 0.6002117991447449,
-      "learning_rate": 4.7610568178496405e-05,
-      "loss": 0.5271,
-      "step": 524
-    },
-    {
-      "epoch": 2.7061855670103094,
-      "grad_norm": 0.6462201476097107,
-      "learning_rate": 4.744003878400835e-05,
-      "loss": 0.6236,
-      "step": 525
-    },
-    {
-      "epoch": 2.711340206185567,
-      "grad_norm": 0.6209389567375183,
-      "learning_rate": 4.726953924039451e-05,
-      "loss": 0.6871,
-      "step": 526
-    },
-    {
-      "epoch": 2.716494845360825,
-      "grad_norm": 0.5495082139968872,
-      "learning_rate": 4.709907153579454e-05,
-      "loss": 0.5909,
-      "step": 527
-    },
-    {
-      "epoch": 2.7216494845360826,
-      "grad_norm": 0.6679218411445618,
-      "learning_rate": 4.692863765797683e-05,
-      "loss": 0.6767,
-      "step": 528
-    },
-    {
-      "epoch": 2.7268041237113403,
-      "grad_norm": 0.5396760702133179,
-      "learning_rate": 4.675823959431535e-05,
-      "loss": 0.5304,
-      "step": 529
-    },
-    {
-      "epoch": 2.731958762886598,
-      "grad_norm": 0.6522016525268555,
-      "learning_rate": 4.658787933176646e-05,
-      "loss": 0.5896,
-      "step": 530
-    },
-    {
-      "epoch": 2.7371134020618557,
-      "grad_norm": 0.6318926811218262,
-      "learning_rate": 4.641755885684566e-05,
-      "loss": 0.8534,
-      "step": 531
-    },
-    {
-      "epoch": 2.7422680412371134,
-      "grad_norm": 0.6719669103622437,
-      "learning_rate": 4.624728015560461e-05,
-      "loss": 0.703,
-      "step": 532
-    },
-    {
-      "epoch": 2.747422680412371,
-      "grad_norm": 0.7230966091156006,
-      "learning_rate": 4.607704521360776e-05,
-      "loss": 0.7014,
-      "step": 533
-    },
-    {
-      "epoch": 2.752577319587629,
-      "grad_norm": 0.5768046379089355,
-      "learning_rate": 4.590685601590936e-05,
-      "loss": 0.5369,
-      "step": 534
-    },
-    {
-      "epoch": 2.7577319587628866,
-      "grad_norm": 0.7714840769767761,
-      "learning_rate": 4.57367145470302e-05,
-      "loss": 0.6511,
-      "step": 535
-    },
-    {
-      "epoch": 2.7628865979381443,
-      "grad_norm": 0.660317063331604,
-      "learning_rate": 4.5566622790934604e-05,
-      "loss": 0.6605,
-      "step": 536
-    },
-    {
-      "epoch": 2.768041237113402,
-      "grad_norm": 0.6877678632736206,
-      "learning_rate": 4.5396582731007095e-05,
-      "loss": 0.7078,
-      "step": 537
-    },
-    {
-      "epoch": 2.7731958762886597,
-      "grad_norm": 0.6704768538475037,
-      "learning_rate": 4.52265963500295e-05,
-      "loss": 0.6474,
-      "step": 538
-    },
-    {
-      "epoch": 2.7783505154639174,
-      "grad_norm": 0.7046148777008057,
-      "learning_rate": 4.505666563015763e-05,
-      "loss": 0.5439,
-      "step": 539
-    },
-    {
-      "epoch": 2.783505154639175,
-      "grad_norm": 0.6655017733573914,
-      "learning_rate": 4.4886792552898286e-05,
-      "loss": 0.8095,
-      "step": 540
-    },
-    {
-      "epoch": 2.788659793814433,
-      "grad_norm": 0.7060091495513916,
-      "learning_rate": 4.471697909908613e-05,
-      "loss": 0.66,
-      "step": 541
-    },
-    {
-      "epoch": 2.7938144329896906,
-      "grad_norm": 0.6830212473869324,
-      "learning_rate": 4.454722724886051e-05,
-      "loss": 0.7703,
-      "step": 542
-    },
-    {
-      "epoch": 2.7989690721649483,
-      "grad_norm": 0.7285298109054565,
-      "learning_rate": 4.437753898164254e-05,
-      "loss": 0.7986,
-      "step": 543
-    },
-    {
-      "epoch": 2.804123711340206,
-      "grad_norm": 0.6340286731719971,
-      "learning_rate": 4.420791627611179e-05,
-      "loss": 0.5987,
-      "step": 544
-    },
-    {
-      "epoch": 2.8092783505154637,
-      "grad_norm": 0.6459646224975586,
-      "learning_rate": 4.403836111018346e-05,
-      "loss": 0.7153,
-      "step": 545
-    },
-    {
-      "epoch": 2.8144329896907214,
-      "grad_norm": 0.5636230707168579,
-      "learning_rate": 4.3868875460985085e-05,
-      "loss": 0.5108,
-      "step": 546
-    },
-    {
-      "epoch": 2.819587628865979,
-      "grad_norm": 0.65835040807724,
-      "learning_rate": 4.369946130483364e-05,
-      "loss": 0.6884,
-      "step": 547
-    },
-    {
-      "epoch": 2.824742268041237,
-      "grad_norm": 0.7029833197593689,
-      "learning_rate": 4.353012061721249e-05,
-      "loss": 0.69,
-      "step": 548
-    },
-    {
-      "epoch": 2.829896907216495,
-      "grad_norm": 0.8330891728401184,
-      "learning_rate": 4.336085537274818e-05,
-      "loss": 0.7947,
-      "step": 549
-    },
-    {
-      "epoch": 2.8350515463917527,
-      "grad_norm": 0.5834147334098816,
-      "learning_rate": 4.319166754518768e-05,
-      "loss": 0.5213,
-      "step": 550
-    },
-    {
-      "epoch": 2.8402061855670104,
-      "grad_norm": 0.7061026692390442,
-      "learning_rate": 4.3022559107375106e-05,
-      "loss": 0.7841,
-      "step": 551
-    },
-    {
-      "epoch": 2.845360824742268,
-      "grad_norm": 0.5975331664085388,
-      "learning_rate": 4.285353203122893e-05,
-      "loss": 0.6662,
-      "step": 552
-    },
-    {
-      "epoch": 2.850515463917526,
-      "grad_norm": 0.6245263814926147,
-      "learning_rate": 4.268458828771883e-05,
-      "loss": 0.6272,
-      "step": 553
-    },
-    {
-      "epoch": 2.8556701030927836,
-      "grad_norm": 0.611504852771759,
-      "learning_rate": 4.251572984684281e-05,
-      "loss": 0.6917,
-      "step": 554
-    },
-    {
-      "epoch": 2.8608247422680413,
-      "grad_norm": 0.7087809443473816,
-      "learning_rate": 4.234695867760412e-05,
-      "loss": 0.6055,
-      "step": 555
-    },
-    {
-      "epoch": 2.865979381443299,
-      "grad_norm": 0.6241830587387085,
-      "learning_rate": 4.2178276747988446e-05,
-      "loss": 0.7149,
-      "step": 556
-    },
-    {
-      "epoch": 2.8711340206185567,
-      "grad_norm": 0.6535329222679138,
-      "learning_rate": 4.200968602494087e-05,
-      "loss": 0.628,
-      "step": 557
-    },
-    {
-      "epoch": 2.8762886597938144,
-      "grad_norm": 0.515200138092041,
-      "learning_rate": 4.18411884743429e-05,
-      "loss": 0.5361,
-      "step": 558
-    },
-    {
-      "epoch": 2.881443298969072,
-      "grad_norm": 0.7569003105163574,
-      "learning_rate": 4.16727860609896e-05,
-      "loss": 0.6474,
-      "step": 559
-    },
-    {
-      "epoch": 2.88659793814433,
-      "grad_norm": 0.6560523509979248,
-      "learning_rate": 4.150448074856667e-05,
-      "loss": 0.7475,
-      "step": 560
-    },
-    {
-      "epoch": 2.8917525773195876,
-      "grad_norm": 0.7731671333312988,
-      "learning_rate": 4.1336274499627596e-05,
-      "loss": 0.7221,
-      "step": 561
-    },
-    {
