{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2456094134176326,
  "eval_steps": 500,
  "global_step": 2797,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008781173164734809,
      "grad_norm": 0.7306153178215027,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 0.6122,
      "step": 10
    },
    {
      "epoch": 0.0017562346329469617,
      "grad_norm": 0.9519491195678711,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 0.603,
      "step": 20
    },
    {
      "epoch": 0.0026343519494204425,
      "grad_norm": 0.6061155200004578,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.6155,
      "step": 30
    },
    {
      "epoch": 0.0035124692658939235,
      "grad_norm": 0.7508756518363953,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.5866,
      "step": 40
    },
    {
      "epoch": 0.0043905865823674044,
      "grad_norm": 0.719323456287384,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.6008,
      "step": 50
    },
    {
      "epoch": 0.005268703898840885,
      "grad_norm": 0.6478108763694763,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5735,
      "step": 60
    },
    {
      "epoch": 0.006146821215314366,
      "grad_norm": 0.5820965766906738,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.5789,
      "step": 70
    },
    {
      "epoch": 0.007024938531787847,
      "grad_norm": 0.6183585524559021,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.5523,
      "step": 80
    },
    {
      "epoch": 0.007903055848261328,
      "grad_norm": 0.5942079424858093,
      "learning_rate": 1.005e-05,
      "loss": 0.5678,
      "step": 90
    },
    {
      "epoch": 0.008781173164734809,
      "grad_norm": 0.5680508613586426,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.5876,
      "step": 100
    },
    {
      "epoch": 0.00965929048120829,
      "grad_norm": 0.7514785528182983,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5735,
      "step": 110
    },
    {
      "epoch": 0.01053740779768177,
      "grad_norm": 0.6386587023735046,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.5764,
      "step": 120
    },
    {
      "epoch": 0.01141552511415525,
      "grad_norm": 0.6114295125007629,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.5751,
      "step": 130
    },
    {
      "epoch": 0.012293642430628733,
      "grad_norm": 0.706280529499054,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.5765,
      "step": 140
    },
    {
      "epoch": 0.013171759747102213,
      "grad_norm": 0.6190201044082642,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.5834,
      "step": 150
    },
    {
      "epoch": 0.014049877063575694,
      "grad_norm": 0.7192485928535461,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.5922,
      "step": 160
    },
    {
      "epoch": 0.014927994380049174,
      "grad_norm": 0.597809374332428,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.5687,
      "step": 170
    },
    {
      "epoch": 0.015806111696522657,
      "grad_norm": 0.5657880902290344,
      "learning_rate": 2e-05,
      "loss": 0.5759,
      "step": 180
    },
    {
      "epoch": 0.016684229012996137,
      "grad_norm": 0.4955204427242279,
      "learning_rate": 2e-05,
      "loss": 0.5932,
      "step": 190
    },
    {
      "epoch": 0.017562346329469618,
      "grad_norm": 0.476298451423645,
      "learning_rate": 2e-05,
      "loss": 0.5747,
      "step": 200
    },
    {
      "epoch": 0.0184404636459431,
      "grad_norm": 0.4774770438671112,
      "learning_rate": 2e-05,
      "loss": 0.5909,
      "step": 210
    },
    {
      "epoch": 0.01931858096241658,
      "grad_norm": 0.4712103605270386,
      "learning_rate": 2e-05,
      "loss": 0.5686,
      "step": 220
    },
    {
      "epoch": 0.02019669827889006,
      "grad_norm": 0.5302442908287048,
      "learning_rate": 2e-05,
      "loss": 0.5817,
      "step": 230
    },
    {
      "epoch": 0.02107481559536354,
      "grad_norm": 0.45405933260917664,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 240
    },
    {
      "epoch": 0.02195293291183702,
      "grad_norm": 0.4746021032333374,
      "learning_rate": 2e-05,
      "loss": 0.6033,
      "step": 250
    },
    {
      "epoch": 0.0228310502283105,
      "grad_norm": 0.4489727318286896,
      "learning_rate": 2e-05,
      "loss": 0.6024,
      "step": 260
    },
    {
      "epoch": 0.02370916754478398,
      "grad_norm": 0.4666856527328491,
      "learning_rate": 2e-05,
      "loss": 0.5715,
      "step": 270
    },
    {
      "epoch": 0.024587284861257466,
      "grad_norm": 0.5077083110809326,
      "learning_rate": 2e-05,
      "loss": 0.5815,
      "step": 280
    },
    {
      "epoch": 0.025465402177730946,
      "grad_norm": 0.43067094683647156,
      "learning_rate": 2e-05,
      "loss": 0.5723,
      "step": 290
    },
    {
      "epoch": 0.026343519494204427,
      "grad_norm": 0.4907589256763458,
      "learning_rate": 2e-05,
      "loss": 0.5818,
      "step": 300
    },
    {
      "epoch": 0.027221636810677907,
      "grad_norm": 0.3945549428462982,
      "learning_rate": 2e-05,
      "loss": 0.5826,
      "step": 310
    },
    {
      "epoch": 0.028099754127151388,
      "grad_norm": 0.39142847061157227,
      "learning_rate": 2e-05,
      "loss": 0.5826,
      "step": 320
    },
    {
      "epoch": 0.02897787144362487,
      "grad_norm": 0.43336743116378784,
      "learning_rate": 2e-05,
      "loss": 0.5859,
      "step": 330
    },
    {
      "epoch": 0.02985598876009835,
      "grad_norm": 0.3956754803657532,
      "learning_rate": 2e-05,
      "loss": 0.5862,
