{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.49148226203020723,
  "eval_steps": 500,
  "global_step": 5597,
  "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": 0.32561326026916504,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 2550
    },
    {
      "epoch": 0.2247980330172111,
      "grad_norm": 0.3419996201992035,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 2560
    },
    {
      "epoch": 0.22567615033368457,
      "grad_norm": 0.3595596253871918,
      "learning_rate": 2e-05,
      "loss": 0.5846,
      "step": 2570
    },
    {
      "epoch": 0.22655426765015807,
      "grad_norm": 0.35708296298980713,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 2580
    },
    {
      "epoch": 0.22743238496663154,
      "grad_norm": 0.3468819260597229,
      "learning_rate": 2e-05,
      "loss": 0.5667,
      "step": 2590
    },
    {
      "epoch": 0.228310502283105,
      "grad_norm": 0.34723973274230957,
      "learning_rate": 2e-05,
      "loss": 0.5775,
      "step": 2600
    },
    {
      "epoch": 0.2291886195995785,
      "grad_norm": 0.3546499013900757,
      "learning_rate": 2e-05,
      "loss": 0.6174,
      "step": 2610
    },
    {
      "epoch": 0.23006673691605198,
      "grad_norm": 0.4166103005409241,
      "learning_rate": 2e-05,
      "loss": 0.5735,
      "step": 2620
    },
    {
      "epoch": 0.23094485423252548,
      "grad_norm": 0.4527616500854492,
      "learning_rate": 2e-05,
      "loss": 0.5645,
      "step": 2630
    },
    {
      "epoch": 0.23182297154899895,
      "grad_norm": 0.35841286182403564,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 2640
    },
    {
      "epoch": 0.23270108886547242,
      "grad_norm": 0.3600994646549225,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 2650
    },
    {
      "epoch": 0.23357920618194591,
      "grad_norm": 0.3399403393268585,
      "learning_rate": 2e-05,
      "loss": 0.5545,
      "step": 2660
    },
    {
      "epoch": 0.23445732349841938,
      "grad_norm": 0.38476696610450745,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 2670
    },
    {
      "epoch": 0.23533544081489288,
      "grad_norm": 0.34411126375198364,
      "learning_rate": 2e-05,
      "loss": 0.5696,
      "step": 2680
    },
    {
      "epoch": 0.23621355813136635,
      "grad_norm": 0.3096472918987274,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 2690
    },
    {
      "epoch": 0.23709167544783982,
      "grad_norm": 0.3570089340209961,
      "learning_rate": 2e-05,
      "loss": 0.5564,
      "step": 2700
    },
    {
      "epoch": 0.23796979276431332,
      "grad_norm": 0.3330743908882141,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 2710
    },
    {
      "epoch": 0.2388479100807868,
      "grad_norm": 0.38190898299217224,
      "learning_rate": 2e-05,
      "loss": 0.5643,
      "step": 2720
    },
    {
      "epoch": 0.23972602739726026,
      "grad_norm": 0.3937070965766907,
      "learning_rate": 2e-05,
      "loss": 0.5955,
      "step": 2730
    },
    {
      "epoch": 0.24060414471373376,
      "grad_norm": 0.37956365942955017,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 2740
    },
    {
      "epoch": 0.24148226203020723,
      "grad_norm": 0.33479583263397217,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 2750
    },
    {
      "epoch": 0.24236037934668073,
      "grad_norm": 0.5535967946052551,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 2760
    },
    {
      "epoch": 0.2432384966631542,
      "grad_norm": 0.41816988587379456,
      "learning_rate": 2e-05,
      "loss": 0.5846,
      "step": 2770
    },
    {
      "epoch": 0.24411661397962767,
      "grad_norm": 0.36469870805740356,
      "learning_rate": 2e-05,
      "loss": 0.5847,
      "step": 2780
    },
    {
      "epoch": 0.24499473129610116,
      "grad_norm": 0.3915431797504425,
      "learning_rate": 2e-05,
      "loss": 0.5543,
      "step": 2790
    },
    {
      "epoch": 0.24587284861257463,
      "grad_norm": 0.3385593891143799,
      "learning_rate": 2e-05,
      "loss": 0.5332,