-      "epoch": 2.8969072164948453,
-      "grad_norm": 0.748393714427948,
-      "learning_rate": 4.1168169275570635e-05,
-      "loss": 0.6456,
-      "step": 562
-    },
-    {
-      "epoch": 2.902061855670103,
-      "grad_norm": 0.5793806910514832,
-      "learning_rate": 4.1000167036616113e-05,
-      "loss": 0.6022,
-      "step": 563
-    },
-    {
-      "epoch": 2.9072164948453607,
-      "grad_norm": 0.6097270250320435,
-      "learning_rate": 4.083226974178339e-05,
-      "loss": 0.6723,
-      "step": 564
-    },
-    {
-      "epoch": 2.9123711340206184,
-      "grad_norm": 0.7190105319023132,
-      "learning_rate": 4.066447934886819e-05,
-      "loss": 0.6841,
-      "step": 565
-    },
-    {
-      "epoch": 2.917525773195876,
-      "grad_norm": 0.781724750995636,
-      "learning_rate": 4.049679781441965e-05,
-      "loss": 0.6372,
-      "step": 566
-    },
-    {
-      "epoch": 2.9226804123711343,
-      "grad_norm": 0.5826302766799927,
-      "learning_rate": 4.0329227093717515e-05,
-      "loss": 0.617,
-      "step": 567
-    },
-    {
-      "epoch": 2.927835051546392,
-      "grad_norm": 0.5542965531349182,
-      "learning_rate": 4.016176914074944e-05,
-      "loss": 0.6842,
-      "step": 568
-    },
-    {
-      "epoch": 2.9329896907216497,
-      "grad_norm": 0.6216481328010559,
-      "learning_rate": 3.999442590818804e-05,
-      "loss": 0.7328,
-      "step": 569
-    },
-    {
-      "epoch": 2.9381443298969074,
-      "grad_norm": 0.7261362075805664,
-      "learning_rate": 3.982719934736832e-05,
-      "loss": 0.7162,
-      "step": 570
-    },
-    {
-      "epoch": 2.943298969072165,
-      "grad_norm": 0.8381164073944092,
-      "learning_rate": 3.9660091408264716e-05,
-      "loss": 0.6958,
-      "step": 571
-    },
-    {
-      "epoch": 2.948453608247423,
-      "grad_norm": 0.5935752391815186,
-      "learning_rate": 3.949310403946849e-05,
-      "loss": 0.6205,
-      "step": 572
-    },
-    {
-      "epoch": 2.9536082474226806,
-      "grad_norm": 0.7071570754051208,
-      "learning_rate": 3.9326239188165025e-05,
-      "loss": 0.7817,
-      "step": 573
-    },
-    {
-      "epoch": 2.9587628865979383,
-      "grad_norm": 0.6648340821266174,
-      "learning_rate": 3.915949880011096e-05,
-      "loss": 0.7229,
-      "step": 574
-    },
-    {
-      "epoch": 2.963917525773196,
-      "grad_norm": 0.5541516542434692,
-      "learning_rate": 3.899288481961173e-05,
-      "loss": 0.5545,
-      "step": 575
-    },
-    {
-      "epoch": 2.9690721649484537,
-      "grad_norm": 0.7823721766471863,
-      "learning_rate": 3.8826399189498654e-05,
-      "loss": 0.784,
-      "step": 576
-    },
-    {
-      "epoch": 2.9742268041237114,
-      "grad_norm": 0.7309272885322571,
-      "learning_rate": 3.86600438511065e-05,
-      "loss": 0.7443,
-      "step": 577
-    },
-    {
-      "epoch": 2.979381443298969,
-      "grad_norm": 0.6938033699989319,
-      "learning_rate": 3.8493820744250685e-05,
-      "loss": 0.7407,
-      "step": 578
-    },
-    {
-      "epoch": 2.984536082474227,
-      "grad_norm": 0.49541375041007996,
-      "learning_rate": 3.832773180720475e-05,
-      "loss": 0.6053,
-      "step": 579
-    },
-    {
-      "epoch": 2.9896907216494846,
-      "grad_norm": 0.6114165782928467,
-      "learning_rate": 3.8161778976677666e-05,
-      "loss": 0.5553,
-      "step": 580
-    },
-    {
-      "epoch": 2.9948453608247423,
-      "grad_norm": 0.7064359188079834,
-      "learning_rate": 3.79959641877914e-05,
-      "loss": 0.5511,
-      "step": 581
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 0.6850186586380005,
-      "learning_rate": 3.783028937405821e-05,
-      "loss": 0.5452,
-      "step": 582
-    },
-    {
-      "epoch": 3.0,
-      "eval_loss": 0.7680767178535461,
-      "eval_runtime": 23.1292,
-      "eval_samples_per_second": 7.091,
-      "eval_steps_per_second": 1.773,
-      "step": 582
-    },
-    {
-      "epoch": 3.0051546391752577,
-      "grad_norm": 0.592870831489563,
-      "learning_rate": 3.766475646735815e-05,
-      "loss": 0.6334,
-      "step": 583
-    },
-    {
-      "epoch": 3.0103092783505154,
-      "grad_norm": 0.7153403162956238,
-      "learning_rate": 3.7499367397916555e-05,
-      "loss": 0.6852,
-      "step": 584
-    },
-    {
-      "epoch": 3.015463917525773,
-      "grad_norm": 0.6092976927757263,
-      "learning_rate": 3.733412409428148e-05,
-      "loss": 0.5763,
-      "step": 585
-    },
-    {
-      "epoch": 3.020618556701031,
-      "grad_norm": 0.6821876168251038,
-      "learning_rate": 3.716902848330133e-05,
-      "loss": 0.5305,
-      "step": 586
-    },
-    {
-      "epoch": 3.0257731958762886,
-      "grad_norm": 0.6120645403862,
-      "learning_rate": 3.7004082490102226e-05,
-      "loss": 0.6668,
-      "step": 587
-    },
-    {
-      "epoch": 3.0309278350515463,
-      "grad_norm": 0.5387507081031799,
-      "learning_rate": 3.6839288038065734e-05,
-      "loss": 0.5388,
-      "step": 588
-    },
-    {
-      "epoch": 3.036082474226804,
-      "grad_norm": 0.6281033158302307,
-      "learning_rate": 3.667464704880625e-05,
-      "loss": 0.6401,
-      "step": 589
-    },
-    {
-      "epoch": 3.0412371134020617,
-      "grad_norm": 0.7044920921325684,
-      "learning_rate": 3.651016144214878e-05,
-      "loss": 0.5057,
-      "step": 590
-    },
-    {
-      "epoch": 3.0463917525773194,
-      "grad_norm": 0.7941780090332031,
-      "learning_rate": 3.634583313610644e-05,
-      "loss": 0.8192,
-      "step": 591
-    },
-    {
-      "epoch": 3.051546391752577,
-      "grad_norm": 0.7709544897079468,
-      "learning_rate": 3.618166404685805e-05,
-      "loss": 0.4853,
-      "step": 592
-    },
-    {
-      "epoch": 3.056701030927835,
-      "grad_norm": 0.5864732265472412,
-      "learning_rate": 3.601765608872595e-05,
-      "loss": 0.6387,
-      "step": 593
-    },
-    {
-      "epoch": 3.0618556701030926,
-      "grad_norm": 0.8262513875961304,
-      "learning_rate": 3.585381117415349e-05,
-      "loss": 0.602,
-      "step": 594
-    },
-    {
-      "epoch": 3.0670103092783507,
-      "grad_norm": 0.6160711050033569,
-      "learning_rate": 3.5690131213682944e-05,
-      "loss": 0.4651,
-      "step": 595
-    },
-    {
-      "epoch": 3.0721649484536084,
-      "grad_norm": 0.7176710367202759,
-      "learning_rate": 3.5526618115932975e-05,
-      "loss": 0.8127,
-      "step": 596
-    },
-    {
-      "epoch": 3.077319587628866,
-      "grad_norm": 0.7125350832939148,
-      "learning_rate": 3.53632737875766e-05,
-      "loss": 0.4828,
-      "step": 597
-    },
-    {
-      "epoch": 3.082474226804124,
-      "grad_norm": 0.7212585210800171,
