{
  "best_metric": 0.08146847784519196,
  "best_model_checkpoint": "./fine-tuned/checkpoint-25000",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 25000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004,
      "grad_norm": 0.24035809934139252,
      "learning_rate": 2.99412e-05,
      "loss": 0.3973,
      "step": 50
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.39832672476768494,
      "learning_rate": 2.98812e-05,
      "loss": 0.1126,
      "step": 100
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.2938326299190521,
      "learning_rate": 2.9821200000000002e-05,
      "loss": 0.0932,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.18936102092266083,
      "learning_rate": 2.9761200000000002e-05,
      "loss": 0.0897,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.18386273086071014,
      "learning_rate": 2.9701200000000003e-05,
      "loss": 0.092,
      "step": 250
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.1968618482351303,
      "learning_rate": 2.96412e-05,
      "loss": 0.0845,
      "step": 300
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.18028958141803741,
      "learning_rate": 2.95812e-05,
      "loss": 0.0792,
      "step": 350
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.2615596652030945,
      "learning_rate": 2.95212e-05,
      "loss": 0.0796,
      "step": 400
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.1913922131061554,
      "learning_rate": 2.9461200000000002e-05,
      "loss": 0.0798,
      "step": 450
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2351102977991104,
      "learning_rate": 2.9401200000000002e-05,
      "loss": 0.081,
      "step": 500
    },
    {
      "epoch": 0.04,
      "eval_loss": 0.09441258758306503,
      "eval_runtime": 88.0185,
      "eval_samples_per_second": 22.722,
      "eval_steps_per_second": 5.681,
      "step": 500
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.1859619915485382,
      "learning_rate": 2.9341200000000003e-05,
      "loss": 0.077,
      "step": 550
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.11936317384243011,
      "learning_rate": 2.92812e-05,
      "loss": 0.0727,
      "step": 600
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.2207396775484085,
      "learning_rate": 2.92212e-05,
      "loss": 0.0743,
      "step": 650
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.18488994240760803,
      "learning_rate": 2.91612e-05,
      "loss": 0.0824,
      "step": 700
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.22228538990020752,
      "learning_rate": 2.9101200000000002e-05,
      "loss": 0.0716,
      "step": 750
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.24138867855072021,
      "learning_rate": 2.9041200000000002e-05,
      "loss": 0.0814,
      "step": 800
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.25113552808761597,
      "learning_rate": 2.89812e-05,
      "loss": 0.076,
      "step": 850
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.8853724598884583,
      "learning_rate": 2.89212e-05,
      "loss": 0.0781,
      "step": 900
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.1753206104040146,
      "learning_rate": 2.88612e-05,
      "loss": 0.084,
      "step": 950
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1704334318637848,
      "learning_rate": 2.88012e-05,
      "loss": 0.0769,
      "step": 1000
    },
    {
      "epoch": 0.08,
      "eval_loss": 0.09160277992486954,
      "eval_runtime": 88.0495,
      "eval_samples_per_second": 22.714,
      "eval_steps_per_second": 5.679,
      "step": 1000
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.16729697585105896,
      "learning_rate": 2.8741200000000002e-05,
      "loss": 0.0823,
      "step": 1050
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.14851506054401398,
      "learning_rate": 2.86812e-05,
      "loss": 0.0785,
      "step": 1100
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.22481797635555267,
      "learning_rate": 2.86212e-05,
      "loss": 0.0786,
      "step": 1150
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.13808289170265198,
      "learning_rate": 2.85612e-05,
      "loss": 0.0785,
      "step": 1200
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.17833128571510315,
      "learning_rate": 2.85012e-05,
      "loss": 0.0737,
      "step": 1250
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.14926594495773315,
      "learning_rate": 2.84412e-05,
      "loss": 0.0767,
      "step": 1300
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.19346196949481964,
      "learning_rate": 2.8381200000000002e-05,
      "loss": 0.077,
      "step": 1350
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.2675027847290039,
      "learning_rate": 2.83212e-05,
      "loss": 0.0806,
      "step": 1400
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.19048169255256653,
      "learning_rate": 2.82612e-05,
      "loss": 0.0756,
      "step": 1450
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1875162124633789,
      "learning_rate": 2.82012e-05,
      "loss": 0.0823,
      "step": 1500
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.09038107097148895,
      "eval_runtime": 87.8777,
      "eval_samples_per_second": 22.759,
      "eval_steps_per_second": 5.69,
      "step": 1500
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.19083499908447266,
      "learning_rate": 2.81412e-05,
      "loss": 0.0736,
      "step": 1550
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.19123569130897522,
      "learning_rate": 2.80812e-05,
      "loss": 0.0766,
      "step": 1600
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.24691827595233917,
      "learning_rate": 2.80212e-05,
      "loss": 0.0798,
      "step": 1650
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.17910048365592957,
      "learning_rate": 2.79612e-05,
      "loss": 0.0687,
      "step": 1700
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1740667223930359,
      "learning_rate": 2.79012e-05,
      "loss": 0.0758,
      "step": 1750
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.15178219974040985,
      "learning_rate": 2.78412e-05,
      "loss": 0.0732,
      "step": 1800
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.1904926896095276,
      "learning_rate": 2.77812e-05,
      "loss": 0.0734,
      "step": 1850
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.2795208990573883,
      "learning_rate": 2.77212e-05,
      "loss": 0.076,
      "step": 1900
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.18160228431224823,
      "learning_rate": 2.76612e-05,
      "loss": 0.0716,
      "step": 1950
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15877611935138702,
      "learning_rate": 2.76012e-05,
      "loss": 0.0773,
      "step": 2000
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.08910445868968964,
      "eval_runtime": 87.8635,
      "eval_samples_per_second": 22.763,
      "eval_steps_per_second": 5.691,
      "step": 2000
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.2411368191242218,
      "learning_rate": 2.75412e-05,
      "loss": 0.0786,
      "step": 2050
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.16663742065429688,
      "learning_rate": 2.74812e-05,
      "loss": 0.0724,
      "step": 2100
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.23420193791389465,
      "learning_rate": 2.74212e-05,
      "loss": 0.0653,
      "step": 2150
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.1807372272014618,
      "learning_rate": 2.7361199999999998e-05,
      "loss": 0.0676,
      "step": 2200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.16474364697933197,
      "learning_rate": 2.73012e-05,
      "loss": 0.0767,
      "step": 2250
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.17184095084667206,
      "learning_rate": 2.72412e-05,
      "loss": 0.0658,
      "step": 2300
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.16993258893489838,
      "learning_rate": 2.71812e-05,
      "loss": 0.0755,
      "step": 2350
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.1555277407169342,
      "learning_rate": 2.71212e-05,
      "loss": 0.0698,
      "step": 2400
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.09040562808513641,
      "learning_rate": 2.7061199999999998e-05,
      "loss": 0.0757,
      "step": 2450
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12910398840904236,
      "learning_rate": 2.7001199999999998e-05,
      "loss": 0.0688,
      "step": 2500
    },
    {
      "epoch": 0.2,
      "eval_loss": 0.08841572701931,
      "eval_runtime": 87.7555,
      "eval_samples_per_second": 22.791,
      "eval_steps_per_second": 5.698,
      "step": 2500
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.14937053620815277,
      "learning_rate": 2.69412e-05,
      "loss": 0.0727,
      "step": 2550
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.15660254657268524,
      "learning_rate": 2.68812e-05,
      "loss": 0.0656,
      "step": 2600
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.09593763947486877,
      "learning_rate": 2.68212e-05,
      "loss": 0.0726,
      "step": 2650
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.25192323327064514,
      "learning_rate": 2.67624e-05,
      "loss": 0.0753,
      "step": 2700
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1267642378807068,
      "learning_rate": 2.67024e-05,
      "loss": 0.0707,
      "step": 2750
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.13844658434391022,
      "learning_rate": 2.66424e-05,
      "loss": 0.0711,
      "step": 2800
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.15095186233520508,
      "learning_rate": 2.65824e-05,
      "loss": 0.0696,
      "step": 2850
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.09553442895412445,
      "learning_rate": 2.65224e-05,
      "loss": 0.0709,
      "step": 2900
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.21425922214984894,
      "learning_rate": 2.64624e-05,
      "loss": 0.0662,
      "step": 2950
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10706017166376114,
      "learning_rate": 2.64024e-05,
      "loss": 0.0721,
      "step": 3000
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.08755213767290115,
      "eval_runtime": 87.7619,
      "eval_samples_per_second": 22.789,
      "eval_steps_per_second": 5.697,
      "step": 3000
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.18366344273090363,
      "learning_rate": 2.63424e-05,
      "loss": 0.0781,
      "step": 3050
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.15975314378738403,
      "learning_rate": 2.62824e-05,
      "loss": 0.0771,
      "step": 3100
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.14510446786880493,
      "learning_rate": 2.6222399999999998e-05,
      "loss": 0.0734,
      "step": 3150