      "step": 2800
    },
    {
      "epoch": 0.24675096592904813,
      "grad_norm": 0.3149898648262024,
      "learning_rate": 2e-05,
      "loss": 0.5469,
      "step": 2810
    },
    {
      "epoch": 0.2476290832455216,
      "grad_norm": 0.35694247484207153,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 2820
    },
    {
      "epoch": 0.24850720056199507,
      "grad_norm": 0.3620472848415375,
      "learning_rate": 2e-05,
      "loss": 0.5795,
      "step": 2830
    },
    {
      "epoch": 0.24938531787846857,
      "grad_norm": 0.3236044645309448,
      "learning_rate": 2e-05,
      "loss": 0.5761,
      "step": 2840
    },
    {
      "epoch": 0.25026343519494204,
      "grad_norm": 0.4055582880973816,
      "learning_rate": 2e-05,
      "loss": 0.577,
      "step": 2850
    },
    {
      "epoch": 0.2511415525114155,
      "grad_norm": 0.3753984868526459,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 2860
    },
    {
      "epoch": 0.252019669827889,
      "grad_norm": 0.33592119812965393,
      "learning_rate": 2e-05,
      "loss": 0.5618,
      "step": 2870
    },
    {
      "epoch": 0.2528977871443625,
      "grad_norm": 0.3353751599788666,
      "learning_rate": 2e-05,
      "loss": 0.5633,
      "step": 2880
    },
    {
      "epoch": 0.253775904460836,
      "grad_norm": 0.3403836190700531,
      "learning_rate": 2e-05,
      "loss": 0.5644,
      "step": 2890
    },
    {
      "epoch": 0.25465402177730945,
      "grad_norm": 0.32027125358581543,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 2900
    },
    {
      "epoch": 0.2555321390937829,
      "grad_norm": 0.3449241518974304,
      "learning_rate": 2e-05,
      "loss": 0.599,
      "step": 2910
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.3605469763278961,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 2920
    },
    {
      "epoch": 0.2572883737267299,
      "grad_norm": 0.3666762411594391,
      "learning_rate": 2e-05,
      "loss": 0.5835,
      "step": 2930
    },
    {
      "epoch": 0.2581664910432034,
      "grad_norm": 0.34501874446868896,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 2940
    },
    {
      "epoch": 0.25904460835967685,
      "grad_norm": 0.3615836203098297,
      "learning_rate": 2e-05,
      "loss": 0.5911,
      "step": 2950
    },
    {
      "epoch": 0.2599227256761503,
      "grad_norm": 0.3566023111343384,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 2960
    },
    {
      "epoch": 0.2608008429926238,
      "grad_norm": 0.36601170897483826,
      "learning_rate": 2e-05,
      "loss": 0.5908,
      "step": 2970
    },
    {
      "epoch": 0.2616789603090973,
      "grad_norm": 0.3530998229980469,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 2980
    },
    {
      "epoch": 0.2625570776255708,
      "grad_norm": 0.3485164940357208,
      "learning_rate": 2e-05,
      "loss": 0.561,
      "step": 2990
    },
    {
      "epoch": 0.26343519494204426,
      "grad_norm": 0.3305230736732483,
      "learning_rate": 2e-05,
      "loss": 0.5612,
      "step": 3000
    },
    {
      "epoch": 0.26431331225851773,
      "grad_norm": 0.3350762128829956,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 3010
    },
    {
      "epoch": 0.2651914295749912,
      "grad_norm": 0.34024325013160706,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 3020
    },
    {
      "epoch": 0.2660695468914647,
      "grad_norm": 0.3802330791950226,
      "learning_rate": 2e-05,
      "loss": 0.5751,
      "step": 3030
    },
    {
      "epoch": 0.2669476642079382,
      "grad_norm": 0.3376270532608032,
      "learning_rate": 2e-05,
      "loss": 0.5601,
      "step": 3040
    },
    {
      "epoch": 0.26782578152441167,
      "grad_norm": 0.3571026027202606,
      "learning_rate": 2e-05,
      "loss": 0.5728,
      "step": 3050
    },
    {
      "epoch": 0.26870389884088514,
      "grad_norm": 0.33023712038993835,
      "learning_rate": 2e-05,
      "loss": 0.5959,
      "step": 3060
    },
    {
      "epoch": 0.2695820161573586,