-      "learning_rate": 3.5200100133318834e-05,
-      "loss": 0.5628,
-      "step": 598
-    },
-    {
-      "epoch": 3.0876288659793816,
-      "grad_norm": 0.5740189552307129,
-      "learning_rate": 3.5037099055874536e-05,
-      "loss": 0.4883,
-      "step": 599
-    },
-    {
-      "epoch": 3.0927835051546393,
-      "grad_norm": 0.7745830416679382,
-      "learning_rate": 3.487427245594622e-05,
-      "loss": 0.566,
-      "step": 600
-    },
-    {
-      "epoch": 3.097938144329897,
-      "grad_norm": 0.6833390593528748,
-      "learning_rate": 3.47116222322018e-05,
-      "loss": 0.5527,
-      "step": 601
-    },
-    {
-      "epoch": 3.1030927835051547,
-      "grad_norm": 0.6211826205253601,
-      "learning_rate": 3.4549150281252636e-05,
-      "loss": 0.5606,
-      "step": 602
-    },
-    {
-      "epoch": 3.1082474226804124,
-      "grad_norm": 0.6371477246284485,
-      "learning_rate": 3.4386858497631205e-05,
-      "loss": 0.5087,
-      "step": 603
-    },
-    {
-      "epoch": 3.11340206185567,
-      "grad_norm": 0.688470721244812,
-      "learning_rate": 3.422474877376917e-05,
-      "loss": 0.5553,
-      "step": 604
-    },
-    {
-      "epoch": 3.118556701030928,
-      "grad_norm": 0.6175742149353027,
-      "learning_rate": 3.406282299997521e-05,
-      "loss": 0.6158,
-      "step": 605
-    },
-    {
-      "epoch": 3.1237113402061856,
-      "grad_norm": 0.7345418334007263,
-      "learning_rate": 3.3901083064413095e-05,
-      "loss": 0.6047,
-      "step": 606
-    },
-    {
-      "epoch": 3.1288659793814433,
-      "grad_norm": 0.5087317824363708,
-      "learning_rate": 3.3739530853079516e-05,
-      "loss": 0.4364,
-      "step": 607
-    },
-    {
-      "epoch": 3.134020618556701,
-      "grad_norm": 0.6300753355026245,
-      "learning_rate": 3.357816824978222e-05,
-      "loss": 0.5034,
-      "step": 608
-    },
-    {
-      "epoch": 3.1391752577319587,
-      "grad_norm": 0.6126660704612732,
-      "learning_rate": 3.341699713611799e-05,
-      "loss": 0.5927,
-      "step": 609
-    },
-    {
-      "epoch": 3.1443298969072164,
-      "grad_norm": 0.7594387531280518,
-      "learning_rate": 3.325601939145069e-05,
-      "loss": 0.5538,
-      "step": 610
-    },
-    {
-      "epoch": 3.149484536082474,
-      "grad_norm": 0.7265405654907227,
-      "learning_rate": 3.309523689288941e-05,
-      "loss": 0.5529,
-      "step": 611
-    },
-    {
-      "epoch": 3.154639175257732,
-      "grad_norm": 0.6953518986701965,
-      "learning_rate": 3.293465151526649e-05,
-      "loss": 0.6362,
-      "step": 612
-    },
-    {
-      "epoch": 3.1597938144329896,
-      "grad_norm": 0.6807132363319397,
-      "learning_rate": 3.277426513111575e-05,
-      "loss": 0.6603,
-      "step": 613
-    },
-    {
-      "epoch": 3.1649484536082473,
-      "grad_norm": 0.6613560914993286,
-      "learning_rate": 3.261407961065056e-05,
-      "loss": 0.6489,
-      "step": 614
-    },
-    {
-      "epoch": 3.170103092783505,
-      "grad_norm": 0.6142688989639282,
-      "learning_rate": 3.245409682174217e-05,
-      "loss": 0.4752,
-      "step": 615
-    },
-    {
-      "epoch": 3.1752577319587627,
-      "grad_norm": 0.6799762845039368,
-      "learning_rate": 3.229431862989775e-05,
-      "loss": 0.6954,
-      "step": 616
-    },
-    {
-      "epoch": 3.1804123711340204,
-      "grad_norm": 0.732398509979248,
-      "learning_rate": 3.2134746898238774e-05,
-      "loss": 0.5572,
-      "step": 617
-    },
-    {
-      "epoch": 3.1855670103092786,
-      "grad_norm": 0.7211857438087463,
-      "learning_rate": 3.197538348747927e-05,
-      "loss": 0.5869,
-      "step": 618
-    },
-    {
-      "epoch": 3.1907216494845363,
-      "grad_norm": 0.5751710534095764,
-      "learning_rate": 3.181623025590405e-05,
-      "loss": 0.4213,
-      "step": 619
-    },
-    {
-      "epoch": 3.195876288659794,
-      "grad_norm": 0.4845646917819977,
-      "learning_rate": 3.165728905934718e-05,
-      "loss": 0.3849,
-      "step": 620
-    },
-    {
-      "epoch": 3.2010309278350517,
-      "grad_norm": 0.7036471366882324,
-      "learning_rate": 3.149856175117014e-05,
-      "loss": 0.4576,
-      "step": 621
-    },
-    {
-      "epoch": 3.2061855670103094,
-      "grad_norm": 0.7593416571617126,
-      "learning_rate": 3.134005018224044e-05,
-      "loss": 0.6383,
-      "step": 622
-    },
-    {
-      "epoch": 3.211340206185567,
-      "grad_norm": 0.8135774731636047,
-      "learning_rate": 3.118175620090983e-05,
-      "loss": 0.5818,
-      "step": 623
-    },
-    {
-      "epoch": 3.216494845360825,
-      "grad_norm": 0.6283600330352783,
-      "learning_rate": 3.1023681652992926e-05,
-      "loss": 0.6409,
-      "step": 624
-    },
-    {
-      "epoch": 3.2216494845360826,
-      "grad_norm": 0.7765624523162842,
-      "learning_rate": 3.086582838174551e-05,
-      "loss": 0.6392,
-      "step": 625
-    },
-    {
-      "epoch": 3.2268041237113403,
-      "grad_norm": 0.6492854356765747,
-      "learning_rate": 3.070819822784323e-05,
-      "loss": 0.5593,
-      "step": 626
-    },
-    {
-      "epoch": 3.231958762886598,
-      "grad_norm": 0.8030527234077454,
-      "learning_rate": 3.055079302935997e-05,
-      "loss": 0.548,
-      "step": 627
-    },
-    {
-      "epoch": 3.2371134020618557,
-      "grad_norm": 0.7499942779541016,
-      "learning_rate": 3.0393614621746498e-05,
-      "loss": 0.5565,
-      "step": 628
-    },
-    {
-      "epoch": 3.2422680412371134,
-      "grad_norm": 0.7912932634353638,
-      "learning_rate": 3.023666483780905e-05,
-      "loss": 0.6234,
-      "step": 629
-    },
-    {
-      "epoch": 3.247422680412371,
-      "grad_norm": 0.6806544661521912,
-      "learning_rate": 3.007994550768793e-05,
-      "loss": 0.5514,
-      "step": 630
-    },
-    {
-      "epoch": 3.252577319587629,
-      "grad_norm": 0.7280873656272888,
-      "learning_rate": 2.9923458458836258e-05,
-      "loss": 0.6465,
-      "step": 631
-    },
-    {
-      "epoch": 3.2577319587628866,
-      "grad_norm": 0.7153916954994202,
-      "learning_rate": 2.9767205515998518e-05,
-      "loss": 0.4821,
-      "step": 632
-    },
-    {
-      "epoch": 3.2628865979381443,
-      "grad_norm": 0.9249881505966187,
-      "learning_rate": 2.9611188501189435e-05,
-      "loss": 0.6613,
-      "step": 633
-    },
-    {
-      "epoch": 3.268041237113402,
-      "grad_norm": 0.7101225256919861,
-      "learning_rate": 2.9455409233672592e-05,
-      "loss": 0.5088,
-      "step": 634
-    },
-    {
-      "epoch": 3.2731958762886597,
-      "grad_norm": 0.7100300192832947,
-      "learning_rate": 2.929986952993933e-05,
-      "loss": 0.6544,
-      "step": 635
-    },
-    {
-      "epoch": 3.2783505154639174,