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.10040156543254852,
      "learning_rate": 2.61624e-05,
      "loss": 0.0624,
      "step": 3200
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1453912854194641,
      "learning_rate": 2.61024e-05,
      "loss": 0.0661,
      "step": 3250
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.13999666273593903,
      "learning_rate": 2.60424e-05,
      "loss": 0.0694,
      "step": 3300
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.13396582007408142,
      "learning_rate": 2.59824e-05,
      "loss": 0.0692,
      "step": 3350
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.1334969699382782,
      "learning_rate": 2.59224e-05,
      "loss": 0.0629,
      "step": 3400
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.16296976804733276,
      "learning_rate": 2.5862399999999998e-05,
      "loss": 0.0679,
      "step": 3450
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1321544647216797,
      "learning_rate": 2.58024e-05,
      "loss": 0.0764,
      "step": 3500
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.0869474709033966,
      "eval_runtime": 87.6679,
      "eval_samples_per_second": 22.813,
      "eval_steps_per_second": 5.703,
      "step": 3500
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.13956592977046967,
      "learning_rate": 2.57424e-05,
      "loss": 0.0749,
      "step": 3550
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.2318839579820633,
      "learning_rate": 2.56824e-05,
      "loss": 0.0658,
      "step": 3600
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.10267651081085205,
      "learning_rate": 2.56224e-05,
      "loss": 0.0651,
      "step": 3650
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.13903647661209106,
      "learning_rate": 2.5562399999999998e-05,
      "loss": 0.0684,
      "step": 3700
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.07369179278612137,
      "learning_rate": 2.5502399999999998e-05,
      "loss": 0.0643,
      "step": 3750
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.18283645808696747,
      "learning_rate": 2.54424e-05,
      "loss": 0.0745,
      "step": 3800
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.1031743511557579,
      "learning_rate": 2.5382400000000003e-05,
      "loss": 0.0678,
      "step": 3850
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.2246129959821701,
      "learning_rate": 2.5322400000000003e-05,
      "loss": 0.0701,
      "step": 3900
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.1434032917022705,
      "learning_rate": 2.52624e-05,
      "loss": 0.071,
      "step": 3950
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11730857193470001,
      "learning_rate": 2.52024e-05,
      "loss": 0.0696,
      "step": 4000
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.0867869183421135,
      "eval_runtime": 87.7801,
      "eval_samples_per_second": 22.784,
      "eval_steps_per_second": 5.696,
      "step": 4000
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.07968447357416153,
      "learning_rate": 2.5142400000000002e-05,
      "loss": 0.0635,
      "step": 4050
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.1641731858253479,
      "learning_rate": 2.5082400000000002e-05,
      "loss": 0.0667,
      "step": 4100
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.15083415806293488,
      "learning_rate": 2.5022400000000003e-05,
      "loss": 0.071,
      "step": 4150
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.16534006595611572,
      "learning_rate": 2.4962400000000003e-05,
      "loss": 0.0735,
      "step": 4200
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1629945933818817,
      "learning_rate": 2.49024e-05,
      "loss": 0.0641,
      "step": 4250
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.11393357813358307,
      "learning_rate": 2.48424e-05,
      "loss": 0.0679,
      "step": 4300
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.10581399500370026,
      "learning_rate": 2.4782400000000002e-05,
      "loss": 0.0666,
      "step": 4350
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.1571837067604065,
      "learning_rate": 2.4722400000000002e-05,
      "loss": 0.0619,
      "step": 4400
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.11944606155157089,
      "learning_rate": 2.4662400000000003e-05,
      "loss": 0.0666,
      "step": 4450
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.17318391799926758,
      "learning_rate": 2.46024e-05,
      "loss": 0.0684,
      "step": 4500
    },
    {
      "epoch": 0.36,
      "eval_loss": 0.08653330057859421,
      "eval_runtime": 87.7823,
      "eval_samples_per_second": 22.784,
      "eval_steps_per_second": 5.696,
      "step": 4500
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.09011202305555344,
      "learning_rate": 2.45424e-05,
      "loss": 0.0699,
      "step": 4550
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.1441943347454071,
      "learning_rate": 2.44824e-05,
      "loss": 0.0634,
      "step": 4600
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.17736917734146118,
      "learning_rate": 2.4422400000000002e-05,
      "loss": 0.0688,
      "step": 4650
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.16143649816513062,
      "learning_rate": 2.4362400000000002e-05,
      "loss": 0.0575,
      "step": 4700
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1789257675409317,
      "learning_rate": 2.43024e-05,
      "loss": 0.067,
      "step": 4750
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.13589045405387878,
      "learning_rate": 2.42424e-05,
      "loss": 0.0713,
      "step": 4800
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.12760789692401886,
      "learning_rate": 2.41824e-05,
      "loss": 0.0644,
      "step": 4850
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.17507490515708923,
      "learning_rate": 2.41224e-05,
      "loss": 0.0634,
      "step": 4900
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.15503354370594025,
      "learning_rate": 2.4062400000000002e-05,
      "loss": 0.0654,
      "step": 4950
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12309867143630981,
      "learning_rate": 2.4002400000000002e-05,
      "loss": 0.0698,
      "step": 5000
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.0860014408826828,
      "eval_runtime": 87.7287,
      "eval_samples_per_second": 22.798,
      "eval_steps_per_second": 5.699,
      "step": 5000
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.1547604650259018,
      "learning_rate": 2.39424e-05,
      "loss": 0.0691,
      "step": 5050
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.17712494730949402,
      "learning_rate": 2.38824e-05,
      "loss": 0.064,
      "step": 5100
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.16606800258159637,
      "learning_rate": 2.38224e-05,
      "loss": 0.0687,
      "step": 5150
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.18862195312976837,
      "learning_rate": 2.37624e-05,
      "loss": 0.0662,
      "step": 5200
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1778397411108017,
      "learning_rate": 2.3702400000000002e-05,
      "loss": 0.0632,
      "step": 5250
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.15363118052482605,
      "learning_rate": 2.36424e-05,
      "loss": 0.0617,
      "step": 5300
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.1690954864025116,
      "learning_rate": 2.35824e-05,
      "loss": 0.0717,
      "step": 5350
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.19384829699993134,
      "learning_rate": 2.35224e-05,
      "loss": 0.0669,
      "step": 5400
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.20759643614292145,
      "learning_rate": 2.34624e-05,
      "loss": 0.0662,
      "step": 5450
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16528765857219696,
      "learning_rate": 2.34024e-05,
      "loss": 0.0737,
      "step": 5500
    },
    {
      "epoch": 0.44,
      "eval_loss": 0.08539459854364395,
      "eval_runtime": 87.7109,
      "eval_samples_per_second": 22.802,
      "eval_steps_per_second": 5.701,
      "step": 5500
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.09632231295108795,
      "learning_rate": 2.3342400000000002e-05,
      "loss": 0.0621,
      "step": 5550
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.15080232918262482,
      "learning_rate": 2.32824e-05,
      "loss": 0.0674,
      "step": 5600
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.14794333279132843,
      "learning_rate": 2.32224e-05,
      "loss": 0.0662,
      "step": 5650
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.15014077723026276,
      "learning_rate": 2.31624e-05,
      "loss": 0.0574,
      "step": 5700
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.10953331738710403,
      "learning_rate": 2.31024e-05,
      "loss": 0.0684,
      "step": 5750
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.17434537410736084,
      "learning_rate": 2.30424e-05,
      "loss": 0.0687,
      "step": 5800
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.12245655059814453,
      "learning_rate": 2.29824e-05,
      "loss": 0.0689,
      "step": 5850
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.10118559747934341,
      "learning_rate": 2.29224e-05,
      "loss": 0.0609,
      "step": 5900
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.14768172800540924,
      "learning_rate": 2.28624e-05,
      "loss": 0.0717,
      "step": 5950
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.18510164320468903,
      "learning_rate": 2.28024e-05,
      "loss": 0.0716,
      "step": 6000
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.08568704128265381,
      "eval_runtime": 87.7082,
      "eval_samples_per_second": 22.803,
      "eval_steps_per_second": 5.701,
      "step": 6000
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.09803249686956406,
      "learning_rate": 2.27424e-05,
      "loss": 0.0693,
      "step": 6050
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.1030198410153389,
      "learning_rate": 2.2682399999999998e-05,
      "loss": 0.065,
      "step": 6100
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.15723823010921478,
      "learning_rate": 2.26224e-05,
      "loss": 0.063,
      "step": 6150
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.1828642040491104,
      "learning_rate": 2.25624e-05,
      "loss": 0.0665,
      "step": 6200
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11824677884578705,
      "learning_rate": 2.25024e-05,
      "loss": 0.0656,
      "step": 6250
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.07376304268836975,
      "learning_rate": 2.24424e-05,
      "loss": 0.0664,
      "step": 6300
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.09230540692806244,
      "learning_rate": 2.23824e-05,