      "step": 340
    },
    {
      "epoch": 0.03073410607657183,
      "grad_norm": 0.37747862935066223,
      "learning_rate": 2e-05,
      "loss": 0.5866,
      "step": 350
    },
    {
      "epoch": 0.03161222339304531,
      "grad_norm": 0.4057143032550812,
      "learning_rate": 2e-05,
      "loss": 0.5888,
      "step": 360
    },
    {
      "epoch": 0.03249034070951879,
      "grad_norm": 0.3896559774875641,
      "learning_rate": 2e-05,
      "loss": 0.5819,
      "step": 370
    },
    {
      "epoch": 0.033368458025992274,
      "grad_norm": 0.333257257938385,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 380
    },
    {
      "epoch": 0.03424657534246575,
      "grad_norm": 0.41028985381126404,
      "learning_rate": 2e-05,
      "loss": 0.5906,
      "step": 390
    },
    {
      "epoch": 0.035124692658939236,
      "grad_norm": 0.4036722481250763,
      "learning_rate": 2e-05,
      "loss": 0.5687,
      "step": 400
    },
    {
      "epoch": 0.03600280997541271,
      "grad_norm": 0.40243393182754517,
      "learning_rate": 2e-05,
      "loss": 0.578,
      "step": 410
    },
    {
      "epoch": 0.0368809272918862,
      "grad_norm": 0.4444314241409302,
      "learning_rate": 2e-05,
      "loss": 0.5886,
      "step": 420
    },
    {
      "epoch": 0.037759044608359674,
      "grad_norm": 0.3521971106529236,
      "learning_rate": 2e-05,
      "loss": 0.5807,
      "step": 430
    },
    {
      "epoch": 0.03863716192483316,
      "grad_norm": 0.3873286545276642,
      "learning_rate": 2e-05,
      "loss": 0.5656,
      "step": 440
    },
    {
      "epoch": 0.03951527924130664,
      "grad_norm": 0.3987811207771301,
      "learning_rate": 2e-05,
      "loss": 0.5804,
      "step": 450
    },
    {
      "epoch": 0.04039339655778012,
      "grad_norm": 0.4175412356853485,
      "learning_rate": 2e-05,
      "loss": 0.5695,
      "step": 460
    },
    {
      "epoch": 0.0412715138742536,
      "grad_norm": 0.37522923946380615,
      "learning_rate": 2e-05,
      "loss": 0.5957,
      "step": 470
    },
    {
      "epoch": 0.04214963119072708,
      "grad_norm": 0.44698572158813477,
      "learning_rate": 2e-05,
      "loss": 0.5958,
      "step": 480
    },
    {
      "epoch": 0.043027748507200564,
      "grad_norm": 0.3578386604785919,
      "learning_rate": 2e-05,
      "loss": 0.5666,
      "step": 490
    },
    {
      "epoch": 0.04390586582367404,
      "grad_norm": 0.40327927470207214,
      "learning_rate": 2e-05,
      "loss": 0.5962,
      "step": 500
    },
    {
      "epoch": 0.044783983140147525,
      "grad_norm": 0.4901353716850281,
      "learning_rate": 2e-05,
      "loss": 0.6052,
      "step": 510
    },
    {
      "epoch": 0.045662100456621,
      "grad_norm": 0.3535653054714203,
      "learning_rate": 2e-05,
      "loss": 0.5766,
      "step": 520
    },
    {
      "epoch": 0.046540217773094486,
      "grad_norm": 0.5284445285797119,
      "learning_rate": 2e-05,
      "loss": 0.5685,
      "step": 530
    },
    {
      "epoch": 0.04741833508956796,
      "grad_norm": 0.38909244537353516,
      "learning_rate": 2e-05,
      "loss": 0.6218,
      "step": 540
    },
    {
      "epoch": 0.04829645240604145,
      "grad_norm": 0.4495580494403839,
      "learning_rate": 2e-05,
      "loss": 0.6113,
      "step": 550
    },
    {
      "epoch": 0.04917456972251493,
      "grad_norm": 0.3668941855430603,
      "learning_rate": 2e-05,
      "loss": 0.5851,
      "step": 560
    },
    {
      "epoch": 0.05005268703898841,
      "grad_norm": 0.3919713497161865,
      "learning_rate": 2e-05,
      "loss": 0.5783,
      "step": 570
    },
    {
      "epoch": 0.05093080435546189,
      "grad_norm": 0.36708441376686096,
      "learning_rate": 2e-05,
      "loss": 0.5812,
      "step": 580
    },
    {
      "epoch": 0.05180892167193537,
      "grad_norm": 0.4177910387516022,
      "learning_rate": 2e-05,
      "loss": 0.5918,
      "step": 590
    },
    {
      "epoch": 0.05268703898840885,
      "grad_norm": 0.39156895875930786,
      "learning_rate": 2e-05,
      "loss": 0.5892,
      "step": 600
    },
    {
      "epoch": 0.05356515630488233,
      "grad_norm": 0.4563087224960327,
      "learning_rate": 2e-05,
      "loss": 0.5873,
      "step": 610
    },
    {
      "epoch": 0.054443273621355814,
      "grad_norm": 0.35439926385879517,
      "learning_rate": 2e-05,
      "loss": 0.5862,
      "step": 620
    },
    {
      "epoch": 0.05532139093782929,
      "grad_norm": 0.3745802342891693,
      "learning_rate": 2e-05,
      "loss": 0.5776,
      "step": 630
    },
    {
      "epoch": 0.056199508254302775,
      "grad_norm": 0.3608105480670929,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 640
    },
    {
      "epoch": 0.05707762557077625,
      "grad_norm": 0.43292662501335144,
      "learning_rate": 2e-05,
      "loss": 0.5719,
      "step": 650
    },
    {
      "epoch": 0.05795574288724974,
      "grad_norm": 0.34552648663520813,
      "learning_rate": 2e-05,
      "loss": 0.5921,
      "step": 660
    },
    {
      "epoch": 0.05883386020372322,
      "grad_norm": 0.41297587752342224,
      "learning_rate": 2e-05,
      "loss": 0.6026,
      "step": 670
    },
    {
      "epoch": 0.0597119775201967,
      "grad_norm": 0.38442927598953247,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 680
    },
    {
      "epoch": 0.06059009483667018,
      "grad_norm": 0.36447006464004517,
      "learning_rate": 2e-05,
      "loss": 0.5888,
      "step": 690
    },
    {
      "epoch": 0.06146821215314366,
      "grad_norm": 0.36359819769859314,