-      "grad_norm": 0.8734927177429199,
-      "learning_rate": 2.9144571203687476e-05,
-      "loss": 0.5974,
-      "step": 636
-    },
-    {
-      "epoch": 3.283505154639175,
-      "grad_norm": 0.7215682864189148,
-      "learning_rate": 2.8989516065800238e-05,
-      "loss": 0.565,
-      "step": 637
-    },
-    {
-      "epoch": 3.288659793814433,
-      "grad_norm": 0.7901187539100647,
-      "learning_rate": 2.8834705924325118e-05,
-      "loss": 0.6572,
-      "step": 638
-    },
-    {
-      "epoch": 3.2938144329896906,
-      "grad_norm": 0.6137844920158386,
-      "learning_rate": 2.8680142584452742e-05,
-      "loss": 0.5265,
-      "step": 639
-    },
-    {
-      "epoch": 3.2989690721649483,
-      "grad_norm": 0.7701372504234314,
-      "learning_rate": 2.8525827848495913e-05,
-      "loss": 0.68,
-      "step": 640
-    },
-    {
-      "epoch": 3.304123711340206,
-      "grad_norm": 0.681596040725708,
-      "learning_rate": 2.83717635158685e-05,
-      "loss": 0.6126,
-      "step": 641
-    },
-    {
-      "epoch": 3.3092783505154637,
-      "grad_norm": 0.7642390131950378,
-      "learning_rate": 2.8217951383064544e-05,
-      "loss": 0.7179,
-      "step": 642
-    },
-    {
-      "epoch": 3.3144329896907214,
-      "grad_norm": 0.6880099773406982,
-      "learning_rate": 2.8064393243637222e-05,
-      "loss": 0.5507,
-      "step": 643
-    },
-    {
-      "epoch": 3.319587628865979,
-      "grad_norm": 0.6608781814575195,
-      "learning_rate": 2.791109088817803e-05,
-      "loss": 0.4494,
-      "step": 644
-    },
-    {
-      "epoch": 3.3247422680412373,
-      "grad_norm": 0.685088038444519,
-      "learning_rate": 2.7758046104295797e-05,
-      "loss": 0.6755,
-      "step": 645
-    },
-    {
-      "epoch": 3.329896907216495,
-      "grad_norm": 0.8195577263832092,
-      "learning_rate": 2.760526067659591e-05,
-      "loss": 0.5689,
-      "step": 646
-    },
-    {
-      "epoch": 3.3350515463917527,
-      "grad_norm": 0.801682710647583,
-      "learning_rate": 2.7452736386659516e-05,
-      "loss": 0.7178,
-      "step": 647
-    },
-    {
-      "epoch": 3.3402061855670104,
-      "grad_norm": 0.653991162776947,
-      "learning_rate": 2.7300475013022663e-05,
-      "loss": 0.571,
-      "step": 648
-    },
-    {
-      "epoch": 3.345360824742268,
-      "grad_norm": 0.6898659467697144,
-      "learning_rate": 2.7148478331155702e-05,
-      "loss": 0.5027,
-      "step": 649
-    },
-    {
-      "epoch": 3.350515463917526,
-      "grad_norm": 0.7034061551094055,
-      "learning_rate": 2.6996748113442394e-05,
-      "loss": 0.4802,
-      "step": 650
-    },
-    {
-      "epoch": 3.3556701030927836,
-      "grad_norm": 0.6520366072654724,
-      "learning_rate": 2.6845286129159464e-05,
-      "loss": 0.4855,
-      "step": 651
-    },
-    {
-      "epoch": 3.3608247422680413,
-      "grad_norm": 0.6832021474838257,
-      "learning_rate": 2.669409414445574e-05,
-      "loss": 0.5634,
-      "step": 652
-    },
-    {
-      "epoch": 3.365979381443299,
-      "grad_norm": 0.8851689696311951,
-      "learning_rate": 2.6543173922331743e-05,
-      "loss": 0.7024,
-      "step": 653
-    },
-    {
-      "epoch": 3.3711340206185567,
-      "grad_norm": 0.6908465623855591,
-      "learning_rate": 2.639252722261908e-05,
-      "loss": 0.5088,
-      "step": 654
-    },
-    {
-      "epoch": 3.3762886597938144,
-      "grad_norm": 0.7850653529167175,
-      "learning_rate": 2.624215580195981e-05,
-      "loss": 0.5754,
-      "step": 655
-    },
-    {
-      "epoch": 3.381443298969072,
-      "grad_norm": 0.7114818096160889,
-      "learning_rate": 2.6092061413786156e-05,
-      "loss": 0.5193,
-      "step": 656
-    },
-    {
-      "epoch": 3.38659793814433,
-      "grad_norm": 0.7497783899307251,
-      "learning_rate": 2.5942245808299886e-05,
-      "loss": 0.5521,
-      "step": 657
-    },
-    {
-      "epoch": 3.3917525773195876,
-      "grad_norm": 0.7602093815803528,
-      "learning_rate": 2.5792710732451997e-05,
-      "loss": 0.6786,
-      "step": 658
-    },
-    {
-      "epoch": 3.3969072164948453,
-      "grad_norm": 0.6801767945289612,
-      "learning_rate": 2.56434579299223e-05,
-      "loss": 0.6006,
-      "step": 659
-    },
-    {
-      "epoch": 3.402061855670103,
-      "grad_norm": 0.8054990172386169,
-      "learning_rate": 2.5494489141099153e-05,
-      "loss": 0.6026,
-      "step": 660
-    },
-    {
-      "epoch": 3.4072164948453607,
-      "grad_norm": 0.5966134071350098,
-      "learning_rate": 2.534580610305909e-05,
-      "loss": 0.4467,
-      "step": 661
-    },
-    {
-      "epoch": 3.4123711340206184,
-      "grad_norm": 0.6399171352386475,
-      "learning_rate": 2.5197410549546595e-05,
-      "loss": 0.3739,
-      "step": 662
-    },
-    {
-      "epoch": 3.417525773195876,
-      "grad_norm": 0.6846553683280945,
-      "learning_rate": 2.5049304210953933e-05,
-      "loss": 0.5254,
-      "step": 663
-    },
-    {
-      "epoch": 3.422680412371134,
-      "grad_norm": 0.7457582950592041,
-      "learning_rate": 2.4901488814300856e-05,
-      "loss": 0.5307,
-      "step": 664
-    },
-    {
-      "epoch": 3.4278350515463916,
-      "grad_norm": 0.7888783812522888,
-      "learning_rate": 2.4753966083214615e-05,
-      "loss": 0.6,
-      "step": 665
-    },
-    {
-      "epoch": 3.4329896907216497,
-      "grad_norm": 0.7287904024124146,
-      "learning_rate": 2.4606737737909697e-05,
-      "loss": 0.5551,
-      "step": 666
-    },
-    {
-      "epoch": 3.4381443298969074,
-      "grad_norm": 0.852834939956665,
-      "learning_rate": 2.4459805495167942e-05,
-      "loss": 0.651,
-      "step": 667
-    },
-    {
-      "epoch": 3.443298969072165,
-      "grad_norm": 0.7840898633003235,
-      "learning_rate": 2.4313171068318357e-05,
-      "loss": 0.6072,
-      "step": 668
-    },
-    {
-      "epoch": 3.448453608247423,
-      "grad_norm": 0.8575794696807861,
-      "learning_rate": 2.4166836167217283e-05,
-      "loss": 0.6781,
-      "step": 669
-    },
-    {
-      "epoch": 3.4536082474226806,
-      "grad_norm": 0.6471691131591797,
-      "learning_rate": 2.4020802498228335e-05,
-      "loss": 0.5674,
-      "step": 670
-    },
-    {
-      "epoch": 3.4587628865979383,
-      "grad_norm": 0.7855519652366638,
-      "learning_rate": 2.3875071764202563e-05,
-      "loss": 0.5972,
-      "step": 671
-    },
-    {
-      "epoch": 3.463917525773196,
-      "grad_norm": 0.8715437054634094,
-      "learning_rate": 2.3729645664458638e-05,
-      "loss": 0.6441,
-      "step": 672
-    },
-    {
-      "epoch": 3.4690721649484537,
-      "grad_norm": 0.6959824562072754,
-      "learning_rate": 2.3584525894762928e-05,