      "grad_norm": 0.3749672770500183,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 3070
    },
    {
      "epoch": 0.27046013347383213,
      "grad_norm": 0.3238333463668823,
      "learning_rate": 2e-05,
      "loss": 0.5606,
      "step": 3080
    },
    {
      "epoch": 0.2713382507903056,
      "grad_norm": 0.35644635558128357,
      "learning_rate": 2e-05,
      "loss": 0.5763,
      "step": 3090
    },
    {
      "epoch": 0.27221636810677907,
      "grad_norm": 0.3674982786178589,
      "learning_rate": 2e-05,
      "loss": 0.5593,
      "step": 3100
    },
    {
      "epoch": 0.27309448542325254,
      "grad_norm": 0.3481002449989319,
      "learning_rate": 2e-05,
      "loss": 0.5621,
      "step": 3110
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 0.35084569454193115,
      "learning_rate": 2e-05,
      "loss": 0.5815,
      "step": 3120
    },
    {
      "epoch": 0.2748507200561995,
      "grad_norm": 0.41141173243522644,
      "learning_rate": 2e-05,
      "loss": 0.5774,
      "step": 3130
    },
    {
      "epoch": 0.275728837372673,
      "grad_norm": 0.3726920485496521,
      "learning_rate": 2e-05,
      "loss": 0.556,
      "step": 3140
    },
    {
      "epoch": 0.2766069546891465,
      "grad_norm": 0.3815704584121704,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 3150
    },
    {
      "epoch": 0.27748507200561995,
      "grad_norm": 0.3393665850162506,
      "learning_rate": 2e-05,
      "loss": 0.5789,
      "step": 3160
    },
    {
      "epoch": 0.2783631893220934,
      "grad_norm": 0.3703439235687256,
      "learning_rate": 2e-05,
      "loss": 0.6027,
      "step": 3170
    },
    {
      "epoch": 0.2792413066385669,
      "grad_norm": 0.3668241798877716,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 3180
    },
    {
      "epoch": 0.2801194239550404,
      "grad_norm": 0.3330550491809845,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 3190
    },
    {
      "epoch": 0.2809975412715139,
      "grad_norm": 0.333389550447464,
      "learning_rate": 2e-05,
      "loss": 0.5444,
      "step": 3200
    },
    {
      "epoch": 0.28187565858798735,
      "grad_norm": 0.3240389823913574,
      "learning_rate": 2e-05,
      "loss": 0.5779,
      "step": 3210
    },
    {
      "epoch": 0.2827537759044608,
      "grad_norm": 0.3601751923561096,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 3220
    },
    {
      "epoch": 0.2836318932209343,
      "grad_norm": 0.38156554102897644,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 3230
    },
    {
      "epoch": 0.2845100105374078,
      "grad_norm": 0.3225893974304199,
      "learning_rate": 2e-05,
      "loss": 0.5822,
      "step": 3240
    },
    {
      "epoch": 0.2853881278538813,
      "grad_norm": 0.3563006520271301,
      "learning_rate": 2e-05,
      "loss": 0.5876,
      "step": 3250
    },
    {
      "epoch": 0.28626624517035476,
      "grad_norm": 0.3583647310733795,
      "learning_rate": 2e-05,
      "loss": 0.5689,
      "step": 3260
    },
    {
      "epoch": 0.28714436248682823,
      "grad_norm": 0.3173294961452484,
      "learning_rate": 2e-05,
      "loss": 0.5335,
      "step": 3270
    },
    {
      "epoch": 0.2880224798033017,
      "grad_norm": 0.3422831892967224,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 3280
    },
    {
      "epoch": 0.2889005971197752,
      "grad_norm": 0.33035796880722046,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 3290
    },
    {
      "epoch": 0.2897787144362487,
      "grad_norm": 0.393304705619812,
      "learning_rate": 2e-05,
      "loss": 0.5907,
      "step": 3300
    },
    {
      "epoch": 0.29065683175272217,
      "grad_norm": 0.5257350206375122,
      "learning_rate": 2e-05,
      "loss": 0.5571,
      "step": 3310
    },
    {
      "epoch": 0.29153494906919564,
      "grad_norm": 0.3335498571395874,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 3320
    },
    {
      "epoch": 0.2924130663856691,
      "grad_norm": 0.35538333654403687,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 3330
    },