      "learning_rate": 2e-05,
      "loss": 0.5778,
      "step": 700
    },
    {
      "epoch": 0.06234632946961714,
      "grad_norm": 0.3486689627170563,
      "learning_rate": 2e-05,
      "loss": 0.5999,
      "step": 710
    },
    {
      "epoch": 0.06322444678609063,
      "grad_norm": 0.3839160203933716,
      "learning_rate": 2e-05,
      "loss": 0.5783,
      "step": 720
    },
    {
      "epoch": 0.0641025641025641,
      "grad_norm": 0.42905154824256897,
      "learning_rate": 2e-05,
      "loss": 0.5919,
      "step": 730
    },
    {
      "epoch": 0.06498068141903758,
      "grad_norm": 0.3768789768218994,
      "learning_rate": 2e-05,
      "loss": 0.5689,
      "step": 740
    },
    {
      "epoch": 0.06585879873551106,
      "grad_norm": 0.33872702717781067,
      "learning_rate": 2e-05,
      "loss": 0.569,
      "step": 750
    },
    {
      "epoch": 0.06673691605198455,
      "grad_norm": 0.3960425555706024,
      "learning_rate": 2e-05,
      "loss": 0.5803,
      "step": 760
    },
    {
      "epoch": 0.06761503336845803,
      "grad_norm": 0.8060588836669922,
      "learning_rate": 2e-05,
      "loss": 0.5979,
      "step": 770
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 0.42517799139022827,
      "learning_rate": 2e-05,
      "loss": 0.6079,
      "step": 780
    },
    {
      "epoch": 0.06937126800140499,
      "grad_norm": 0.40368008613586426,
      "learning_rate": 2e-05,
      "loss": 0.5885,
      "step": 790
    },
    {
      "epoch": 0.07024938531787847,
      "grad_norm": 0.36727896332740784,
      "learning_rate": 2e-05,
      "loss": 0.5804,
      "step": 800
    },
    {
      "epoch": 0.07112750263435196,
      "grad_norm": 0.357287734746933,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 810
    },
    {
      "epoch": 0.07200561995082543,
      "grad_norm": 0.39184826612472534,
      "learning_rate": 2e-05,
      "loss": 0.5832,
      "step": 820
    },
    {
      "epoch": 0.07288373726729891,
      "grad_norm": 0.3752371370792389,
      "learning_rate": 2e-05,
      "loss": 0.5528,
      "step": 830
    },
    {
      "epoch": 0.0737618545837724,
      "grad_norm": 0.3477722704410553,
      "learning_rate": 2e-05,
      "loss": 0.5861,
      "step": 840
    },
    {
      "epoch": 0.07463997190024588,
      "grad_norm": 0.3644666075706482,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 850
    },
    {
      "epoch": 0.07551808921671935,
      "grad_norm": 0.3730826675891876,
      "learning_rate": 2e-05,
      "loss": 0.6121,
      "step": 860
    },
    {
      "epoch": 0.07639620653319283,
      "grad_norm": 0.369928777217865,
      "learning_rate": 2e-05,
      "loss": 0.5858,
      "step": 870
    },
    {
      "epoch": 0.07727432384966632,
      "grad_norm": 0.40936708450317383,
      "learning_rate": 2e-05,
      "loss": 0.6146,
      "step": 880
    },
    {
      "epoch": 0.0781524411661398,
      "grad_norm": 0.3968866169452667,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 890
    },
    {
      "epoch": 0.07903055848261328,
      "grad_norm": 0.34718722105026245,
      "learning_rate": 2e-05,
      "loss": 0.5706,
      "step": 900
    },
    {
      "epoch": 0.07990867579908675,
      "grad_norm": 0.3978041410446167,
      "learning_rate": 2e-05,
      "loss": 0.5865,
      "step": 910
    },
    {
      "epoch": 0.08078679311556024,
      "grad_norm": 0.3254806399345398,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 920
    },
    {
      "epoch": 0.08166491043203372,
      "grad_norm": 0.35298028588294983,
      "learning_rate": 2e-05,
      "loss": 0.6019,
      "step": 930
    },
    {
      "epoch": 0.0825430277485072,
      "grad_norm": 0.42407262325286865,
      "learning_rate": 2e-05,
      "loss": 0.6136,
      "step": 940
    },
    {
      "epoch": 0.08342114506498068,
      "grad_norm": 0.3866136074066162,
      "learning_rate": 2e-05,
      "loss": 0.5884,
      "step": 950
    },
    {
      "epoch": 0.08429926238145416,
      "grad_norm": 0.3518330454826355,
      "learning_rate": 2e-05,
      "loss": 0.5958,
      "step": 960
    },
    {
      "epoch": 0.08517737969792764,
      "grad_norm": 0.34305262565612793,
      "learning_rate": 2e-05,
      "loss": 0.5255,
      "step": 970
    },
    {
      "epoch": 0.08605549701440113,
      "grad_norm": 0.3526217043399811,
      "learning_rate": 2e-05,
      "loss": 0.5693,
      "step": 980
    },
    {
      "epoch": 0.08693361433087461,
      "grad_norm": 0.40801969170570374,
      "learning_rate": 2e-05,
      "loss": 0.5732,
      "step": 990
    },
    {
      "epoch": 0.08781173164734808,
      "grad_norm": 0.39768823981285095,
      "learning_rate": 2e-05,
      "loss": 0.5989,
      "step": 1000
    },
    {
      "epoch": 0.08868984896382157,
      "grad_norm": 0.4879983067512512,
      "learning_rate": 2e-05,
      "loss": 0.5825,
      "step": 1010
    },
    {
      "epoch": 0.08956796628029505,
      "grad_norm": 0.34764784574508667,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 1020
    },
    {
      "epoch": 0.09044608359676853,
      "grad_norm": 0.33165058493614197,
      "learning_rate": 2e-05,
      "loss": 0.6006,
      "step": 1030
    },
    {
      "epoch": 0.091324200913242,
      "grad_norm": 0.3657977283000946,
      "learning_rate": 2e-05,
      "loss": 0.5654,
      "step": 1040
    },
    {
      "epoch": 0.09220231822971549,
      "grad_norm": 0.3795984387397766,
      "learning_rate": 2e-05,
      "loss": 0.5812,
      "step": 1050
    },
    {
      "epoch": 0.09308043554618897,