-      "loss": 0.4333,
-      "step": 673
-    },
-    {
-      "epoch": 3.4742268041237114,
-      "grad_norm": 0.6880579590797424,
-      "learning_rate": 2.3439714147309845e-05,
-      "loss": 0.5053,
-      "step": 674
-    },
-    {
-      "epoch": 3.479381443298969,
-      "grad_norm": 0.7788439393043518,
-      "learning_rate": 2.329521211070199e-05,
-      "loss": 0.5467,
-      "step": 675
-    },
-    {
-      "epoch": 3.484536082474227,
-      "grad_norm": 0.7182348966598511,
-      "learning_rate": 2.3151021469930613e-05,
-      "loss": 0.6482,
-      "step": 676
-    },
-    {
-      "epoch": 3.4896907216494846,
-      "grad_norm": 0.794500470161438,
-      "learning_rate": 2.3007143906355767e-05,
-      "loss": 0.5767,
-      "step": 677
-    },
-    {
-      "epoch": 3.4948453608247423,
-      "grad_norm": 0.9387006759643555,
-      "learning_rate": 2.2863581097686925e-05,
-      "loss": 0.8179,
-      "step": 678
-    },
-    {
-      "epoch": 3.5,
-      "grad_norm": 0.6931821703910828,
-      "learning_rate": 2.2720334717963222e-05,
-      "loss": 0.5789,
-      "step": 679
-    },
-    {
-      "epoch": 3.5,
-      "eval_loss": 0.7876521944999695,
-      "eval_runtime": 23.1822,
-      "eval_samples_per_second": 7.074,
-      "eval_steps_per_second": 1.769,
-      "step": 679
-    },
-    {
-      "epoch": 3.5051546391752577,
-      "grad_norm": 0.6769412159919739,
-      "learning_rate": 2.2577406437534054e-05,
-      "loss": 0.5329,
-      "step": 680
-    },
-    {
-      "epoch": 3.5103092783505154,
-      "grad_norm": 0.683117151260376,
-      "learning_rate": 2.2434797923039598e-05,
-      "loss": 0.5284,
-      "step": 681
-    },
-    {
-      "epoch": 3.515463917525773,
-      "grad_norm": 0.790727972984314,
-      "learning_rate": 2.2292510837391267e-05,
-      "loss": 0.612,
-      "step": 682
-    },
-    {
-      "epoch": 3.520618556701031,
-      "grad_norm": 0.7108373641967773,
-      "learning_rate": 2.2150546839752438e-05,
-      "loss": 0.6705,
-      "step": 683
-    },
-    {
-      "epoch": 3.5257731958762886,
-      "grad_norm": 0.8616272807121277,
-      "learning_rate": 2.2008907585519095e-05,
-      "loss": 0.5653,
-      "step": 684
-    },
-    {
-      "epoch": 3.5309278350515463,
-      "grad_norm": 0.8035832047462463,
-      "learning_rate": 2.186759472630045e-05,
-      "loss": 0.8008,
-      "step": 685
-    },
-    {
-      "epoch": 3.536082474226804,
-      "grad_norm": 0.7284449338912964,
-      "learning_rate": 2.172660990989971e-05,
-      "loss": 0.56,
-      "step": 686
-    },
-    {
-      "epoch": 3.5412371134020617,
-      "grad_norm": 0.803911566734314,
-      "learning_rate": 2.1585954780294947e-05,
-      "loss": 0.6312,
-      "step": 687
-    },
-    {
-      "epoch": 3.5463917525773194,
-      "grad_norm": 0.8152459859848022,
-      "learning_rate": 2.144563097761984e-05,
-      "loss": 0.5366,
-      "step": 688
-    },
-    {
-      "epoch": 3.551546391752577,
-      "grad_norm": 0.7566794157028198,
-      "learning_rate": 2.130564013814453e-05,
-      "loss": 0.6201,
-      "step": 689
-    },
-    {
-      "epoch": 3.556701030927835,
-      "grad_norm": 0.6466950178146362,
-      "learning_rate": 2.1165983894256647e-05,
-      "loss": 0.5122,
-      "step": 690
-    },
-    {
-      "epoch": 3.5618556701030926,
-      "grad_norm": 0.8573073148727417,
-      "learning_rate": 2.102666387444215e-05,
-      "loss": 0.6787,
-      "step": 691
-    },
-    {
-      "epoch": 3.5670103092783503,
-      "grad_norm": 0.6495526432991028,
-      "learning_rate": 2.0887681703266453e-05,
-      "loss": 0.4294,
-      "step": 692
-    },
-    {
-      "epoch": 3.572164948453608,
-      "grad_norm": 0.6471603512763977,
-      "learning_rate": 2.0749039001355375e-05,
-      "loss": 0.5999,
-      "step": 693
-    },
-    {
-      "epoch": 3.5773195876288657,
-      "grad_norm": 0.7354085445404053,
-      "learning_rate": 2.061073738537635e-05,
-      "loss": 0.5685,
-      "step": 694
-    },
-    {
-      "epoch": 3.582474226804124,
-      "grad_norm": 0.7679943442344666,
-      "learning_rate": 2.0472778468019454e-05,
-      "loss": 0.5547,
-      "step": 695
-    },
-    {
-      "epoch": 3.5876288659793816,
-      "grad_norm": 0.6642536520957947,
-      "learning_rate": 2.0335163857978744e-05,
-      "loss": 0.4903,
-      "step": 696
-    },
-    {
-      "epoch": 3.5927835051546393,
-      "grad_norm": 0.927869439125061,
-      "learning_rate": 2.019789515993336e-05,
-      "loss": 0.6191,
-      "step": 697
-    },
-    {
-      "epoch": 3.597938144329897,
-      "grad_norm": 0.718420684337616,
-      "learning_rate": 2.0060973974528874e-05,
-      "loss": 0.4906,
-      "step": 698
-    },
-    {
-      "epoch": 3.6030927835051547,
-      "grad_norm": 0.7137137651443481,
-      "learning_rate": 1.992440189835869e-05,
-      "loss": 0.7192,
-      "step": 699
-    },
-    {
-      "epoch": 3.6082474226804124,
-      "grad_norm": 0.8115518093109131,
-      "learning_rate": 1.9788180523945277e-05,
-      "loss": 0.5906,
-      "step": 700
-    },
-    {
-      "epoch": 3.61340206185567,
-      "grad_norm": 0.8083131313323975,
-      "learning_rate": 1.9652311439721764e-05,
-      "loss": 0.6711,
-      "step": 701
-    },
-    {
-      "epoch": 3.618556701030928,
-      "grad_norm": 0.6981199383735657,
-      "learning_rate": 1.9516796230013272e-05,
-      "loss": 0.598,
-      "step": 702
-    },
-    {
-      "epoch": 3.6237113402061856,
-      "grad_norm": 0.6912441253662109,
-      "learning_rate": 1.9381636475018577e-05,
-      "loss": 0.5254,
-      "step": 703
-    },
-    {
-      "epoch": 3.6288659793814433,
-      "grad_norm": 0.6640885472297668,
-      "learning_rate": 1.9246833750791526e-05,
-      "loss": 0.4806,
-      "step": 704
-    },
-    {
-      "epoch": 3.634020618556701,
-      "grad_norm": 0.6553522944450378,
-      "learning_rate": 1.9112389629222823e-05,
-      "loss": 0.6355,
-      "step": 705
-    },
-    {
-      "epoch": 3.6391752577319587,
-      "grad_norm": 0.8154643177986145,
-      "learning_rate": 1.8978305678021595e-05,
-      "loss": 0.6503,
-      "step": 706
-    },
-    {
-      "epoch": 3.6443298969072164,
-      "grad_norm": 0.6225295066833496,
-      "learning_rate": 1.884458346069713e-05,
-      "loss": 0.5938,
-      "step": 707
-    },
-    {
-      "epoch": 3.649484536082474,
-      "grad_norm": 0.8098758459091187,
-      "learning_rate": 1.8711224536540678e-05,
-      "loss": 0.5258,
-      "step": 708
-    },
-    {
-      "epoch": 3.654639175257732,
-      "grad_norm": 0.7377306818962097,
-      "learning_rate": 1.857823046060722e-05,
-      "loss": 0.6682,
-      "step": 709
-    },
-    {
-      "epoch": 3.6597938144329896,