      "loss": 0.0668,
      "step": 6350
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.1208396852016449,
      "learning_rate": 2.2322399999999998e-05,
      "loss": 0.0642,
      "step": 6400
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.12650001049041748,
      "learning_rate": 2.22624e-05,
      "loss": 0.0656,
      "step": 6450
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11461616307497025,
      "learning_rate": 2.22024e-05,
      "loss": 0.0676,
      "step": 6500
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.08497656136751175,
      "eval_runtime": 87.668,
      "eval_samples_per_second": 22.813,
      "eval_steps_per_second": 5.703,
      "step": 6500
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.24468739330768585,
      "learning_rate": 2.21424e-05,
      "loss": 0.0688,
      "step": 6550
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.17887485027313232,
      "learning_rate": 2.20824e-05,
      "loss": 0.0645,
      "step": 6600
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.12986980378627777,
      "learning_rate": 2.2022399999999998e-05,
      "loss": 0.0609,
      "step": 6650
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.25361281633377075,
      "learning_rate": 2.1962399999999998e-05,
      "loss": 0.0603,
      "step": 6700
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1815791130065918,
      "learning_rate": 2.19024e-05,
      "loss": 0.0659,
      "step": 6750
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.12782719731330872,
      "learning_rate": 2.18424e-05,
      "loss": 0.0641,
      "step": 6800
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.1801528036594391,
      "learning_rate": 2.17824e-05,
      "loss": 0.0666,
      "step": 6850
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.1247314065694809,
      "learning_rate": 2.17224e-05,
      "loss": 0.0592,
      "step": 6900
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.19411933422088623,
      "learning_rate": 2.16624e-05,
      "loss": 0.0688,
      "step": 6950
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11316727846860886,
      "learning_rate": 2.1602400000000002e-05,
      "loss": 0.0635,
      "step": 7000
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.08500248938798904,
      "eval_runtime": 87.7238,
      "eval_samples_per_second": 22.799,
      "eval_steps_per_second": 5.7,
      "step": 7000
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.14090943336486816,
      "learning_rate": 2.1542400000000002e-05,
      "loss": 0.0648,
      "step": 7050
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.16401338577270508,
      "learning_rate": 2.1482400000000003e-05,
      "loss": 0.064,
      "step": 7100
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.14420969784259796,
      "learning_rate": 2.1422400000000003e-05,
      "loss": 0.0627,
      "step": 7150
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.3247956931591034,
      "learning_rate": 2.13624e-05,
      "loss": 0.0621,
      "step": 7200
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1672086864709854,
      "learning_rate": 2.13024e-05,
      "loss": 0.0673,
      "step": 7250
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.16130150854587555,
      "learning_rate": 2.1242400000000002e-05,
      "loss": 0.0604,
      "step": 7300
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.201412633061409,
      "learning_rate": 2.1182400000000002e-05,
      "loss": 0.0687,
      "step": 7350
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.17173218727111816,
      "learning_rate": 2.1122400000000003e-05,
      "loss": 0.0626,
      "step": 7400
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.14102816581726074,
      "learning_rate": 2.10624e-05,
      "loss": 0.0662,
      "step": 7450
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12381123006343842,
      "learning_rate": 2.10024e-05,
      "loss": 0.0676,
      "step": 7500
    },
    {
      "epoch": 0.6,
      "eval_loss": 0.08412499725818634,
      "eval_runtime": 87.8551,
      "eval_samples_per_second": 22.765,
      "eval_steps_per_second": 5.691,
      "step": 7500
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.14278633892536163,
      "learning_rate": 2.09424e-05,
      "loss": 0.0665,
      "step": 7550
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.19127364456653595,
      "learning_rate": 2.0882400000000002e-05,
      "loss": 0.0652,
      "step": 7600
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.18026478588581085,
      "learning_rate": 2.0822400000000002e-05,
      "loss": 0.0673,
      "step": 7650
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.1525663435459137,
      "learning_rate": 2.0762400000000003e-05,
      "loss": 0.0596,
      "step": 7700
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1647537797689438,
      "learning_rate": 2.07024e-05,
      "loss": 0.0638,
      "step": 7750
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.14861586689949036,
      "learning_rate": 2.06424e-05,
      "loss": 0.0616,
      "step": 7800
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.13468614220619202,
      "learning_rate": 2.05824e-05,
      "loss": 0.0645,
      "step": 7850
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.22103475034236908,
      "learning_rate": 2.0522400000000002e-05,
      "loss": 0.0618,
      "step": 7900
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.21467621624469757,
      "learning_rate": 2.0462400000000002e-05,
      "loss": 0.0609,
      "step": 7950
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.09938893467187881,
      "learning_rate": 2.04024e-05,
      "loss": 0.0671,
      "step": 8000
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.08401757478713989,
      "eval_runtime": 88.0793,
      "eval_samples_per_second": 22.707,
      "eval_steps_per_second": 5.677,
      "step": 8000
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.09193145483732224,
      "learning_rate": 2.03424e-05,
      "loss": 0.0592,
      "step": 8050
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.13023436069488525,
      "learning_rate": 2.02824e-05,
      "loss": 0.0627,
      "step": 8100
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.12572939693927765,
      "learning_rate": 2.02224e-05,
      "loss": 0.0696,
      "step": 8150
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.08949209004640579,
      "learning_rate": 2.0162400000000002e-05,
      "loss": 0.0633,
      "step": 8200
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.18614652752876282,
      "learning_rate": 2.01024e-05,
      "loss": 0.0683,
      "step": 8250
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.1969350129365921,
      "learning_rate": 2.00424e-05,
      "loss": 0.0629,
      "step": 8300
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.18870118260383606,
      "learning_rate": 1.99824e-05,
      "loss": 0.0697,
      "step": 8350
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.13722488284111023,
      "learning_rate": 1.99224e-05,
      "loss": 0.0654,
      "step": 8400
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.12897425889968872,
      "learning_rate": 1.98624e-05,
      "loss": 0.0632,
      "step": 8450
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.16306981444358826,
      "learning_rate": 1.9802400000000002e-05,
      "loss": 0.0606,
      "step": 8500
    },
    {
      "epoch": 0.68,
      "eval_loss": 0.08417751640081406,
      "eval_runtime": 88.0586,
      "eval_samples_per_second": 22.712,
      "eval_steps_per_second": 5.678,
      "step": 8500
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.113871268928051,
      "learning_rate": 1.97424e-05,
      "loss": 0.0651,
      "step": 8550
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.12488531321287155,
      "learning_rate": 1.96824e-05,
      "loss": 0.0624,
      "step": 8600
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.16593888401985168,
      "learning_rate": 1.96224e-05,
      "loss": 0.0619,
      "step": 8650
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.14381399750709534,
      "learning_rate": 1.95624e-05,
      "loss": 0.0631,
      "step": 8700
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.19177569448947906,
      "learning_rate": 1.95024e-05,
      "loss": 0.0584,
      "step": 8750
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.18918590247631073,
      "learning_rate": 1.94424e-05,
      "loss": 0.0563,
      "step": 8800
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.1318553388118744,
      "learning_rate": 1.93824e-05,
      "loss": 0.0644,
      "step": 8850
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.11326153576374054,
      "learning_rate": 1.93224e-05,
      "loss": 0.0678,
      "step": 8900
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.1167350560426712,
      "learning_rate": 1.92624e-05,
      "loss": 0.0558,
      "step": 8950
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.10959301143884659,
      "learning_rate": 1.92024e-05,
      "loss": 0.0618,
      "step": 9000
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.0838891863822937,
      "eval_runtime": 88.0828,
      "eval_samples_per_second": 22.706,
      "eval_steps_per_second": 5.676,
      "step": 9000
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.152841717004776,
      "learning_rate": 1.91424e-05,
      "loss": 0.0611,
      "step": 9050
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.16737787425518036,
      "learning_rate": 1.90824e-05,
      "loss": 0.0626,
      "step": 9100
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.1456371396780014,
      "learning_rate": 1.90224e-05,
      "loss": 0.0599,
      "step": 9150
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.1772635579109192,
      "learning_rate": 1.89624e-05,
      "loss": 0.0577,
      "step": 9200
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11802179366350174,
      "learning_rate": 1.89024e-05,
      "loss": 0.0645,
      "step": 9250
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.1235092505812645,
      "learning_rate": 1.88424e-05,
      "loss": 0.0584,
      "step": 9300
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.11901592463254929,
      "learning_rate": 1.8782399999999998e-05,
      "loss": 0.06,
      "step": 9350
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.09477788209915161,
      "learning_rate": 1.87224e-05,
      "loss": 0.0637,
      "step": 9400
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.12917232513427734,
      "learning_rate": 1.86624e-05,
      "loss": 0.0648,
      "step": 9450
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.17290474474430084,
      "learning_rate": 1.86024e-05,
      "loss": 0.0638,
      "step": 9500
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.08371420204639435,