      "grad_norm": 0.3213583827018738,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 1060
    },
    {
      "epoch": 0.09395855286266246,
      "grad_norm": 0.36173421144485474,
      "learning_rate": 2e-05,
      "loss": 0.5715,
      "step": 1070
    },
    {
      "epoch": 0.09483667017913593,
      "grad_norm": 0.42429402470588684,
      "learning_rate": 2e-05,
      "loss": 0.6046,
      "step": 1080
    },
    {
      "epoch": 0.09571478749560941,
      "grad_norm": 0.3333086669445038,
      "learning_rate": 2e-05,
      "loss": 0.5758,
      "step": 1090
    },
    {
      "epoch": 0.0965929048120829,
      "grad_norm": 0.3732471764087677,
      "learning_rate": 2e-05,
      "loss": 0.574,
      "step": 1100
    },
    {
      "epoch": 0.09747102212855638,
      "grad_norm": 0.3796260952949524,
      "learning_rate": 2e-05,
      "loss": 0.5723,
      "step": 1110
    },
    {
      "epoch": 0.09834913944502986,
      "grad_norm": 0.33387455344200134,
      "learning_rate": 2e-05,
      "loss": 0.581,
      "step": 1120
    },
    {
      "epoch": 0.09922725676150333,
      "grad_norm": 0.35619693994522095,
      "learning_rate": 2e-05,
      "loss": 0.5716,
      "step": 1130
    },
    {
      "epoch": 0.10010537407797682,
      "grad_norm": 0.37752023339271545,
      "learning_rate": 2e-05,
      "loss": 0.5797,
      "step": 1140
    },
    {
      "epoch": 0.1009834913944503,
      "grad_norm": 0.3690735399723053,
      "learning_rate": 2e-05,
      "loss": 0.5857,
      "step": 1150
    },
    {
      "epoch": 0.10186160871092378,
      "grad_norm": 0.34846580028533936,
      "learning_rate": 2e-05,
      "loss": 0.568,
      "step": 1160
    },
    {
      "epoch": 0.10273972602739725,
      "grad_norm": 0.3458128571510315,
      "learning_rate": 2e-05,
      "loss": 0.5988,
      "step": 1170
    },
    {
      "epoch": 0.10361784334387074,
      "grad_norm": 0.3719422221183777,
      "learning_rate": 2e-05,
      "loss": 0.582,
      "step": 1180
    },
    {
      "epoch": 0.10449596066034422,
      "grad_norm": 0.4103449881076813,
      "learning_rate": 2e-05,
      "loss": 0.5956,
      "step": 1190
    },
    {
      "epoch": 0.1053740779768177,
      "grad_norm": 0.4256537854671478,
      "learning_rate": 2e-05,
      "loss": 0.6203,
      "step": 1200
    },
    {
      "epoch": 0.10625219529329119,
      "grad_norm": 0.41839829087257385,
      "learning_rate": 2e-05,
      "loss": 0.5895,
      "step": 1210
    },
    {
      "epoch": 0.10713031260976466,
      "grad_norm": 0.3410065770149231,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 1220
    },
    {
      "epoch": 0.10800842992623814,
      "grad_norm": 0.4618186056613922,
      "learning_rate": 2e-05,
      "loss": 0.5837,
      "step": 1230
    },
    {
      "epoch": 0.10888654724271163,
      "grad_norm": 0.3834467828273773,
      "learning_rate": 2e-05,
      "loss": 0.587,
      "step": 1240
    },
    {
      "epoch": 0.10976466455918511,
      "grad_norm": 0.3355961740016937,
      "learning_rate": 2e-05,
      "loss": 0.5671,
      "step": 1250
    },
    {
      "epoch": 0.11064278187565858,
      "grad_norm": 0.35168391466140747,
      "learning_rate": 2e-05,
      "loss": 0.586,
      "step": 1260
    },
    {
      "epoch": 0.11152089919213207,
      "grad_norm": 0.3857669532299042,
      "learning_rate": 2e-05,
      "loss": 0.5881,
      "step": 1270
    },
    {
      "epoch": 0.11239901650860555,
      "grad_norm": 0.3423902690410614,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 1280
    },
    {
      "epoch": 0.11327713382507903,
      "grad_norm": 0.34977564215660095,
      "learning_rate": 2e-05,
      "loss": 0.5787,
      "step": 1290
    },
    {
      "epoch": 0.1141552511415525,
      "grad_norm": 0.3350295126438141,
      "learning_rate": 2e-05,
      "loss": 0.5937,
      "step": 1300
    },
    {
      "epoch": 0.11503336845802599,
      "grad_norm": 0.35885995626449585,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 1310
    },
    {
      "epoch": 0.11591148577449947,
      "grad_norm": 0.3675665855407715,
      "learning_rate": 2e-05,
      "loss": 0.5828,
      "step": 1320
    },
    {
      "epoch": 0.11678960309097296,
      "grad_norm": 0.36600708961486816,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 1330
    },
    {
      "epoch": 0.11766772040744644,
      "grad_norm": 0.361731618642807,
      "learning_rate": 2e-05,
      "loss": 0.6,
      "step": 1340
    },
    {
      "epoch": 0.11854583772391991,
      "grad_norm": 0.33314457535743713,
      "learning_rate": 2e-05,
      "loss": 0.5847,
      "step": 1350
    },
    {
      "epoch": 0.1194239550403934,
      "grad_norm": 0.44293349981307983,
      "learning_rate": 2e-05,
      "loss": 0.5818,
      "step": 1360
    },
    {
      "epoch": 0.12030207235686688,
      "grad_norm": 0.36806032061576843,
      "learning_rate": 2e-05,
      "loss": 0.5899,
      "step": 1370
    },
    {
      "epoch": 0.12118018967334036,
      "grad_norm": 0.3873898386955261,
      "learning_rate": 2e-05,
      "loss": 0.5722,
      "step": 1380
    },
    {
      "epoch": 0.12205830698981383,
      "grad_norm": 0.32653969526290894,
      "learning_rate": 2e-05,
      "loss": 0.5735,
      "step": 1390
    },
    {
      "epoch": 0.12293642430628732,
      "grad_norm": 0.37764307856559753,
      "learning_rate": 2e-05,
      "loss": 0.5844,
      "step": 1400
    },
    {
      "epoch": 0.1238145416227608,
      "grad_norm": 0.39646756649017334,
      "learning_rate": 2e-05,
      "loss": 0.5779,
      "step": 1410
    },
    {