-      "grad_norm": 0.7113179564476013,
-      "learning_rate": 1.8445602783697374e-05,
-      "loss": 0.5529,
-      "step": 710
-    },
-    {
-      "epoch": 3.6649484536082473,
-      "grad_norm": 0.7635980248451233,
-      "learning_rate": 1.831334305233928e-05,
-      "loss": 0.5538,
-      "step": 711
-    },
-    {
-      "epoch": 3.670103092783505,
-      "grad_norm": 0.6800134181976318,
-      "learning_rate": 1.8181452808770637e-05,
-      "loss": 0.5837,
-      "step": 712
-    },
-    {
-      "epoch": 3.675257731958763,
-      "grad_norm": 0.6108891367912292,
-      "learning_rate": 1.804993359092059e-05,
-      "loss": 0.5765,
-      "step": 713
-    },
-    {
-      "epoch": 3.680412371134021,
-      "grad_norm": 0.6556738018989563,
-      "learning_rate": 1.7918786932391944e-05,
-      "loss": 0.5426,
-      "step": 714
-    },
-    {
-      "epoch": 3.6855670103092786,
-      "grad_norm": 0.7539121508598328,
-      "learning_rate": 1.778801436244319e-05,
-      "loss": 0.6723,
-      "step": 715
-    },
-    {
-      "epoch": 3.6907216494845363,
-      "grad_norm": 0.7283245325088501,
-      "learning_rate": 1.765761740597065e-05,
-      "loss": 0.6965,
-      "step": 716
-    },
-    {
-      "epoch": 3.695876288659794,
-      "grad_norm": 0.7831310033798218,
-      "learning_rate": 1.7527597583490822e-05,
-      "loss": 0.6255,
-      "step": 717
-    },
-    {
-      "epoch": 3.7010309278350517,
-      "grad_norm": 0.5926414132118225,
-      "learning_rate": 1.739795641112248e-05,
-      "loss": 0.4947,
-      "step": 718
-    },
-    {
-      "epoch": 3.7061855670103094,
-      "grad_norm": 0.8325262069702148,
-      "learning_rate": 1.726869540056915e-05,
-      "loss": 0.6241,
-      "step": 719
-    },
-    {
-      "epoch": 3.711340206185567,
-      "grad_norm": 0.69277423620224,
-      "learning_rate": 1.713981605910137e-05,
-      "loss": 0.6247,
-      "step": 720
-    },
-    {
-      "epoch": 3.716494845360825,
-      "grad_norm": 0.7472510933876038,
-      "learning_rate": 1.70113198895392e-05,
-      "loss": 0.525,
-      "step": 721
-    },
-    {
-      "epoch": 3.7216494845360826,
-      "grad_norm": 0.8484485149383545,
-      "learning_rate": 1.6883208390234628e-05,
-      "loss": 0.5437,
-      "step": 722
-    },
-    {
-      "epoch": 3.7268041237113403,
-      "grad_norm": 0.6205007433891296,
-      "learning_rate": 1.6755483055054105e-05,
-      "loss": 0.4387,
-      "step": 723
-    },
-    {
-      "epoch": 3.731958762886598,
-      "grad_norm": 0.7974679470062256,
-      "learning_rate": 1.662814537336122e-05,
-      "loss": 0.5756,
-      "step": 724
-    },
-    {
-      "epoch": 3.7371134020618557,
-      "grad_norm": 0.9275572896003723,
-      "learning_rate": 1.650119682999918e-05,
-      "loss": 0.7086,
-      "step": 725
-    },
-    {
-      "epoch": 3.7422680412371134,
-      "grad_norm": 0.5912320017814636,
-      "learning_rate": 1.6374638905273643e-05,
-      "loss": 0.4894,
-      "step": 726
-    },
-    {
-      "epoch": 3.747422680412371,
-      "grad_norm": 0.8025346994400024,
-      "learning_rate": 1.624847307493534e-05,
-      "loss": 0.5677,
-      "step": 727
-    },
-    {
-      "epoch": 3.752577319587629,
-      "grad_norm": 0.8356587886810303,
-      "learning_rate": 1.6122700810162966e-05,
-      "loss": 0.6478,
-      "step": 728
-    },
-    {
-      "epoch": 3.7577319587628866,
-      "grad_norm": 0.7288328409194946,
-      "learning_rate": 1.5997323577545915e-05,
-      "loss": 0.6042,
-      "step": 729
-    },
-    {
-      "epoch": 3.7628865979381443,
-      "grad_norm": 0.7492033839225769,
-      "learning_rate": 1.5872342839067306e-05,
-      "loss": 0.5439,
-      "step": 730
-    },
-    {
-      "epoch": 3.768041237113402,
-      "grad_norm": 0.7219738364219666,
-      "learning_rate": 1.5747760052086803e-05,
-      "loss": 0.7247,
-      "step": 731
-    },
-    {
-      "epoch": 3.7731958762886597,
-      "grad_norm": 0.7529350519180298,
-      "learning_rate": 1.5623576669323743e-05,
-      "loss": 0.595,
-      "step": 732
-    },
-    {
-      "epoch": 3.7783505154639174,
-      "grad_norm": 0.613925576210022,
-      "learning_rate": 1.5499794138840122e-05,
-      "loss": 0.5348,
-      "step": 733
-    },
-    {
-      "epoch": 3.783505154639175,
-      "grad_norm": 0.7281404733657837,
-      "learning_rate": 1.5376413904023722e-05,
-      "loss": 0.431,
-      "step": 734
-    },
-    {
-      "epoch": 3.788659793814433,
-      "grad_norm": 0.6786961555480957,
-      "learning_rate": 1.525343740357128e-05,
-      "loss": 0.5334,
-      "step": 735
-    },
-    {
-      "epoch": 3.7938144329896906,
-      "grad_norm": 0.7997889518737793,
-      "learning_rate": 1.5130866071471717e-05,
-      "loss": 0.5712,
-      "step": 736
-    },
-    {
-      "epoch": 3.7989690721649483,
-      "grad_norm": 0.8244373798370361,
-      "learning_rate": 1.500870133698945e-05,
-      "loss": 0.6237,
-      "step": 737
-    },
-    {
-      "epoch": 3.804123711340206,
-      "grad_norm": 0.7046230435371399,
-      "learning_rate": 1.4886944624647647e-05,
-      "loss": 0.5052,
-      "step": 738
-    },
-    {
-      "epoch": 3.8092783505154637,
-      "grad_norm": 0.6828339695930481,
-      "learning_rate": 1.4765597354211713e-05,
-      "loss": 0.4986,
-      "step": 739
-    },
-    {
-      "epoch": 3.8144329896907214,
-      "grad_norm": 0.7806636691093445,
-      "learning_rate": 1.4644660940672627e-05,
-      "loss": 0.5256,
-      "step": 740
-    },
-    {
-      "epoch": 3.819587628865979,
-      "grad_norm": 0.7063342928886414,
-      "learning_rate": 1.4524136794230547e-05,
-      "loss": 0.6,
-      "step": 741
-    },
-    {
-      "epoch": 3.824742268041237,
-      "grad_norm": 0.6564049124717712,
-      "learning_rate": 1.4404026320278318e-05,
-      "loss": 0.4549,
-      "step": 742
-    },
-    {
-      "epoch": 3.829896907216495,
-      "grad_norm": 0.737053394317627,
-      "learning_rate": 1.4284330919385036e-05,
-      "loss": 0.5977,
-      "step": 743
-    },
-    {
-      "epoch": 3.8350515463917527,
-      "grad_norm": 0.8713696002960205,
-      "learning_rate": 1.4165051987279831e-05,
-      "loss": 0.7163,
-      "step": 744
-    },
-    {
-      "epoch": 3.8402061855670104,
-      "grad_norm": 0.8080871105194092,
-      "learning_rate": 1.404619091483546e-05,
-      "loss": 0.728,
-      "step": 745
-    },
-    {
-      "epoch": 3.845360824742268,
-      "grad_norm": 0.7074190378189087,
-      "learning_rate": 1.3927749088052217e-05,
-      "loss": 0.5355,
-      "step": 746
-    },
-    {
-      "epoch": 3.850515463917526,
-      "grad_norm": 0.887178897857666,
-      "learning_rate": 1.3809727888041668e-05,
-      "loss": 0.6736,