      "eval_runtime": 88.0684,
      "eval_samples_per_second": 22.71,
      "eval_steps_per_second": 5.677,
      "step": 9500
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.07568900287151337,
      "learning_rate": 1.85424e-05,
      "loss": 0.0648,
      "step": 9550
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.11715036630630493,
      "learning_rate": 1.8482399999999998e-05,
      "loss": 0.0644,
      "step": 9600
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.18466413021087646,
      "learning_rate": 1.84224e-05,
      "loss": 0.0641,
      "step": 9650
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.15552838146686554,
      "learning_rate": 1.83624e-05,
      "loss": 0.0538,
      "step": 9700
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1408630758523941,
      "learning_rate": 1.83024e-05,
      "loss": 0.0589,
      "step": 9750
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.1492939293384552,
      "learning_rate": 1.82424e-05,
      "loss": 0.0599,
      "step": 9800
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.08888087421655655,
      "learning_rate": 1.81824e-05,
      "loss": 0.0591,
      "step": 9850
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.11865179240703583,
      "learning_rate": 1.8122399999999998e-05,
      "loss": 0.0581,
      "step": 9900
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.18105269968509674,
      "learning_rate": 1.80624e-05,
      "loss": 0.0592,
      "step": 9950
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11952503770589828,
      "learning_rate": 1.80024e-05,
      "loss": 0.0567,
      "step": 10000
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.08371368050575256,
      "eval_runtime": 88.0598,
      "eval_samples_per_second": 22.712,
      "eval_steps_per_second": 5.678,
      "step": 10000
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.16905057430267334,
      "learning_rate": 1.79424e-05,
      "loss": 0.065,
      "step": 10050
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.11961635947227478,
      "learning_rate": 1.7882400000000003e-05,
      "loss": 0.0677,
      "step": 10100
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.13067485392093658,
      "learning_rate": 1.78224e-05,
      "loss": 0.0611,
      "step": 10150
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.12020887434482574,
      "learning_rate": 1.77624e-05,
      "loss": 0.0582,
      "step": 10200
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1354004591703415,
      "learning_rate": 1.7702400000000002e-05,
      "loss": 0.0592,
      "step": 10250
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.22322942316532135,
      "learning_rate": 1.7642400000000002e-05,
      "loss": 0.0644,
      "step": 10300
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.12816545367240906,
      "learning_rate": 1.7582400000000003e-05,
      "loss": 0.065,
      "step": 10350
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.07579780369997025,
      "learning_rate": 1.75224e-05,
      "loss": 0.057,
      "step": 10400
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.10635198652744293,
      "learning_rate": 1.74624e-05,
      "loss": 0.062,
      "step": 10450
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.10341114550828934,
      "learning_rate": 1.74024e-05,
      "loss": 0.0593,
      "step": 10500
    },
    {
      "epoch": 0.84,
      "eval_loss": 0.08319947868585587,
      "eval_runtime": 88.0121,
      "eval_samples_per_second": 22.724,
      "eval_steps_per_second": 5.681,
      "step": 10500
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.16710689663887024,
      "learning_rate": 1.7342400000000002e-05,
      "loss": 0.0616,
      "step": 10550
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.2513883411884308,
      "learning_rate": 1.7282400000000002e-05,
      "loss": 0.0638,
      "step": 10600
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.182452991604805,
      "learning_rate": 1.7222400000000003e-05,
      "loss": 0.0588,
      "step": 10650
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.16390974819660187,
      "learning_rate": 1.71624e-05,
      "loss": 0.059,
      "step": 10700
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13543175160884857,
      "learning_rate": 1.71024e-05,
      "loss": 0.0648,
      "step": 10750
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.12370152771472931,
      "learning_rate": 1.70424e-05,
      "loss": 0.0625,
      "step": 10800
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.23662041127681732,
      "learning_rate": 1.6982400000000002e-05,
      "loss": 0.0667,
      "step": 10850
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.1502092331647873,
      "learning_rate": 1.6922400000000003e-05,
      "loss": 0.0652,
      "step": 10900
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.1745540052652359,
      "learning_rate": 1.68624e-05,
      "loss": 0.063,
      "step": 10950
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20543019473552704,
      "learning_rate": 1.68024e-05,
      "loss": 0.0605,
      "step": 11000
    },
    {
      "epoch": 0.88,
      "eval_loss": 0.08301527053117752,
      "eval_runtime": 88.0655,
      "eval_samples_per_second": 22.71,
      "eval_steps_per_second": 5.678,
      "step": 11000
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.22617511451244354,
      "learning_rate": 1.67424e-05,
      "loss": 0.066,
      "step": 11050
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.15583598613739014,
      "learning_rate": 1.66824e-05,
      "loss": 0.0663,
      "step": 11100
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.06502597779035568,
      "learning_rate": 1.6622400000000002e-05,
      "loss": 0.0611,
      "step": 11150
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.1777944415807724,
      "learning_rate": 1.6562400000000003e-05,
      "loss": 0.0618,
      "step": 11200
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14777645468711853,
      "learning_rate": 1.65024e-05,
      "loss": 0.0616,
      "step": 11250
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.1988598108291626,
      "learning_rate": 1.64424e-05,
      "loss": 0.0683,
      "step": 11300
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.16424456238746643,
      "learning_rate": 1.63824e-05,
      "loss": 0.0616,
      "step": 11350
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.17275363206863403,
      "learning_rate": 1.63224e-05,
      "loss": 0.0603,
      "step": 11400
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.1208202987909317,
      "learning_rate": 1.6262400000000002e-05,
      "loss": 0.0623,
      "step": 11450
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12853899598121643,
      "learning_rate": 1.62024e-05,
      "loss": 0.0597,
      "step": 11500
    },
    {
      "epoch": 0.92,
      "eval_loss": 0.08293735980987549,
      "eval_runtime": 88.063,
      "eval_samples_per_second": 22.711,
      "eval_steps_per_second": 5.678,
      "step": 11500
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.12364531308412552,
      "learning_rate": 1.61424e-05,
      "loss": 0.0624,
      "step": 11550
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.15913207828998566,
      "learning_rate": 1.60824e-05,
      "loss": 0.061,
      "step": 11600
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.16585472226142883,
      "learning_rate": 1.60224e-05,
      "loss": 0.066,
      "step": 11650
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.13831599056720734,
      "learning_rate": 1.59624e-05,
      "loss": 0.0578,
      "step": 11700
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1819075047969818,
      "learning_rate": 1.59024e-05,
      "loss": 0.0593,
      "step": 11750
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.16260235011577606,
      "learning_rate": 1.58424e-05,
      "loss": 0.0647,
      "step": 11800
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.1420051008462906,
      "learning_rate": 1.57824e-05,
      "loss": 0.0556,
      "step": 11850
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.17357395589351654,
      "learning_rate": 1.57224e-05,
      "loss": 0.0585,
      "step": 11900
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.1465808004140854,
      "learning_rate": 1.56624e-05,
      "loss": 0.0633,
      "step": 11950
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18425939977169037,
      "learning_rate": 1.56024e-05,
      "loss": 0.061,
      "step": 12000
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.082674041390419,
      "eval_runtime": 88.0509,
      "eval_samples_per_second": 22.714,
      "eval_steps_per_second": 5.679,
      "step": 12000
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.1296168565750122,
      "learning_rate": 1.55424e-05,
      "loss": 0.0638,
      "step": 12050
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.14450989663600922,
      "learning_rate": 1.54824e-05,
      "loss": 0.0623,
      "step": 12100
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.09367354214191437,
      "learning_rate": 1.54224e-05,
      "loss": 0.0618,
      "step": 12150
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.13332900404930115,
      "learning_rate": 1.53624e-05,
      "loss": 0.0608,
      "step": 12200
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.14300012588500977,
      "learning_rate": 1.53024e-05,
      "loss": 0.0695,
      "step": 12250
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.14829818904399872,
      "learning_rate": 1.52424e-05,
      "loss": 0.0589,
      "step": 12300
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.08889272063970566,
      "learning_rate": 1.5182399999999999e-05,
      "loss": 0.0578,
      "step": 12350
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.18994523584842682,
      "learning_rate": 1.51224e-05,
      "loss": 0.0667,
      "step": 12400
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.16152743995189667,
      "learning_rate": 1.50624e-05,
      "loss": 0.0552,
      "step": 12450
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2399597018957138,
      "learning_rate": 1.5002399999999999e-05,
      "loss": 0.0632,
      "step": 12500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.0824647843837738,
      "eval_runtime": 88.0514,
      "eval_samples_per_second": 22.714,
      "eval_steps_per_second": 5.678,
      "step": 12500
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.12758083641529083,
      "learning_rate": 1.4942400000000001e-05,
      "loss": 0.0553,
      "step": 12550
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.14093191921710968,
      "learning_rate": 1.4882400000000002e-05,
      "loss": 0.0572,
      "step": 12600
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.20115911960601807,