      "epoch": 0.12469265893923429,
      "grad_norm": 0.35311469435691833,
      "learning_rate": 2e-05,
      "loss": 0.562,
      "step": 1420
    },
    {
      "epoch": 0.12557077625570776,
      "grad_norm": 0.39760711789131165,
      "learning_rate": 2e-05,
      "loss": 0.5804,
      "step": 1430
    },
    {
      "epoch": 0.12644889357218125,
      "grad_norm": 0.34020617604255676,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 1440
    },
    {
      "epoch": 0.12732701088865472,
      "grad_norm": 0.37063539028167725,
      "learning_rate": 2e-05,
      "loss": 0.5901,
      "step": 1450
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 0.3460864722728729,
      "learning_rate": 2e-05,
      "loss": 0.5765,
      "step": 1460
    },
    {
      "epoch": 0.1290832455216017,
      "grad_norm": 0.35517415404319763,
      "learning_rate": 2e-05,
      "loss": 0.5759,
      "step": 1470
    },
    {
      "epoch": 0.12996136283807516,
      "grad_norm": 0.400665283203125,
      "learning_rate": 2e-05,
      "loss": 0.5975,
      "step": 1480
    },
    {
      "epoch": 0.13083948015454866,
      "grad_norm": 0.37015774846076965,
      "learning_rate": 2e-05,
      "loss": 0.5772,
      "step": 1490
    },
    {
      "epoch": 0.13171759747102213,
      "grad_norm": 0.6685165762901306,
      "learning_rate": 2e-05,
      "loss": 0.5989,
      "step": 1500
    },
    {
      "epoch": 0.1325957147874956,
      "grad_norm": 0.449286550283432,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1510
    },
    {
      "epoch": 0.1334738321039691,
      "grad_norm": 0.3733328580856323,
      "learning_rate": 2e-05,
      "loss": 0.5837,
      "step": 1520
    },
    {
      "epoch": 0.13435194942044257,
      "grad_norm": 0.40340012311935425,
      "learning_rate": 2e-05,
      "loss": 0.5848,
      "step": 1530
    },
    {
      "epoch": 0.13523006673691607,
      "grad_norm": 0.392964631319046,
      "learning_rate": 2e-05,
      "loss": 0.5842,
      "step": 1540
    },
    {
      "epoch": 0.13610818405338954,
      "grad_norm": 0.33298441767692566,
      "learning_rate": 2e-05,
      "loss": 0.5757,
      "step": 1550
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 0.36373716592788696,
      "learning_rate": 2e-05,
      "loss": 0.5782,
      "step": 1560
    },
    {
      "epoch": 0.1378644186863365,
      "grad_norm": 0.3489643335342407,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1570
    },
    {
      "epoch": 0.13874253600280997,
      "grad_norm": 0.3226819932460785,
      "learning_rate": 2e-05,
      "loss": 0.5985,
      "step": 1580
    },
    {
      "epoch": 0.13962065331928344,
      "grad_norm": 0.40060415863990784,
      "learning_rate": 2e-05,
      "loss": 0.5766,
      "step": 1590
    },
    {
      "epoch": 0.14049877063575694,
      "grad_norm": 0.3596484661102295,
      "learning_rate": 2e-05,
      "loss": 0.5792,
      "step": 1600
    },
    {
      "epoch": 0.1413768879522304,
      "grad_norm": 0.3642957806587219,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 1610
    },
    {
      "epoch": 0.1422550052687039,
      "grad_norm": 0.9815397262573242,
      "learning_rate": 2e-05,
      "loss": 0.5826,
      "step": 1620
    },
    {
      "epoch": 0.14313312258517738,
      "grad_norm": 0.3498055934906006,
      "learning_rate": 2e-05,
      "loss": 0.6066,
      "step": 1630
    },
    {
      "epoch": 0.14401123990165085,
      "grad_norm": 0.31737828254699707,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 1640
    },
    {
      "epoch": 0.14488935721812435,
      "grad_norm": 0.3356111943721771,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 1650
    },
    {
      "epoch": 0.14576747453459782,
      "grad_norm": 0.3234397768974304,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 1660
    },
    {
      "epoch": 0.14664559185107132,
      "grad_norm": 0.3743995726108551,
      "learning_rate": 2e-05,
      "loss": 0.5963,
      "step": 1670
    },
    {
      "epoch": 0.1475237091675448,
      "grad_norm": 0.31582197546958923,
      "learning_rate": 2e-05,
      "loss": 0.5809,
      "step": 1680
    },
    {
      "epoch": 0.14840182648401826,
      "grad_norm": 0.37730827927589417,
      "learning_rate": 2e-05,
      "loss": 0.5648,
      "step": 1690
    },
    {
      "epoch": 0.14927994380049175,
      "grad_norm": 0.5714032649993896,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 1700
    },
    {
      "epoch": 0.15015806111696522,
      "grad_norm": 0.3735637366771698,
      "learning_rate": 2e-05,
      "loss": 0.5839,
      "step": 1710
    },
    {
      "epoch": 0.1510361784334387,
      "grad_norm": 0.34317269921302795,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 1720
    },
    {
      "epoch": 0.1519142957499122,
      "grad_norm": 0.37194734811782837,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 1730
    },
    {
      "epoch": 0.15279241306638566,
      "grad_norm": 0.36889639496803284,
      "learning_rate": 2e-05,
      "loss": 0.5762,
      "step": 1740
    },
    {
      "epoch": 0.15367053038285916,
      "grad_norm": 0.5078188180923462,
      "learning_rate": 2e-05,
      "loss": 0.5804,
      "step": 1750
    },
    {
      "epoch": 0.15454864769933263,
      "grad_norm": 0.3582494854927063,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 1760
    },
    {
      "epoch": 0.1554267650158061,
      "grad_norm": 0.3619888424873352,
      "learning_rate": 2e-05,
      "loss": 0.5952,
      "step": 1770
    },
    {
      "epoch": 0.1563048823322796,