-      "step": 747
-    },
-    {
-      "epoch": 3.8556701030927836,
-      "grad_norm": 0.7117917537689209,
-      "learning_rate": 1.3692128691010592e-05,
-      "loss": 0.5781,
-      "step": 748
-    },
-    {
-      "epoch": 3.8608247422680413,
-      "grad_norm": 1.025062918663025,
-      "learning_rate": 1.3574952868244922e-05,
-      "loss": 0.6942,
-      "step": 749
-    },
-    {
-      "epoch": 3.865979381443299,
-      "grad_norm": 0.8618969917297363,
-      "learning_rate": 1.3458201786093794e-05,
-      "loss": 0.6821,
-      "step": 750
-    },
-    {
-      "epoch": 3.8711340206185567,
-      "grad_norm": 0.7979894876480103,
-      "learning_rate": 1.334187680595358e-05,
-      "loss": 0.5799,
-      "step": 751
-    },
-    {
-      "epoch": 3.8762886597938144,
-      "grad_norm": 0.6518426537513733,
-      "learning_rate": 1.3225979284251954e-05,
-      "loss": 0.5948,
-      "step": 752
-    },
-    {
-      "epoch": 3.881443298969072,
-      "grad_norm": 0.8365257382392883,
-      "learning_rate": 1.3110510572432221e-05,
-      "loss": 0.7303,
-      "step": 753
-    },
-    {
-      "epoch": 3.88659793814433,
-      "grad_norm": 0.7343977093696594,
-      "learning_rate": 1.2995472016937404e-05,
-      "loss": 0.5519,
-      "step": 754
-    },
-    {
-      "epoch": 3.8917525773195876,
-      "grad_norm": 0.7648212909698486,
-      "learning_rate": 1.2880864959194665e-05,
-      "loss": 0.6492,
-      "step": 755
-    },
-    {
-      "epoch": 3.8969072164948453,
-      "grad_norm": 0.7290052771568298,
-      "learning_rate": 1.2766690735599568e-05,
-      "loss": 0.5893,
-      "step": 756
-    },
-    {
-      "epoch": 3.902061855670103,
-      "grad_norm": 0.5812563300132751,
-      "learning_rate": 1.2652950677500574e-05,
-      "loss": 0.5686,
-      "step": 757
-    },
-    {
-      "epoch": 3.9072164948453607,
-      "grad_norm": 0.6635538339614868,
-      "learning_rate": 1.253964611118345e-05,
-      "loss": 0.4879,
-      "step": 758
-    },
-    {
-      "epoch": 3.9123711340206184,
-      "grad_norm": 0.6084232926368713,
-      "learning_rate": 1.2426778357855873e-05,
-      "loss": 0.5395,
-      "step": 759
-    },
-    {
-      "epoch": 3.917525773195876,
-      "grad_norm": 0.7855112552642822,
-      "learning_rate": 1.2314348733631959e-05,
-      "loss": 0.6793,
-      "step": 760
-    },
-    {
-      "epoch": 3.9226804123711343,
-      "grad_norm": 0.7807559370994568,
-      "learning_rate": 1.2202358549516923e-05,
-      "loss": 0.6824,
-      "step": 761
-    },
-    {
-      "epoch": 3.927835051546392,
-      "grad_norm": 0.7157235741615295,
-      "learning_rate": 1.209080911139187e-05,
-      "loss": 0.5723,
-      "step": 762
-    },
-    {
-      "epoch": 3.9329896907216497,
-      "grad_norm": 0.7934853434562683,
-      "learning_rate": 1.1979701719998453e-05,
-      "loss": 0.6191,
-      "step": 763
-    },
-    {
-      "epoch": 3.9381443298969074,
-      "grad_norm": 0.7907642722129822,
-      "learning_rate": 1.1869037670923815e-05,
-      "loss": 0.7329,
-      "step": 764
-    },
-    {
-      "epoch": 3.943298969072165,
-      "grad_norm": 0.7775223851203918,
-      "learning_rate": 1.1758818254585369e-05,
-      "loss": 0.6165,
-      "step": 765
-    },
-    {
-      "epoch": 3.948453608247423,
-      "grad_norm": 0.7809783220291138,
-      "learning_rate": 1.164904475621587e-05,
-      "loss": 0.681,
-      "step": 766
-    },
-    {
-      "epoch": 3.9536082474226806,
-      "grad_norm": 0.8733285069465637,
-      "learning_rate": 1.1539718455848309e-05,
-      "loss": 0.4686,
-      "step": 767
-    },
-    {
-      "epoch": 3.9587628865979383,
-      "grad_norm": 0.7542945146560669,
-      "learning_rate": 1.1430840628301093e-05,
-      "loss": 0.5553,
-      "step": 768
-    },
-    {
-      "epoch": 3.963917525773196,
-      "grad_norm": 0.790187656879425,
-      "learning_rate": 1.1322412543163135e-05,
-      "loss": 0.6635,
-      "step": 769
-    },
-    {
-      "epoch": 3.9690721649484537,
-      "grad_norm": 0.721284806728363,
-      "learning_rate": 1.1214435464779006e-05,
-      "loss": 0.6948,
-      "step": 770
-    },
-    {
-      "epoch": 3.9742268041237114,
-      "grad_norm": 0.6923427581787109,
-      "learning_rate": 1.1106910652234276e-05,
-      "loss": 0.5599,
-      "step": 771
-    },
-    {
-      "epoch": 3.979381443298969,
-      "grad_norm": 0.7891685962677002,
-      "learning_rate": 1.099983935934077e-05,
-      "loss": 0.6763,
-      "step": 772
-    },
-    {
-      "epoch": 3.984536082474227,
-      "grad_norm": 0.8703778386116028,
-      "learning_rate": 1.089322283462197e-05,
-      "loss": 0.6884,
-      "step": 773
-    },
-    {
-      "epoch": 3.9896907216494846,
-      "grad_norm": 0.629092812538147,
-      "learning_rate": 1.0787062321298442e-05,
-      "loss": 0.5368,
-      "step": 774
-    },
-    {
-      "epoch": 3.9948453608247423,
-      "grad_norm": 0.7654589414596558,
-      "learning_rate": 1.0681359057273388e-05,
-      "loss": 0.6477,
-      "step": 775
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 1.1668832302093506,
-      "learning_rate": 1.0576114275118131e-05,
-      "loss": 0.6873,
-      "step": 776
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.7821245193481445,
-      "eval_runtime": 23.21,
-      "eval_samples_per_second": 7.066,
-      "eval_steps_per_second": 1.766,
-      "step": 776
-    },
-    {
-      "epoch": 4.005154639175258,
-      "grad_norm": 0.6904922127723694,
-      "learning_rate": 1.0471329202057823e-05,
-      "loss": 0.4518,
-      "step": 777
-    },
-    {
-      "epoch": 4.010309278350515,
-      "grad_norm": 0.5978090763092041,
-      "learning_rate": 1.0367005059957096e-05,
-      "loss": 0.4717,
-      "step": 778
-    },
-    {
-      "epoch": 4.015463917525773,
-      "grad_norm": 0.800383985042572,
-      "learning_rate": 1.0263143065305769e-05,
-      "loss": 0.4304,
-      "step": 779
-    },
-    {
-      "epoch": 4.020618556701031,
-      "grad_norm": 0.693153440952301,
-      "learning_rate": 1.0159744429204777e-05,
-      "loss": 0.5344,
-      "step": 780
-    },
-    {
-      "epoch": 4.025773195876289,
-      "grad_norm": 0.6497305035591125,
-      "learning_rate": 1.005681035735192e-05,
-      "loss": 0.5306,
-      "step": 781
-    },
-    {
-      "epoch": 4.030927835051546,
-      "grad_norm": 0.7015767097473145,
-      "learning_rate": 9.954342050027921e-06,
-      "loss": 0.5403,
-      "step": 782
-    },
-    {
-      "epoch": 4.036082474226804,
-      "grad_norm": 0.6772336363792419,
-      "learning_rate": 9.852340702082318e-06,
-      "loss": 0.4334,
-      "step": 783
-    },
-    {
-      "epoch": 4.041237113402062,