      "learning_rate": 1.48224e-05,
      "loss": 0.055,
      "step": 12650
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.2974820137023926,
      "learning_rate": 1.4762400000000001e-05,
      "loss": 0.0548,
      "step": 12700
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.06170056387782097,
      "learning_rate": 1.47024e-05,
      "loss": 0.0528,
      "step": 12750
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.18193961679935455,
      "learning_rate": 1.46424e-05,
      "loss": 0.0551,
      "step": 12800
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.11086717993021011,
      "learning_rate": 1.4582400000000001e-05,
      "loss": 0.0533,
      "step": 12850
    },
    {
      "epoch": 1.032,
      "grad_norm": 0.1452319473028183,
      "learning_rate": 1.45224e-05,
      "loss": 0.0525,
      "step": 12900
    },
    {
      "epoch": 1.036,
      "grad_norm": 0.16397640109062195,
      "learning_rate": 1.44624e-05,
      "loss": 0.0622,
      "step": 12950
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.09709367156028748,
      "learning_rate": 1.4402400000000001e-05,
      "loss": 0.0559,
      "step": 13000
    },
    {
      "epoch": 1.04,
      "eval_loss": 0.08302706480026245,
      "eval_runtime": 88.0122,
      "eval_samples_per_second": 22.724,
      "eval_steps_per_second": 5.681,
      "step": 13000
    },
    {
      "epoch": 1.044,
      "grad_norm": 0.09841930866241455,
      "learning_rate": 1.43424e-05,
      "loss": 0.0534,
      "step": 13050
    },
    {
      "epoch": 1.048,
      "grad_norm": 0.1224198266863823,
      "learning_rate": 1.42824e-05,
      "loss": 0.0564,
      "step": 13100
    },
    {
      "epoch": 1.052,
      "grad_norm": 0.17220191657543182,
      "learning_rate": 1.42224e-05,
      "loss": 0.0567,
      "step": 13150
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.08674409985542297,
      "learning_rate": 1.41636e-05,
      "loss": 0.0508,
      "step": 13200
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.12293367087841034,
      "learning_rate": 1.41036e-05,
      "loss": 0.0541,
      "step": 13250
    },
    {
      "epoch": 1.064,
      "grad_norm": 0.12341846525669098,
      "learning_rate": 1.40436e-05,
      "loss": 0.055,
      "step": 13300
    },
    {
      "epoch": 1.068,
      "grad_norm": 0.10165009647607803,
      "learning_rate": 1.39836e-05,
      "loss": 0.0589,
      "step": 13350
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.18138067424297333,
      "learning_rate": 1.39236e-05,
      "loss": 0.053,
      "step": 13400
    },
    {
      "epoch": 1.076,
      "grad_norm": 0.18615098297595978,
      "learning_rate": 1.3863599999999999e-05,
      "loss": 0.0558,
      "step": 13450
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.06511889398097992,
      "learning_rate": 1.38036e-05,
      "loss": 0.062,
      "step": 13500
    },
    {
      "epoch": 1.08,
      "eval_loss": 0.08285626024007797,
      "eval_runtime": 88.0463,
      "eval_samples_per_second": 22.715,
      "eval_steps_per_second": 5.679,
      "step": 13500
    },
    {
      "epoch": 1.084,
      "grad_norm": 0.1615404486656189,
      "learning_rate": 1.3743600000000002e-05,
      "loss": 0.0572,
      "step": 13550
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.1500886082649231,
      "learning_rate": 1.36836e-05,
      "loss": 0.0557,
      "step": 13600
    },
    {
      "epoch": 1.092,
      "grad_norm": 0.08288303017616272,
      "learning_rate": 1.3623600000000001e-05,
      "loss": 0.0567,
      "step": 13650
    },
    {
      "epoch": 1.096,
      "grad_norm": 0.12978018820285797,
      "learning_rate": 1.3563600000000002e-05,
      "loss": 0.051,
      "step": 13700
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.17383359372615814,
      "learning_rate": 1.35036e-05,
      "loss": 0.0527,
      "step": 13750
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.21195685863494873,
      "learning_rate": 1.3443600000000001e-05,
      "loss": 0.0526,
      "step": 13800
    },
    {
      "epoch": 1.108,
      "grad_norm": 0.15831385552883148,
      "learning_rate": 1.33836e-05,
      "loss": 0.0595,
      "step": 13850
    },
    {
      "epoch": 1.112,
      "grad_norm": 0.23136693239212036,
      "learning_rate": 1.33236e-05,
      "loss": 0.0523,
      "step": 13900
    },
    {
      "epoch": 1.116,
      "grad_norm": 0.12520194053649902,
      "learning_rate": 1.3263600000000001e-05,
      "loss": 0.0603,
      "step": 13950
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.10329103469848633,
      "learning_rate": 1.32036e-05,
      "loss": 0.0494,
      "step": 14000
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.083070769906044,
      "eval_runtime": 88.0459,
      "eval_samples_per_second": 22.715,
      "eval_steps_per_second": 5.679,
      "step": 14000
    },
    {
      "epoch": 1.124,
      "grad_norm": 0.17533883452415466,
      "learning_rate": 1.31436e-05,
      "loss": 0.0552,
      "step": 14050
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 0.10817945748567581,
      "learning_rate": 1.3083600000000001e-05,
      "loss": 0.049,
      "step": 14100
    },
    {
      "epoch": 1.1320000000000001,
      "grad_norm": 0.09337913990020752,
      "learning_rate": 1.30236e-05,
      "loss": 0.0573,
      "step": 14150
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.15710942447185516,
      "learning_rate": 1.29636e-05,
      "loss": 0.0605,
      "step": 14200
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 0.10915792733430862,
      "learning_rate": 1.29036e-05,
      "loss": 0.0581,
      "step": 14250
    },
    {
      "epoch": 1.144,
      "grad_norm": 0.10125772655010223,
      "learning_rate": 1.28436e-05,
      "loss": 0.0599,
      "step": 14300
    },
    {
      "epoch": 1.148,
      "grad_norm": 0.0998956710100174,
      "learning_rate": 1.27836e-05,
      "loss": 0.0479,
      "step": 14350
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.13762612640857697,
      "learning_rate": 1.27236e-05,
      "loss": 0.0589,
      "step": 14400
    },
    {
      "epoch": 1.156,
      "grad_norm": 0.11048023402690887,
      "learning_rate": 1.26636e-05,
      "loss": 0.0591,
      "step": 14450
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.14803436398506165,
      "learning_rate": 1.26036e-05,
      "loss": 0.0553,
      "step": 14500
    },
    {
      "epoch": 1.16,
      "eval_loss": 0.08270228654146194,
      "eval_runtime": 88.0514,
      "eval_samples_per_second": 22.714,
      "eval_steps_per_second": 5.678,
      "step": 14500
    },
    {
      "epoch": 1.164,
      "grad_norm": 0.06833196431398392,
      "learning_rate": 1.25436e-05,
      "loss": 0.0586,
      "step": 14550
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.16051365435123444,
      "learning_rate": 1.24836e-05,
      "loss": 0.056,
      "step": 14600
    },
    {
      "epoch": 1.172,
      "grad_norm": 0.18909616768360138,
      "learning_rate": 1.24236e-05,
      "loss": 0.057,
      "step": 14650
    },
    {
      "epoch": 1.176,
      "grad_norm": 0.17333486676216125,
      "learning_rate": 1.23636e-05,
      "loss": 0.0562,
      "step": 14700
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0701974630355835,
      "learning_rate": 1.23036e-05,
      "loss": 0.0545,
      "step": 14750
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.1582074612379074,
      "learning_rate": 1.2243599999999999e-05,
      "loss": 0.06,
      "step": 14800
    },
    {
      "epoch": 1.188,
      "grad_norm": 0.13948781788349152,
      "learning_rate": 1.21836e-05,
      "loss": 0.0529,
      "step": 14850
    },
    {
      "epoch": 1.192,
      "grad_norm": 0.0946699008345604,
      "learning_rate": 1.21236e-05,
      "loss": 0.0668,
      "step": 14900
    },
    {
      "epoch": 1.196,
      "grad_norm": 0.1089014783501625,
      "learning_rate": 1.20636e-05,
      "loss": 0.0562,
      "step": 14950
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.09682592004537582,
      "learning_rate": 1.20036e-05,
      "loss": 0.0548,
      "step": 15000
    },
    {
      "epoch": 1.2,
      "eval_loss": 0.08243728429079056,
      "eval_runtime": 88.0302,
      "eval_samples_per_second": 22.719,
      "eval_steps_per_second": 5.68,
      "step": 15000
    },
    {
      "epoch": 1.204,
      "grad_norm": 0.15492039918899536,
      "learning_rate": 1.19436e-05,
      "loss": 0.0535,
      "step": 15050
    },
    {
      "epoch": 1.208,
      "grad_norm": 0.0744326263666153,
      "learning_rate": 1.1883600000000001e-05,
      "loss": 0.0572,
      "step": 15100
    },
    {
      "epoch": 1.212,
      "grad_norm": 0.20596753060817719,
      "learning_rate": 1.1823600000000001e-05,
      "loss": 0.0539,
      "step": 15150
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.15914830565452576,
      "learning_rate": 1.17636e-05,
      "loss": 0.0583,
      "step": 15200
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.16685660183429718,
      "learning_rate": 1.1703600000000001e-05,
      "loss": 0.0566,
      "step": 15250
    },
    {
      "epoch": 1.224,
      "grad_norm": 0.11385693401098251,
      "learning_rate": 1.1643600000000001e-05,
      "loss": 0.0542,
      "step": 15300
    },
    {
      "epoch": 1.228,
      "grad_norm": 0.12850815057754517,
      "learning_rate": 1.15836e-05,
      "loss": 0.0622,
      "step": 15350
    },
    {
      "epoch": 1.232,
      "grad_norm": 0.17093044519424438,
      "learning_rate": 1.1523600000000001e-05,
      "loss": 0.0536,
      "step": 15400
    },
    {
      "epoch": 1.236,
      "grad_norm": 0.18593502044677734,
      "learning_rate": 1.1463600000000002e-05,
      "loss": 0.0541,
      "step": 15450
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.16509443521499634,
      "learning_rate": 1.14036e-05,
      "loss": 0.0501,
      "step": 15500
    },
    {
      "epoch": 1.24,
      "eval_loss": 0.08269675076007843,
      "eval_runtime": 88.0811,
      "eval_samples_per_second": 22.706,
      "eval_steps_per_second": 5.677,
      "step": 15500
    },
    {
      "epoch": 1.244,
      "grad_norm": 0.15668867528438568,
      "learning_rate": 1.1343600000000001e-05,
      "loss": 0.0534,
      "step": 15550
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.08703252673149109,
      "learning_rate": 1.12836e-05,
      "loss": 0.0534,
      "step": 15600
    },
    {
      "epoch": 1.252,
      "grad_norm": 0.1255914270877838,
      "learning_rate": 1.12248e-05,
      "loss": 0.0514,
      "step": 15650
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.11858811974525452,
      "learning_rate": 1.11648e-05,
      "loss": 0.0527,
      "step": 15700
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.13111431896686554,
      "learning_rate": 1.1104799999999999e-05,
      "loss": 0.0494,
      "step": 15750