      "grad_norm": 0.3266761898994446,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 1780
    },
    {
      "epoch": 0.15718299964875307,
      "grad_norm": 0.4891241192817688,
      "learning_rate": 2e-05,
      "loss": 0.5671,
      "step": 1790
    },
    {
      "epoch": 0.15806111696522657,
      "grad_norm": 0.4156396985054016,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 1800
    },
    {
      "epoch": 0.15893923428170004,
      "grad_norm": 0.3631264269351959,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 1810
    },
    {
      "epoch": 0.1598173515981735,
      "grad_norm": 0.3365376889705658,
      "learning_rate": 2e-05,
      "loss": 0.5891,
      "step": 1820
    },
    {
      "epoch": 0.160695468914647,
      "grad_norm": 0.3454102873802185,
      "learning_rate": 2e-05,
      "loss": 0.6014,
      "step": 1830
    },
    {
      "epoch": 0.16157358623112048,
      "grad_norm": 0.3719184696674347,
      "learning_rate": 2e-05,
      "loss": 0.588,
      "step": 1840
    },
    {
      "epoch": 0.16245170354759395,
      "grad_norm": 0.42239874601364136,
      "learning_rate": 2e-05,
      "loss": 0.5661,
      "step": 1850
    },
    {
      "epoch": 0.16332982086406744,
      "grad_norm": 0.3765469789505005,
      "learning_rate": 2e-05,
      "loss": 0.5827,
      "step": 1860
    },
    {
      "epoch": 0.1642079381805409,
      "grad_norm": 0.3562774956226349,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 1870
    },
    {
      "epoch": 0.1650860554970144,
      "grad_norm": 0.36806225776672363,
      "learning_rate": 2e-05,
      "loss": 0.5603,
      "step": 1880
    },
    {
      "epoch": 0.16596417281348788,
      "grad_norm": 0.36471012234687805,
      "learning_rate": 2e-05,
      "loss": 0.5665,
      "step": 1890
    },
    {
      "epoch": 0.16684229012996135,
      "grad_norm": 0.3576139509677887,
      "learning_rate": 2e-05,
      "loss": 0.6038,
      "step": 1900
    },
    {
      "epoch": 0.16772040744643485,
      "grad_norm": 0.3674548268318176,
      "learning_rate": 2e-05,
      "loss": 0.5799,
      "step": 1910
    },
    {
      "epoch": 0.16859852476290832,
      "grad_norm": 0.44370803236961365,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 1920
    },
    {
      "epoch": 0.16947664207938182,
      "grad_norm": 0.3467773199081421,
      "learning_rate": 2e-05,
      "loss": 0.5763,
      "step": 1930
    },
    {
      "epoch": 0.1703547593958553,
      "grad_norm": 0.3192700445652008,
      "learning_rate": 2e-05,
      "loss": 0.5752,
      "step": 1940
    },
    {
      "epoch": 0.17123287671232876,
      "grad_norm": 0.3663754165172577,
      "learning_rate": 2e-05,
      "loss": 0.578,
      "step": 1950
    },
    {
      "epoch": 0.17211099402880226,
      "grad_norm": 0.3520260751247406,
      "learning_rate": 2e-05,
      "loss": 0.5913,
      "step": 1960
    },
    {
      "epoch": 0.17298911134527573,
      "grad_norm": 0.3636491298675537,
      "learning_rate": 2e-05,
      "loss": 0.581,
      "step": 1970
    },
    {
      "epoch": 0.17386722866174922,
      "grad_norm": 0.3422032594680786,
      "learning_rate": 2e-05,
      "loss": 0.5583,
      "step": 1980
    },
    {
      "epoch": 0.1747453459782227,
      "grad_norm": 0.3622123897075653,
      "learning_rate": 2e-05,
      "loss": 0.5784,
      "step": 1990
    },
    {
      "epoch": 0.17562346329469616,
      "grad_norm": 0.3840397000312805,
      "learning_rate": 2e-05,
      "loss": 0.5793,
      "step": 2000
    },
    {
      "epoch": 0.17650158061116966,
      "grad_norm": 0.3582792580127716,
      "learning_rate": 2e-05,
      "loss": 0.5635,
      "step": 2010
    },
    {
      "epoch": 0.17737969792764313,
      "grad_norm": 0.3363848328590393,
      "learning_rate": 2e-05,
      "loss": 0.5786,
      "step": 2020
    },
    {
      "epoch": 0.1782578152441166,
      "grad_norm": 0.3684748411178589,
      "learning_rate": 2e-05,
      "loss": 0.5798,
      "step": 2030
    },
    {
      "epoch": 0.1791359325605901,
      "grad_norm": 0.37635061144828796,
      "learning_rate": 2e-05,
      "loss": 0.6066,
      "step": 2040
    },
    {
      "epoch": 0.18001404987706357,
      "grad_norm": 0.3775244653224945,
      "learning_rate": 2e-05,
      "loss": 0.573,
      "step": 2050
    },
    {
      "epoch": 0.18089216719353707,
      "grad_norm": 0.3410365581512451,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 2060
    },
    {
      "epoch": 0.18177028451001054,
      "grad_norm": 0.3407672941684723,
      "learning_rate": 2e-05,
      "loss": 0.5668,
      "step": 2070
    },
    {
      "epoch": 0.182648401826484,
      "grad_norm": 0.35885336995124817,
      "learning_rate": 2e-05,
      "loss": 0.583,
      "step": 2080
    },
    {
      "epoch": 0.1835265191429575,
      "grad_norm": 0.35528841614723206,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 2090
    },
    {
      "epoch": 0.18440463645943098,
      "grad_norm": 0.3379683792591095,
      "learning_rate": 2e-05,
      "loss": 0.5682,
      "step": 2100
    },
    {
      "epoch": 0.18528275377590447,
      "grad_norm": 0.3734015226364136,
      "learning_rate": 2e-05,
      "loss": 0.5755,
      "step": 2110
    },
    {
      "epoch": 0.18616087109237794,
      "grad_norm": 0.34907132387161255,
      "learning_rate": 2e-05,
      "loss": 0.609,
      "step": 2120
    },
    {
      "epoch": 0.18703898840885141,
      "grad_norm": 0.3928704559803009,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 2130
    },
    {
      "epoch": 0.1879171057253249,
      "grad_norm": 0.3575749397277832,