-      "grad_norm": 0.6541298627853394,
-      "learning_rate": 9.750807502919652e-06,
-      "loss": 0.5742,
-      "step": 784
-    },
-    {
-      "epoch": 4.046391752577319,
-      "grad_norm": 0.7539450526237488,
-      "learning_rate": 9.64974363648548e-06,
-      "loss": 0.7293,
-      "step": 785
-    },
-    {
-      "epoch": 4.051546391752577,
-      "grad_norm": 0.783802330493927,
-      "learning_rate": 9.549150281252633e-06,
-      "loss": 0.5296,
-      "step": 786
-    },
-    {
-      "epoch": 4.056701030927835,
-      "grad_norm": 0.6802383661270142,
-      "learning_rate": 9.449028610207494e-06,
-      "loss": 0.5059,
-      "step": 787
-    },
-    {
-      "epoch": 4.061855670103093,
-      "grad_norm": 0.8454770445823669,
-      "learning_rate": 9.349379790836243e-06,
-      "loss": 0.6427,
-      "step": 788
-    },
-    {
-      "epoch": 4.06701030927835,
-      "grad_norm": 0.7464700937271118,
-      "learning_rate": 9.25020498511135e-06,
-      "loss": 0.6154,
-      "step": 789
-    },
-    {
-      "epoch": 4.072164948453608,
-      "grad_norm": 0.7701639533042908,
-      "learning_rate": 9.151505349477902e-06,
-      "loss": 0.5741,
-      "step": 790
-    },
-    {
-      "epoch": 4.077319587628866,
-      "grad_norm": 0.6774284243583679,
-      "learning_rate": 9.053282034840238e-06,
-      "loss": 0.5858,
-      "step": 791
-    },
-    {
-      "epoch": 4.082474226804123,
-      "grad_norm": 0.637546718120575,
-      "learning_rate": 8.955536186548425e-06,
-      "loss": 0.5001,
-      "step": 792
-    },
-    {
-      "epoch": 4.087628865979381,
-      "grad_norm": 0.7409276366233826,
-      "learning_rate": 8.858268944384995e-06,
-      "loss": 0.5082,
-      "step": 793
-    },
-    {
-      "epoch": 4.092783505154639,
-      "grad_norm": 0.8038662075996399,
-      "learning_rate": 8.761481442551573e-06,
-      "loss": 0.6967,
-      "step": 794
-    },
-    {
-      "epoch": 4.097938144329897,
-      "grad_norm": 0.7973031997680664,
-      "learning_rate": 8.665174809655708e-06,
-      "loss": 0.6006,
-      "step": 795
-    },
-    {
-      "epoch": 4.103092783505154,
-      "grad_norm": 0.7798672318458557,
-      "learning_rate": 8.569350168697704e-06,
-      "loss": 0.6076,
-      "step": 796
-    },
-    {
-      "epoch": 4.108247422680412,
-      "grad_norm": 0.7757202982902527,
-      "learning_rate": 8.474008637057478e-06,
-      "loss": 0.5881,
-      "step": 797
-    },
-    {
-      "epoch": 4.11340206185567,
-      "grad_norm": 0.8978819251060486,
-      "learning_rate": 8.379151326481587e-06,
-      "loss": 0.5522,
-      "step": 798
-    },
-    {
-      "epoch": 4.118556701030927,
-      "grad_norm": 0.8775174617767334,
-      "learning_rate": 8.284779343070265e-06,
-      "loss": 0.5599,
-      "step": 799
-    },
-    {
-      "epoch": 4.123711340206185,
-      "grad_norm": 0.6384215950965881,
-      "learning_rate": 8.19089378726447e-06,
-      "loss": 0.466,
-      "step": 800
-    },
-    {
-      "epoch": 4.128865979381443,
-      "grad_norm": 0.7570362091064453,
-      "learning_rate": 8.097495753833078e-06,
-      "loss": 0.6081,
-      "step": 801
-    },
-    {
-      "epoch": 4.134020618556701,
-      "grad_norm": 0.8209741711616516,
-      "learning_rate": 8.004586331860175e-06,
-      "loss": 0.6016,
-      "step": 802
-    },
-    {
-      "epoch": 4.139175257731959,
-      "grad_norm": 0.7955187559127808,
-      "learning_rate": 7.91216660473228e-06,
-      "loss": 0.5968,
-      "step": 803
-    },
-    {
-      "epoch": 4.144329896907217,
-      "grad_norm": 0.7099952101707458,
-      "learning_rate": 7.820237650125712e-06,
-      "loss": 0.6445,
-      "step": 804
-    },
-    {
-      "epoch": 4.149484536082475,
-      "grad_norm": 0.923795759677887,
-      "learning_rate": 7.728800539994113e-06,
-      "loss": 0.7184,
-      "step": 805
-    },
-    {
-      "epoch": 4.154639175257732,
-      "grad_norm": 0.7572796940803528,
-      "learning_rate": 7.637856340555822e-06,
-      "loss": 0.5521,
-      "step": 806
-    },
-    {
-      "epoch": 4.15979381443299,
-      "grad_norm": 0.823486864566803,
-      "learning_rate": 7.547406112281557e-06,
-      "loss": 0.5439,
-      "step": 807
-    },
-    {
-      "epoch": 4.164948453608248,
-      "grad_norm": 0.7394459247589111,
-      "learning_rate": 7.457450909881969e-06,
-      "loss": 0.499,
-      "step": 808
-    },
-    {
-      "epoch": 4.170103092783505,
-      "grad_norm": 0.8787947297096252,
-      "learning_rate": 7.367991782295391e-06,
-      "loss": 0.4774,
-      "step": 809
-    },
-    {
-      "epoch": 4.175257731958763,
-      "grad_norm": 0.931903600692749,
-      "learning_rate": 7.2790297726755716e-06,
-      "loss": 0.4642,
-      "step": 810
-    },
-    {
-      "epoch": 4.180412371134021,
-      "grad_norm": 0.8398101329803467,
-      "learning_rate": 7.190565918379549e-06,
-      "loss": 0.636,
-      "step": 811
-    },
-    {
-      "epoch": 4.185567010309279,
-      "grad_norm": 0.8237578868865967,
-      "learning_rate": 7.1026012509555265e-06,
-      "loss": 0.5771,
-      "step": 812
-    },
-    {
-      "epoch": 4.190721649484536,
-      "grad_norm": 0.8667364716529846,
-      "learning_rate": 7.015136796130828e-06,
-      "loss": 0.6154,
-      "step": 813
-    },
-    {
-      "epoch": 4.195876288659794,
-      "grad_norm": 0.803131103515625,
-      "learning_rate": 6.928173573800006e-06,
-      "loss": 0.5299,
-      "step": 814
-    },
-    {
-      "epoch": 4.201030927835052,
-      "grad_norm": 0.8482971787452698,
-      "learning_rate": 6.8417125980128675e-06,
-      "loss": 0.6238,
-      "step": 815
-    },
-    {
-      "epoch": 4.206185567010309,
-      "grad_norm": 0.7783701419830322,
-      "learning_rate": 6.755754876962711e-06,
-      "loss": 0.5216,
-      "step": 816
-    },
-    {
-      "epoch": 4.211340206185567,
-      "grad_norm": 0.6552199125289917,
-      "learning_rate": 6.670301412974511e-06,
-      "loss": 0.4832,
-      "step": 817
-    },
-    {
-      "epoch": 4.216494845360825,
-      "grad_norm": 0.8719028830528259,
-      "learning_rate": 6.585353202493322e-06,
-      "loss": 0.575,
-      "step": 818
-    },
-    {
-      "epoch": 4.221649484536083,
-      "grad_norm": 0.7946240305900574,
-      "learning_rate": 6.500911236072532e-06,
-      "loss": 0.648,
-      "step": 819
-    },
-    {
-      "epoch": 4.22680412371134,
-      "grad_norm": 0.7797208428382874,
-      "learning_rate": 6.416976498362432e-06,
-      "loss": 0.4849,
-      "step": 820
     }
   ],
   "logging_steps": 1,
@@ -5838,7 +69,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.171226860461752e+17,
+  "total_flos": 1935445139128320.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null