    },
    {
      "epoch": 1.264,
      "grad_norm": 0.17452576756477356,
      "learning_rate": 1.10448e-05,
      "loss": 0.0562,
      "step": 15800
    },
    {
      "epoch": 1.268,
      "grad_norm": 0.13699005544185638,
      "learning_rate": 1.09848e-05,
      "loss": 0.051,
      "step": 15850
    },
    {
      "epoch": 1.272,
      "grad_norm": 0.18750061094760895,
      "learning_rate": 1.0924799999999999e-05,
      "loss": 0.0545,
      "step": 15900
    },
    {
      "epoch": 1.276,
      "grad_norm": 0.1259056031703949,
      "learning_rate": 1.08648e-05,
      "loss": 0.0529,
      "step": 15950
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.10538329184055328,
      "learning_rate": 1.08048e-05,
      "loss": 0.058,
      "step": 16000
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.08262528479099274,
      "eval_runtime": 88.074,
      "eval_samples_per_second": 22.708,
      "eval_steps_per_second": 5.677,
      "step": 16000
    },
    {
      "epoch": 1.284,
      "grad_norm": 0.14214898645877838,
      "learning_rate": 1.07448e-05,
      "loss": 0.056,
      "step": 16050
    },
    {
      "epoch": 1.288,
      "grad_norm": 0.13236215710639954,
      "learning_rate": 1.06848e-05,
      "loss": 0.0566,
      "step": 16100
    },
    {
      "epoch": 1.292,
      "grad_norm": 0.1475580334663391,
      "learning_rate": 1.06248e-05,
      "loss": 0.0507,
      "step": 16150
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.13150113821029663,
      "learning_rate": 1.05648e-05,
      "loss": 0.0488,
      "step": 16200
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.0869784876704216,
      "learning_rate": 1.0504800000000001e-05,
      "loss": 0.0498,
      "step": 16250
    },
    {
      "epoch": 1.304,
      "grad_norm": 0.15695451200008392,
      "learning_rate": 1.04448e-05,
      "loss": 0.0517,
      "step": 16300
    },
    {
      "epoch": 1.308,
      "grad_norm": 0.1383635401725769,
      "learning_rate": 1.0384800000000001e-05,
      "loss": 0.0509,
      "step": 16350
    },
    {
      "epoch": 1.312,
      "grad_norm": 0.17300955951213837,
      "learning_rate": 1.0324800000000002e-05,
      "loss": 0.0587,
      "step": 16400
    },
    {
      "epoch": 1.316,
      "grad_norm": 0.1608356237411499,
      "learning_rate": 1.02648e-05,
      "loss": 0.0523,
      "step": 16450
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.1598045974969864,
      "learning_rate": 1.0204800000000001e-05,
      "loss": 0.0524,
      "step": 16500
    },
    {
      "epoch": 1.32,
      "eval_loss": 0.08258219808340073,
      "eval_runtime": 88.0858,
      "eval_samples_per_second": 22.705,
      "eval_steps_per_second": 5.676,
      "step": 16500
    },
    {
      "epoch": 1.324,
      "grad_norm": 0.10573418438434601,
      "learning_rate": 1.0144800000000002e-05,
      "loss": 0.058,
      "step": 16550
    },
    {
      "epoch": 1.328,
      "grad_norm": 0.1370289921760559,
      "learning_rate": 1.00848e-05,
      "loss": 0.0552,
      "step": 16600
    },
    {
      "epoch": 1.332,
      "grad_norm": 0.15583133697509766,
      "learning_rate": 1.0024800000000001e-05,
      "loss": 0.0505,
      "step": 16650
    },
    {
      "epoch": 1.336,
      "grad_norm": 0.12616191804409027,
      "learning_rate": 9.9648e-06,
      "loss": 0.0563,
      "step": 16700
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.1533484160900116,
      "learning_rate": 9.9048e-06,
      "loss": 0.0564,
      "step": 16750
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.11609365046024323,
      "learning_rate": 9.844800000000001e-06,
      "loss": 0.0538,
      "step": 16800
    },
    {
      "epoch": 1.3479999999999999,
      "grad_norm": 0.12423556298017502,
      "learning_rate": 9.7848e-06,
      "loss": 0.0519,
      "step": 16850
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 0.13122333586215973,
      "learning_rate": 9.7248e-06,
      "loss": 0.0518,
      "step": 16900
    },
    {
      "epoch": 1.3559999999999999,
      "grad_norm": 0.17409804463386536,
      "learning_rate": 9.664800000000001e-06,
      "loss": 0.0544,
      "step": 16950
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.13961808383464813,
      "learning_rate": 9.6048e-06,
      "loss": 0.0523,
      "step": 17000
    },
    {
      "epoch": 1.3599999999999999,
      "eval_loss": 0.08219126611948013,
      "eval_runtime": 88.051,
      "eval_samples_per_second": 22.714,
      "eval_steps_per_second": 5.679,
      "step": 17000
    },
    {
      "epoch": 1.3639999999999999,
      "grad_norm": 0.19439847767353058,
      "learning_rate": 9.5448e-06,
      "loss": 0.0566,
      "step": 17050
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 0.1604561060667038,
      "learning_rate": 9.4848e-06,
      "loss": 0.0532,
      "step": 17100
    },
    {
      "epoch": 1.3719999999999999,
      "grad_norm": 0.13188685476779938,
      "learning_rate": 9.4248e-06,
      "loss": 0.0547,
      "step": 17150
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.09787939488887787,
      "learning_rate": 9.3648e-06,
      "loss": 0.0476,
      "step": 17200
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.3960016369819641,
      "learning_rate": 9.3048e-06,
      "loss": 0.0607,
      "step": 17250
    },
    {
      "epoch": 1.384,
      "grad_norm": 0.17494530975818634,
      "learning_rate": 9.2448e-06,
      "loss": 0.0579,
      "step": 17300
    },
    {
      "epoch": 1.388,
      "grad_norm": 0.17870362102985382,
      "learning_rate": 9.1848e-06,
      "loss": 0.0481,
      "step": 17350
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.1152658611536026,
      "learning_rate": 9.1248e-06,
      "loss": 0.0507,
      "step": 17400
    },
    {
      "epoch": 1.396,
      "grad_norm": 0.10560191422700882,
      "learning_rate": 9.0648e-06,
      "loss": 0.0509,
      "step": 17450
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.09202779084444046,
      "learning_rate": 9.004799999999999e-06,
      "loss": 0.0571,
      "step": 17500
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.0821109265089035,
      "eval_runtime": 88.0678,
      "eval_samples_per_second": 22.71,
      "eval_steps_per_second": 5.677,
      "step": 17500
    },
    {
      "epoch": 1.404,
      "grad_norm": 0.1634778082370758,
      "learning_rate": 8.9448e-06,
      "loss": 0.0532,
      "step": 17550
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.12161219865083694,
      "learning_rate": 8.8848e-06,
      "loss": 0.0541,
      "step": 17600
    },
    {
      "epoch": 1.412,
      "grad_norm": 0.15347328782081604,
      "learning_rate": 8.824799999999999e-06,
      "loss": 0.0535,
      "step": 17650
    },
    {
      "epoch": 1.416,
      "grad_norm": 0.07917541265487671,
      "learning_rate": 8.7648e-06,
      "loss": 0.0552,
      "step": 17700
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.15406325459480286,
      "learning_rate": 8.7048e-06,
      "loss": 0.0584,
      "step": 17750
    },
    {
      "epoch": 1.424,
      "grad_norm": 0.21300417184829712,
      "learning_rate": 8.6448e-06,
      "loss": 0.0565,
      "step": 17800
    },
    {
      "epoch": 1.428,
      "grad_norm": 0.20691223442554474,
      "learning_rate": 8.584800000000001e-06,
      "loss": 0.0504,
      "step": 17850
    },
    {
      "epoch": 1.432,
      "grad_norm": 0.1366143375635147,
      "learning_rate": 8.5248e-06,
      "loss": 0.0573,
      "step": 17900
    },
    {
      "epoch": 1.436,
      "grad_norm": 0.11994505673646927,
      "learning_rate": 8.4648e-06,
      "loss": 0.0576,
      "step": 17950
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.06461376696825027,
      "learning_rate": 8.404800000000001e-06,
      "loss": 0.0595,
      "step": 18000
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.08217783272266388,
      "eval_runtime": 88.0706,
      "eval_samples_per_second": 22.709,
      "eval_steps_per_second": 5.677,
      "step": 18000
    },
    {
      "epoch": 1.444,
      "grad_norm": 0.150528684258461,
      "learning_rate": 8.3448e-06,
      "loss": 0.0519,
      "step": 18050
    },
    {
      "epoch": 1.448,
      "grad_norm": 0.08878117799758911,
      "learning_rate": 8.2848e-06,
      "loss": 0.0544,
      "step": 18100
    },
    {
      "epoch": 1.452,
      "grad_norm": 0.12337563931941986,
      "learning_rate": 8.224800000000001e-06,
      "loss": 0.0509,
      "step": 18150
    },
    {
      "epoch": 1.456,
      "grad_norm": 0.12858565151691437,
      "learning_rate": 8.1648e-06,
      "loss": 0.0561,
      "step": 18200
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.14318782091140747,
      "learning_rate": 8.1048e-06,
      "loss": 0.0539,
      "step": 18250
    },
    {
      "epoch": 1.464,
      "grad_norm": 0.16946159303188324,
      "learning_rate": 8.044800000000001e-06,
      "loss": 0.0545,
      "step": 18300
    },
    {
      "epoch": 1.468,
      "grad_norm": 0.11605024337768555,
      "learning_rate": 7.9848e-06,
      "loss": 0.0563,
      "step": 18350
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.20185694098472595,
      "learning_rate": 7.9248e-06,
      "loss": 0.0536,
      "step": 18400
    },
    {
      "epoch": 1.476,
      "grad_norm": 0.16189858317375183,
      "learning_rate": 7.8648e-06,
      "loss": 0.0523,
      "step": 18450
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.24366913735866547,
      "learning_rate": 7.8048e-06,
      "loss": 0.057,
      "step": 18500
    },
    {
      "epoch": 1.48,
      "eval_loss": 0.08215340971946716,
      "eval_runtime": 88.0726,
      "eval_samples_per_second": 22.709,
      "eval_steps_per_second": 5.677,
      "step": 18500
    },
    {
      "epoch": 1.484,
      "grad_norm": 0.13516181707382202,
      "learning_rate": 7.7448e-06,
      "loss": 0.0516,
      "step": 18550
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.14996372163295746,
      "learning_rate": 7.6848e-06,
      "loss": 0.0545,
      "step": 18600
    },
    {
      "epoch": 1.492,
      "grad_norm": 0.11676355451345444,
      "learning_rate": 7.6248e-06,
      "loss": 0.0546,
      "step": 18650
    },
    {
      "epoch": 1.496,
      "grad_norm": 0.0986240953207016,
      "learning_rate": 7.5648e-06,
      "loss": 0.054,
      "step": 18700
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.16591283679008484,
      "learning_rate": 7.5048e-06,
      "loss": 0.0554,
      "step": 18750
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.13870297372341156,
      "learning_rate": 7.4448e-06,
      "loss": 0.0486,
      "step": 18800
    },
    {
      "epoch": 1.508,
      "grad_norm": 0.1334107369184494,
      "learning_rate": 7.3848e-06,
      "loss": 0.0532,
      "step": 18850
    },
    {
      "epoch": 1.512,
      "grad_norm": 0.1280679702758789,
      "learning_rate": 7.3248e-06,