      "learning_rate": 2e-05,
      "loss": 0.5745,
      "step": 2140
    },
    {
      "epoch": 0.18879522304179838,
      "grad_norm": 0.3316913843154907,
      "learning_rate": 2e-05,
      "loss": 0.5696,
      "step": 2150
    },
    {
      "epoch": 0.18967334035827185,
      "grad_norm": 0.35107994079589844,
      "learning_rate": 2e-05,
      "loss": 0.5581,
      "step": 2160
    },
    {
      "epoch": 0.19055145767474535,
      "grad_norm": 0.33369842171669006,
      "learning_rate": 2e-05,
      "loss": 0.575,
      "step": 2170
    },
    {
      "epoch": 0.19142957499121882,
      "grad_norm": 0.3088982403278351,
      "learning_rate": 2e-05,
      "loss": 0.558,
      "step": 2180
    },
    {
      "epoch": 0.19230769230769232,
      "grad_norm": 0.34588614106178284,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 2190
    },
    {
      "epoch": 0.1931858096241658,
      "grad_norm": 0.36946386098861694,
      "learning_rate": 2e-05,
      "loss": 0.5861,
      "step": 2200
    },
    {
      "epoch": 0.19406392694063926,
      "grad_norm": 0.3971337080001831,
      "learning_rate": 2e-05,
      "loss": 0.5911,
      "step": 2210
    },
    {
      "epoch": 0.19494204425711276,
      "grad_norm": 0.40218257904052734,
      "learning_rate": 2e-05,
      "loss": 0.5684,
      "step": 2220
    },
    {
      "epoch": 0.19582016157358623,
      "grad_norm": 0.3587040603160858,
      "learning_rate": 2e-05,
      "loss": 0.5593,
      "step": 2230
    },
    {
      "epoch": 0.19669827889005972,
      "grad_norm": 0.3934634029865265,
      "learning_rate": 2e-05,
      "loss": 0.5904,
      "step": 2240
    },
    {
      "epoch": 0.1975763962065332,
      "grad_norm": 0.37961840629577637,
      "learning_rate": 2e-05,
      "loss": 0.5791,
      "step": 2250
    },
    {
      "epoch": 0.19845451352300666,
      "grad_norm": 0.3834126889705658,
      "learning_rate": 2e-05,
      "loss": 0.5977,
      "step": 2260
    },
    {
      "epoch": 0.19933263083948016,
      "grad_norm": 0.34186139702796936,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 2270
    },
    {
      "epoch": 0.20021074815595363,
      "grad_norm": 0.4032439589500427,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 2280
    },
    {
      "epoch": 0.2010888654724271,
      "grad_norm": 0.3661344349384308,
      "learning_rate": 2e-05,
      "loss": 0.5981,
      "step": 2290
    },
    {
      "epoch": 0.2019669827889006,
      "grad_norm": 0.3640836775302887,
      "learning_rate": 2e-05,
      "loss": 0.5806,
      "step": 2300
    },
    {
      "epoch": 0.20284510010537407,
      "grad_norm": 0.36551639437675476,
      "learning_rate": 2e-05,
      "loss": 0.5811,
      "step": 2310
    },
    {
      "epoch": 0.20372321742184757,
      "grad_norm": 0.37858396768569946,
      "learning_rate": 2e-05,
      "loss": 0.6034,
      "step": 2320
    },
    {
      "epoch": 0.20460133473832104,
      "grad_norm": 0.3028313219547272,
      "learning_rate": 2e-05,
      "loss": 0.5519,
      "step": 2330
    },
    {
      "epoch": 0.2054794520547945,
      "grad_norm": 0.3321625590324402,
      "learning_rate": 2e-05,
      "loss": 0.5731,
      "step": 2340
    },
    {
      "epoch": 0.206357569371268,
      "grad_norm": 0.31999725103378296,
      "learning_rate": 2e-05,
      "loss": 0.5603,
      "step": 2350
    },
    {
      "epoch": 0.20723568668774148,
      "grad_norm": 0.4247737526893616,
      "learning_rate": 2e-05,
      "loss": 0.57,
      "step": 2360
    },
    {
      "epoch": 0.20811380400421498,
      "grad_norm": 0.3929758071899414,
      "learning_rate": 2e-05,
      "loss": 0.5747,
      "step": 2370
    },
    {
      "epoch": 0.20899192132068845,
      "grad_norm": 0.37259769439697266,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 2380
    },
    {
      "epoch": 0.20987003863716192,
      "grad_norm": 0.3574110269546509,
      "learning_rate": 2e-05,
      "loss": 0.5677,
      "step": 2390
    },
    {
      "epoch": 0.2107481559536354,
      "grad_norm": 0.31467244029045105,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 2400
    },
    {
      "epoch": 0.21162627327010888,
      "grad_norm": 0.3683948516845703,
      "learning_rate": 2e-05,
      "loss": 0.5786,
      "step": 2410
    },
    {
      "epoch": 0.21250439058658238,
      "grad_norm": 0.3609021008014679,
      "learning_rate": 2e-05,
      "loss": 0.5764,
      "step": 2420
    },
    {
      "epoch": 0.21338250790305585,
      "grad_norm": 0.3559974133968353,
      "learning_rate": 2e-05,
      "loss": 0.5764,
      "step": 2430
    },
    {
      "epoch": 0.21426062521952932,
      "grad_norm": 0.31802019476890564,
      "learning_rate": 2e-05,
      "loss": 0.5882,
      "step": 2440
    },
    {
      "epoch": 0.21513874253600282,
      "grad_norm": 0.3425753116607666,
      "learning_rate": 2e-05,
      "loss": 0.5723,
      "step": 2450
    },
    {
      "epoch": 0.2160168598524763,
      "grad_norm": 0.37798556685447693,
      "learning_rate": 2e-05,
      "loss": 0.572,
      "step": 2460
    },
    {
      "epoch": 0.21689497716894976,
      "grad_norm": 0.375946968793869,
      "learning_rate": 2e-05,
      "loss": 0.5609,
      "step": 2470
    },
    {
      "epoch": 0.21777309448542326,
      "grad_norm": 0.3843799829483032,
      "learning_rate": 2e-05,
      "loss": 0.5751,
      "step": 2480
    },
    {
      "epoch": 0.21865121180189673,
      "grad_norm": 0.33661726117134094,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 2490
    },
    {
      "epoch": 0.21952932911837023,