      "loss": 0.0543,
      "step": 18900
    },
    {
      "epoch": 1.516,
      "grad_norm": 0.08572965115308762,
      "learning_rate": 7.2647999999999995e-06,
      "loss": 0.0516,
      "step": 18950
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.10845118761062622,
      "learning_rate": 7.204800000000001e-06,
      "loss": 0.0554,
      "step": 19000
    },
    {
      "epoch": 1.52,
      "eval_loss": 0.08186879754066467,
      "eval_runtime": 88.0768,
      "eval_samples_per_second": 22.707,
      "eval_steps_per_second": 5.677,
      "step": 19000
    },
    {
      "epoch": 1.524,
      "grad_norm": 0.09976433962583542,
      "learning_rate": 7.144800000000001e-06,
      "loss": 0.0496,
      "step": 19050
    },
    {
      "epoch": 1.528,
      "grad_norm": 0.19336478412151337,
      "learning_rate": 7.0848e-06,
      "loss": 0.0528,
      "step": 19100
    },
    {
      "epoch": 1.532,
      "grad_norm": 0.13555702567100525,
      "learning_rate": 7.0248e-06,
      "loss": 0.0559,
      "step": 19150
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.2016674280166626,
      "learning_rate": 6.964800000000001e-06,
      "loss": 0.0568,
      "step": 19200
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.8713797330856323,
      "learning_rate": 6.9048e-06,
      "loss": 0.0522,
      "step": 19250
    },
    {
      "epoch": 1.544,
      "grad_norm": 0.13156233727931976,
      "learning_rate": 6.8448e-06,
      "loss": 0.0557,
      "step": 19300
    },
    {
      "epoch": 1.548,
      "grad_norm": 0.1713368147611618,
      "learning_rate": 6.7848e-06,
      "loss": 0.0574,
      "step": 19350
    },
    {
      "epoch": 1.552,
      "grad_norm": 0.13423492014408112,
      "learning_rate": 6.7248e-06,
      "loss": 0.0494,
      "step": 19400
    },
    {
      "epoch": 1.556,
      "grad_norm": 0.1513233631849289,
      "learning_rate": 6.6648e-06,
      "loss": 0.0528,
      "step": 19450
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.1609751433134079,
      "learning_rate": 6.606000000000001e-06,
      "loss": 0.0512,
      "step": 19500
    },
    {
      "epoch": 1.56,
      "eval_loss": 0.08218736946582794,
      "eval_runtime": 88.094,
      "eval_samples_per_second": 22.703,
      "eval_steps_per_second": 5.676,
      "step": 19500
    },
    {
      "epoch": 1.564,
      "grad_norm": 0.22682276368141174,
      "learning_rate": 6.5472e-06,
      "loss": 0.0603,
      "step": 19550
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.13181114196777344,
      "learning_rate": 6.4871999999999995e-06,
      "loss": 0.0497,
      "step": 19600
    },
    {
      "epoch": 1.572,
      "grad_norm": 0.1349440962076187,
      "learning_rate": 6.427200000000001e-06,
      "loss": 0.0507,
      "step": 19650
    },
    {
      "epoch": 1.576,
      "grad_norm": 0.1361471712589264,
      "learning_rate": 6.367200000000001e-06,
      "loss": 0.0501,
      "step": 19700
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.25193363428115845,
      "learning_rate": 6.3072e-06,
      "loss": 0.0565,
      "step": 19750
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.12969471514225006,
      "learning_rate": 6.2472e-06,
      "loss": 0.0616,
      "step": 19800
    },
    {
      "epoch": 1.588,
      "grad_norm": 0.18332916498184204,
      "learning_rate": 6.187200000000001e-06,
      "loss": 0.0513,
      "step": 19850
    },
    {
      "epoch": 1.592,
      "grad_norm": 0.19354714453220367,
      "learning_rate": 6.1272e-06,
      "loss": 0.0612,
      "step": 19900
    },
    {
      "epoch": 1.596,
      "grad_norm": 0.17420926690101624,
      "learning_rate": 6.0672e-06,
      "loss": 0.0604,
      "step": 19950
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.09988817572593689,
      "learning_rate": 6.0072e-06,
      "loss": 0.0523,
      "step": 20000
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.08189179003238678,
      "eval_runtime": 88.0418,
      "eval_samples_per_second": 22.716,
      "eval_steps_per_second": 5.679,
      "step": 20000
    },
    {
      "epoch": 1.604,
      "grad_norm": 0.18323849141597748,
      "learning_rate": 5.9472e-06,
      "loss": 0.0555,
      "step": 20050
    },
    {
      "epoch": 1.608,
      "grad_norm": 0.10411707311868668,
      "learning_rate": 5.8872e-06,
      "loss": 0.0506,
      "step": 20100
    },
    {
      "epoch": 1.612,
      "grad_norm": 0.18998867273330688,
      "learning_rate": 5.8272e-06,
      "loss": 0.0522,
      "step": 20150
    },
    {
      "epoch": 1.616,
      "grad_norm": 0.159571573138237,
      "learning_rate": 5.7672e-06,
      "loss": 0.0527,
      "step": 20200
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.10246822983026505,
      "learning_rate": 5.7072e-06,
      "loss": 0.0517,
      "step": 20250
    },
    {
      "epoch": 1.624,
      "grad_norm": 0.17232735455036163,
      "learning_rate": 5.6472e-06,
      "loss": 0.0559,
      "step": 20300
    },
    {
      "epoch": 1.6280000000000001,
      "grad_norm": 0.13078981637954712,
      "learning_rate": 5.5872e-06,
      "loss": 0.0497,
      "step": 20350
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.06083545461297035,
      "learning_rate": 5.527199999999999e-06,
      "loss": 0.0502,
      "step": 20400
    },
    {
      "epoch": 1.6360000000000001,
      "grad_norm": 0.1311408132314682,
      "learning_rate": 5.467200000000001e-06,
      "loss": 0.0572,
      "step": 20450
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.16154557466506958,
      "learning_rate": 5.4072000000000005e-06,
      "loss": 0.0545,
      "step": 20500
    },
    {
      "epoch": 1.6400000000000001,
      "eval_loss": 0.08202797174453735,
      "eval_runtime": 87.9943,
      "eval_samples_per_second": 22.729,
      "eval_steps_per_second": 5.682,
      "step": 20500
    },
    {
      "epoch": 1.6440000000000001,
      "grad_norm": 0.13027295470237732,
      "learning_rate": 5.3472e-06,
      "loss": 0.0551,
      "step": 20550
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 0.1394919753074646,
      "learning_rate": 5.2872e-06,
      "loss": 0.054,
      "step": 20600
    },
    {
      "epoch": 1.6520000000000001,
      "grad_norm": 0.16753709316253662,
      "learning_rate": 5.2272000000000005e-06,
      "loss": 0.0501,
      "step": 20650
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 0.1509876549243927,
      "learning_rate": 5.1672e-06,
      "loss": 0.0527,
      "step": 20700
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 0.13625292479991913,
      "learning_rate": 5.1072e-06,
      "loss": 0.0508,
      "step": 20750
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 0.1552583873271942,
      "learning_rate": 5.0472000000000006e-06,
      "loss": 0.0548,
      "step": 20800
    },
    {
      "epoch": 1.6680000000000001,
      "grad_norm": 0.1763962060213089,
      "learning_rate": 4.9872e-06,
      "loss": 0.0585,
      "step": 20850
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 0.11216771602630615,
      "learning_rate": 4.9272e-06,
      "loss": 0.0567,
      "step": 20900
    },
    {
      "epoch": 1.6760000000000002,
      "grad_norm": 0.08550629019737244,
      "learning_rate": 4.8672e-06,
      "loss": 0.0523,
      "step": 20950
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.11488083750009537,
      "learning_rate": 4.8072e-06,
      "loss": 0.0503,
      "step": 21000
    },
    {
      "epoch": 1.6800000000000002,
      "eval_loss": 0.0817028358578682,
      "eval_runtime": 88.097,
      "eval_samples_per_second": 22.702,
      "eval_steps_per_second": 5.676,
      "step": 21000
    },
    {
      "epoch": 1.6840000000000002,
      "grad_norm": 0.1464158296585083,
      "learning_rate": 4.7472e-06,
      "loss": 0.0557,
      "step": 21050
    },
    {
      "epoch": 1.688,
      "grad_norm": 0.1558839976787567,
      "learning_rate": 4.6872e-06,
      "loss": 0.0527,
      "step": 21100
    },
    {
      "epoch": 1.692,
      "grad_norm": 0.093449167907238,
      "learning_rate": 4.6271999999999995e-06,
      "loss": 0.0544,
      "step": 21150
    },
    {
      "epoch": 1.696,
      "grad_norm": 0.1413930058479309,
      "learning_rate": 4.5672e-06,
      "loss": 0.0571,
      "step": 21200
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.19774900376796722,
      "learning_rate": 4.507200000000001e-06,
      "loss": 0.0474,
      "step": 21250
    },
    {
      "epoch": 1.704,
      "grad_norm": 0.092808298766613,
      "learning_rate": 4.4472e-06,
      "loss": 0.05,
      "step": 21300
    },
    {
      "epoch": 1.708,
      "grad_norm": 0.09530337899923325,
      "learning_rate": 4.3872e-06,
      "loss": 0.0564,
      "step": 21350
    },
    {
      "epoch": 1.712,
      "grad_norm": 0.1850968301296234,
      "learning_rate": 4.327200000000001e-06,
      "loss": 0.0541,
      "step": 21400
    },
    {
      "epoch": 1.716,
      "grad_norm": 0.23416727781295776,
      "learning_rate": 4.2672e-06,
      "loss": 0.0576,
      "step": 21450
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.11343374848365784,
      "learning_rate": 4.2072e-06,
      "loss": 0.0529,
      "step": 21500
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.08182183653116226,
      "eval_runtime": 88.0562,
      "eval_samples_per_second": 22.713,
      "eval_steps_per_second": 5.678,
      "step": 21500
    },
    {
      "epoch": 1.724,
      "grad_norm": 0.14097870886325836,
      "learning_rate": 4.1472e-06,
      "loss": 0.0488,
      "step": 21550
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.1706978976726532,
      "learning_rate": 4.0872000000000004e-06,
      "loss": 0.0496,
      "step": 21600
    },
    {
      "epoch": 1.732,
      "grad_norm": 0.14371682703495026,
      "learning_rate": 4.0272e-06,
      "loss": 0.0557,
      "step": 21650
    },
    {
      "epoch": 1.736,
      "grad_norm": 0.1176629364490509,
      "learning_rate": 3.9672e-06,
      "loss": 0.0531,
      "step": 21700
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.17272049188613892,
      "learning_rate": 3.9072e-06,
      "loss": 0.0604,
      "step": 21750
    },
    {
      "epoch": 1.744,
      "grad_norm": 0.15085135400295258,
      "learning_rate": 3.8472e-06,
      "loss": 0.0537,
      "step": 21800
    },
    {
      "epoch": 1.748,
      "grad_norm": 0.11613863706588745,
      "learning_rate": 3.7884e-06,
      "loss": 0.0544,
      "step": 21850
    },
    {
      "epoch": 1.752,
      "grad_norm": 0.13247713446617126,
      "learning_rate": 3.7284e-06,
      "loss": 0.0556,
      "step": 21900
    },
    {
      "epoch": 1.756,
      "grad_norm": 0.1755180060863495,
      "learning_rate": 3.6684e-06,
      "loss": 0.0563,
      "step": 21950
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.07918363809585571,
      "learning_rate": 3.6084e-06,