      "grad_norm": 0.3537876009941101,
      "learning_rate": 2e-05,
      "loss": 0.5763,
      "step": 2500
    },
    {
      "epoch": 0.2204074464348437,
      "grad_norm": 0.31671857833862305,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 2510
    },
    {
      "epoch": 0.22128556375131717,
      "grad_norm": 0.3895318806171417,
      "learning_rate": 2e-05,
      "loss": 0.5979,
      "step": 2520
    },
    {
      "epoch": 0.22216368106779066,
      "grad_norm": 0.34495359659194946,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 2530
    },
    {
      "epoch": 0.22304179838426413,
      "grad_norm": 0.3425408601760864,
      "learning_rate": 2e-05,
      "loss": 0.612,
      "step": 2540
    },
    {
      "epoch": 0.22391991570073763,
      "grad_norm": 0.3194979429244995,
      "learning_rate": 1.997261400182901e-05,
      "loss": 0.5391,
      "step": 2550
    },
    {
      "epoch": 0.2247980330172111,
      "grad_norm": 0.3179055452346802,
      "learning_rate": 1.9805800746810423e-05,
      "loss": 0.5613,
      "step": 2560
    },
    {
      "epoch": 0.22567615033368457,
      "grad_norm": 0.3644786775112152,
      "learning_rate": 1.9489934220110274e-05,
      "loss": 0.5845,
      "step": 2570
    },
    {
      "epoch": 0.22655426765015807,
      "grad_norm": 0.35235947370529175,
      "learning_rate": 1.9029840364820392e-05,
      "loss": 0.5552,
      "step": 2580
    },
    {
      "epoch": 0.22743238496663154,
      "grad_norm": 0.3413451313972473,
      "learning_rate": 1.8432548690479908e-05,
      "loss": 0.5665,
      "step": 2590
    },
    {
      "epoch": 0.228310502283105,
      "grad_norm": 0.3342607915401459,
      "learning_rate": 1.7707184873239772e-05,
      "loss": 0.5773,
      "step": 2600
    },
    {
      "epoch": 0.2291886195995785,
      "grad_norm": 0.3714098334312439,
      "learning_rate": 1.6864831329902565e-05,
      "loss": 0.6169,
      "step": 2610
    },
    {
      "epoch": 0.23006673691605198,
      "grad_norm": 0.36180657148361206,
      "learning_rate": 1.591835789604667e-05,
      "loss": 0.5728,
      "step": 2620
    },
    {
      "epoch": 0.23094485423252548,
      "grad_norm": 0.4407656788825989,
      "learning_rate": 1.4882225195206459e-05,
      "loss": 0.5634,
      "step": 2630
    },
    {
      "epoch": 0.23182297154899895,
      "grad_norm": 0.32442378997802734,
      "learning_rate": 1.3772263703317695e-05,
      "loss": 0.5415,
      "step": 2640
    },
    {
      "epoch": 0.23270108886547242,
      "grad_norm": 0.34143486618995667,
      "learning_rate": 1.2605431883975547e-05,
      "loss": 0.562,
      "step": 2650
    },
    {
      "epoch": 0.23357920618194591,
      "grad_norm": 0.35474851727485657,
      "learning_rate": 1.1399557089817574e-05,
      "loss": 0.5526,
      "step": 2660
    },
    {
      "epoch": 0.23445732349841938,
      "grad_norm": 0.3544735610485077,
      "learning_rate": 1.0173063188650463e-05,
      "loss": 0.5526,
      "step": 2670
    },
    {
      "epoch": 0.23533544081489288,
      "grad_norm": 0.3359021246433258,
      "learning_rate": 8.944689075764396e-06,
      "loss": 0.5668,
      "step": 2680
    },
    {
      "epoch": 0.23621355813136635,
      "grad_norm": 0.336708128452301,
      "learning_rate": 7.73320237312362e-06,
      "loss": 0.523,
      "step": 2690
    },
    {
      "epoch": 0.23709167544783982,
      "grad_norm": 0.3341606855392456,
      "learning_rate": 6.557112689658629e-06,
      "loss": 0.5536,
      "step": 2700
    },
    {
      "epoch": 0.23796979276431332,
      "grad_norm": 0.37581685185432434,
      "learning_rate": 5.434388823591213e-06,
      "loss": 0.5679,
      "step": 2710
    },
    {
      "epoch": 0.2388479100807868,
      "grad_norm": 0.36387911438941956,
      "learning_rate": 4.382184227495424e-06,
      "loss": 0.5607,
      "step": 2720
    },
    {
      "epoch": 0.23972602739726026,
      "grad_norm": 0.37615805864334106,
      "learning_rate": 3.4165749305560203e-06,
      "loss": 0.5912,
      "step": 2730
    },
    {
      "epoch": 0.24060414471373376,
      "grad_norm": 0.3798007369041443,
      "learning_rate": 2.5523139221595556e-06,
      "loss": 0.5435,
      "step": 2740
    },
    {
      "epoch": 0.24148226203020723,
      "grad_norm": 0.3351341187953949,
      "learning_rate": 1.8026057494499821e-06,
      "loss": 0.5296,
      "step": 2750
    },
    {
      "epoch": 0.24236037934668073,
      "grad_norm": 0.5009398460388184,
      "learning_rate": 1.178904772643223e-06,
      "loss": 0.5691,
      "step": 2760
    },
    {
      "epoch": 0.2432384966631542,
      "grad_norm": 0.33210980892181396,
      "learning_rate": 6.90740160441754e-07,
      "loss": 0.5799,
      "step": 2770
    },
    {
      "epoch": 0.24411661397962767,
      "grad_norm": 0.350495845079422,
      "learning_rate": 3.45570299343632e-07,
      "loss": 0.5807,
      "step": 2780
    },
    {
      "epoch": 0.24499473129610116,
      "grad_norm": 0.37456074357032776,
      "learning_rate": 1.486688412425152e-07,
      "loss": 0.5501,
      "step": 2790
    },
    {
      "epoch": 0.2456094134176326,
      "step": 2797,
      "total_flos": 1.2485883762455871e+19,
      "train_loss": 0.05090188332272292,
      "train_runtime": 2863.9134,
      "train_samples_per_second": 11.72,
      "train_steps_per_second": 0.977
    }
  ],
  "logging_steps": 10,
  "max_steps": 2797,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2543,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2485883762455871e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}