      "loss": 0.0508,
      "step": 22000
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.08166228979825974,
      "eval_runtime": 88.0895,
      "eval_samples_per_second": 22.704,
      "eval_steps_per_second": 5.676,
      "step": 22000
    },
    {
      "epoch": 1.764,
      "grad_norm": 0.18784619867801666,
      "learning_rate": 3.5483999999999998e-06,
      "loss": 0.0559,
      "step": 22050
    },
    {
      "epoch": 1.768,
      "grad_norm": 0.1378210335969925,
      "learning_rate": 3.4884e-06,
      "loss": 0.0528,
      "step": 22100
    },
    {
      "epoch": 1.772,
      "grad_norm": 0.1482810080051422,
      "learning_rate": 3.4284e-06,
      "loss": 0.0535,
      "step": 22150
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.2096211314201355,
      "learning_rate": 3.3684000000000002e-06,
      "loss": 0.0455,
      "step": 22200
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.16895908117294312,
      "learning_rate": 3.3084000000000004e-06,
      "loss": 0.0527,
      "step": 22250
    },
    {
      "epoch": 1.784,
      "grad_norm": 0.11279798299074173,
      "learning_rate": 3.2484e-06,
      "loss": 0.0524,
      "step": 22300
    },
    {
      "epoch": 1.788,
      "grad_norm": 0.1444096565246582,
      "learning_rate": 3.1884000000000003e-06,
      "loss": 0.0539,
      "step": 22350
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.16793648898601532,
      "learning_rate": 3.1284e-06,
      "loss": 0.0578,
      "step": 22400
    },
    {
      "epoch": 1.796,
      "grad_norm": 0.1865730732679367,
      "learning_rate": 3.0684e-06,
      "loss": 0.0538,
      "step": 22450
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.11493191868066788,
      "learning_rate": 3.0084e-06,
      "loss": 0.0515,
      "step": 22500
    },
    {
      "epoch": 1.8,
      "eval_loss": 0.0815029963850975,
      "eval_runtime": 88.0709,
      "eval_samples_per_second": 22.709,
      "eval_steps_per_second": 5.677,
      "step": 22500
    },
    {
      "epoch": 1.804,
      "grad_norm": 0.10228476673364639,
      "learning_rate": 2.9496e-06,
      "loss": 0.0534,
      "step": 22550
    },
    {
      "epoch": 1.808,
      "grad_norm": 0.12501686811447144,
      "learning_rate": 2.8896000000000003e-06,
      "loss": 0.0529,
      "step": 22600
    },
    {
      "epoch": 1.812,
      "grad_norm": 0.10296665877103806,
      "learning_rate": 2.8296e-06,
      "loss": 0.0575,
      "step": 22650
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 0.15849712491035461,
      "learning_rate": 2.7696e-06,
      "loss": 0.054,
      "step": 22700
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 0.0910249873995781,
      "learning_rate": 2.7096e-06,
      "loss": 0.0533,
      "step": 22750
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.14466793835163116,
      "learning_rate": 2.6496e-06,
      "loss": 0.0512,
      "step": 22800
    },
    {
      "epoch": 1.8279999999999998,
      "grad_norm": 0.21741582453250885,
      "learning_rate": 2.5895999999999997e-06,
      "loss": 0.0574,
      "step": 22850
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 0.14835171401500702,
      "learning_rate": 2.5296000000000003e-06,
      "loss": 0.0569,
      "step": 22900
    },
    {
      "epoch": 1.8359999999999999,
      "grad_norm": 0.13074947893619537,
      "learning_rate": 2.4696e-06,
      "loss": 0.0555,
      "step": 22950
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.14285966753959656,
      "learning_rate": 2.4096e-06,
      "loss": 0.0531,
      "step": 23000
    },
    {
      "epoch": 1.8399999999999999,
      "eval_loss": 0.08154193311929703,
      "eval_runtime": 88.0742,
      "eval_samples_per_second": 22.708,
      "eval_steps_per_second": 5.677,
      "step": 23000
    },
    {
      "epoch": 1.8439999999999999,
      "grad_norm": 0.2026013433933258,
      "learning_rate": 2.3496e-06,
      "loss": 0.0526,
      "step": 23050
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 0.1372475028038025,
      "learning_rate": 2.2896e-06,
      "loss": 0.057,
      "step": 23100
    },
    {
      "epoch": 1.8519999999999999,
      "grad_norm": 0.13629719614982605,
      "learning_rate": 2.2296e-06,
      "loss": 0.0576,
      "step": 23150
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.2135656774044037,
      "learning_rate": 2.1696e-06,
      "loss": 0.0571,
      "step": 23200
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 0.1738504022359848,
      "learning_rate": 2.1095999999999997e-06,
      "loss": 0.054,
      "step": 23250
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 0.17974117398262024,
      "learning_rate": 2.0496000000000002e-06,
      "loss": 0.0551,
      "step": 23300
    },
    {
      "epoch": 1.8679999999999999,
      "grad_norm": 0.16323037445545197,
      "learning_rate": 1.9896e-06,
      "loss": 0.0515,
      "step": 23350
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.11057537794113159,
      "learning_rate": 1.9296e-06,
      "loss": 0.0564,
      "step": 23400
    },
    {
      "epoch": 1.876,
      "grad_norm": 0.1545117348432541,
      "learning_rate": 1.8696e-06,
      "loss": 0.0508,
      "step": 23450
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.25704920291900635,
      "learning_rate": 1.8096e-06,
      "loss": 0.0521,
      "step": 23500
    },
    {
      "epoch": 1.88,
      "eval_loss": 0.08151672035455704,
      "eval_runtime": 88.0934,
      "eval_samples_per_second": 22.703,
      "eval_steps_per_second": 5.676,
      "step": 23500
    },
    {
      "epoch": 1.884,
      "grad_norm": 0.10899360477924347,
      "learning_rate": 1.7496e-06,
      "loss": 0.0558,
      "step": 23550
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.09025511890649796,
      "learning_rate": 1.6896e-06,
      "loss": 0.0582,
      "step": 23600
    },
    {
      "epoch": 1.892,
      "grad_norm": 0.11647246032953262,
      "learning_rate": 1.6296e-06,
      "loss": 0.0518,
      "step": 23650
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.15982107818126678,
      "learning_rate": 1.5696e-06,
      "loss": 0.0535,
      "step": 23700
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.15808264911174774,
      "learning_rate": 1.5096e-06,
      "loss": 0.0557,
      "step": 23750
    },
    {
      "epoch": 1.904,
      "grad_norm": 0.12610608339309692,
      "learning_rate": 1.4496e-06,
      "loss": 0.0547,
      "step": 23800
    },
    {
      "epoch": 1.908,
      "grad_norm": 0.10674013942480087,
      "learning_rate": 1.3896e-06,
      "loss": 0.0506,
      "step": 23850
    },
    {
      "epoch": 1.912,
      "grad_norm": 0.15550707280635834,
      "learning_rate": 1.3296e-06,
      "loss": 0.0576,
      "step": 23900
    },
    {
      "epoch": 1.916,
      "grad_norm": 0.12588605284690857,
      "learning_rate": 1.2696e-06,
      "loss": 0.0566,
      "step": 23950
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.1209346354007721,
      "learning_rate": 1.2096e-06,
      "loss": 0.0564,
      "step": 24000
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.08148019760847092,
      "eval_runtime": 88.0412,
      "eval_samples_per_second": 22.717,
      "eval_steps_per_second": 5.679,
      "step": 24000
    },
    {
      "epoch": 1.924,
      "grad_norm": 0.18958355486392975,
      "learning_rate": 1.1496e-06,
      "loss": 0.054,
      "step": 24050
    },
    {
      "epoch": 1.928,
      "grad_norm": 0.12351219356060028,
      "learning_rate": 1.0896e-06,
      "loss": 0.0614,
      "step": 24100
    },
    {
      "epoch": 1.932,
      "grad_norm": 0.175857812166214,
      "learning_rate": 1.0296e-06,
      "loss": 0.0541,
      "step": 24150
    },
    {
      "epoch": 1.936,
      "grad_norm": 0.15968987345695496,
      "learning_rate": 9.696e-07,
      "loss": 0.0568,
      "step": 24200
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.15781116485595703,
      "learning_rate": 9.096e-07,
      "loss": 0.055,
      "step": 24250
    },
    {
      "epoch": 1.944,
      "grad_norm": 0.17276370525360107,
      "learning_rate": 8.496000000000001e-07,
      "loss": 0.0475,
      "step": 24300
    },
    {
      "epoch": 1.948,
      "grad_norm": 0.09511862695217133,
      "learning_rate": 7.896e-07,
      "loss": 0.0509,
      "step": 24350
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.1636885106563568,
      "learning_rate": 7.296000000000001e-07,
      "loss": 0.0508,
      "step": 24400
    },
    {
      "epoch": 1.956,
      "grad_norm": 0.12255977094173431,
      "learning_rate": 6.696e-07,
      "loss": 0.0467,
      "step": 24450
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.18492284417152405,
      "learning_rate": 6.096000000000001e-07,
      "loss": 0.0559,
      "step": 24500
    },
    {
      "epoch": 1.96,
      "eval_loss": 0.08147666603326797,
      "eval_runtime": 88.0584,
      "eval_samples_per_second": 22.712,
      "eval_steps_per_second": 5.678,
      "step": 24500
    },
    {
      "epoch": 1.964,
      "grad_norm": 0.1472010761499405,
      "learning_rate": 5.496e-07,
      "loss": 0.0485,
      "step": 24550
    },
    {
      "epoch": 1.968,
      "grad_norm": 0.14501795172691345,
      "learning_rate": 4.896e-07,
      "loss": 0.0554,
      "step": 24600
    },
    {
      "epoch": 1.972,
      "grad_norm": 0.09672766178846359,
      "learning_rate": 4.296e-07,
      "loss": 0.0534,
      "step": 24650
    },
    {
      "epoch": 1.976,
      "grad_norm": 0.1332007199525833,
      "learning_rate": 3.696e-07,
      "loss": 0.0513,
      "step": 24700
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.0856068879365921,
      "learning_rate": 3.0959999999999997e-07,
      "loss": 0.0531,
      "step": 24750
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.07875853031873703,
      "learning_rate": 2.4959999999999996e-07,
      "loss": 0.0511,
      "step": 24800
    },
    {
      "epoch": 1.988,
      "grad_norm": 0.1319677084684372,
      "learning_rate": 1.896e-07,
      "loss": 0.0493,
      "step": 24850
    },
    {
      "epoch": 1.992,
      "grad_norm": 0.20934289693832397,
      "learning_rate": 1.296e-07,
      "loss": 0.0559,
      "step": 24900
    },
    {
      "epoch": 1.996,
      "grad_norm": 0.12190008908510208,
      "learning_rate": 6.96e-08,
      "loss": 0.0589,
      "step": 24950
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2332906574010849,
      "learning_rate": 9.600000000000002e-09,
      "loss": 0.0557,
      "step": 25000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.08146847784519196,
      "eval_runtime": 88.0931,
      "eval_samples_per_second": 22.703,
      "eval_steps_per_second": 5.676,
      "step": 25000
    }
  ],
  "logging_steps": 50,
  "max_steps": 25000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.0895789056e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}