    {
      "epoch": 0.29329118370214263,
      "grad_norm": 0.3739526867866516,
      "learning_rate": 2e-05,
      "loss": 0.5581,
      "step": 3340
    },
    {
      "epoch": 0.2941693010186161,
      "grad_norm": 0.38300976157188416,
      "learning_rate": 2e-05,
      "loss": 0.5692,
      "step": 3350
    },
    {
      "epoch": 0.2950474183350896,
      "grad_norm": 0.39358460903167725,
      "learning_rate": 2e-05,
      "loss": 0.5966,
      "step": 3360
    },
    {
      "epoch": 0.29592553565156304,
      "grad_norm": 0.34571489691734314,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 3370
    },
    {
      "epoch": 0.2968036529680365,
      "grad_norm": 0.36871588230133057,
      "learning_rate": 2e-05,
      "loss": 0.5704,
      "step": 3380
    },
    {
      "epoch": 0.29768177028451,
      "grad_norm": 0.45794662833213806,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 3390
    },
    {
      "epoch": 0.2985598876009835,
      "grad_norm": 0.3375195562839508,
      "learning_rate": 2e-05,
      "loss": 0.5595,
      "step": 3400
    },
    {
      "epoch": 0.299438004917457,
      "grad_norm": 0.3880457878112793,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 3410
    },
    {
      "epoch": 0.30031612223393045,
      "grad_norm": 0.34311389923095703,
      "learning_rate": 2e-05,
      "loss": 0.5595,
      "step": 3420
    },
    {
      "epoch": 0.3011942395504039,
      "grad_norm": 0.4251992404460907,
      "learning_rate": 2e-05,
      "loss": 0.5805,
      "step": 3430
    },
    {
      "epoch": 0.3020723568668774,
      "grad_norm": 0.3721943199634552,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 3440
    },
    {
      "epoch": 0.3029504741833509,
      "grad_norm": 0.35362544655799866,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 3450
    },
    {
      "epoch": 0.3038285914998244,
      "grad_norm": 0.37591123580932617,
      "learning_rate": 2e-05,
      "loss": 0.5778,
      "step": 3460
    },
    {
      "epoch": 0.30470670881629786,
      "grad_norm": 0.3445957601070404,
      "learning_rate": 2e-05,
      "loss": 0.5556,
      "step": 3470
    },
    {
      "epoch": 0.3055848261327713,
      "grad_norm": 0.41066107153892517,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 3480
    },
    {
      "epoch": 0.3064629434492448,
      "grad_norm": 0.3666580021381378,
      "learning_rate": 2e-05,
      "loss": 0.5618,
      "step": 3490
    },
    {
      "epoch": 0.3073410607657183,
      "grad_norm": 0.3518039286136627,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 3500
    },
    {
      "epoch": 0.3082191780821918,
      "grad_norm": 0.41677379608154297,
      "learning_rate": 2e-05,
      "loss": 0.5814,
      "step": 3510
    },
    {
      "epoch": 0.30909729539866526,
      "grad_norm": 0.36042261123657227,
      "learning_rate": 2e-05,
      "loss": 0.5981,
      "step": 3520
    },
    {
      "epoch": 0.30997541271513873,
      "grad_norm": 0.31846997141838074,
      "learning_rate": 2e-05,
      "loss": 0.5813,
      "step": 3530
    },
    {
      "epoch": 0.3108535300316122,
      "grad_norm": 0.35640856623649597,
      "learning_rate": 2e-05,
      "loss": 0.5367,
      "step": 3540
    },
    {
      "epoch": 0.3117316473480857,
      "grad_norm": 0.6977357268333435,
      "learning_rate": 2e-05,
      "loss": 0.5531,
      "step": 3550
    },
    {
      "epoch": 0.3126097646645592,
      "grad_norm": 0.35111236572265625,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 3560
    },
    {
      "epoch": 0.31348788198103267,
      "grad_norm": 0.362119197845459,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 3570
    },
    {
      "epoch": 0.31436599929750614,
      "grad_norm": 0.34995242953300476,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 3580
    },
    {
      "epoch": 0.3152441166139796,
      "grad_norm": 0.3599854111671448,
      "learning_rate": 2e-05,
      "loss": 0.5538,
      "step": 3590
    },
    {
      "epoch": 0.31612223393045313,
      "grad_norm": 0.367215096950531,
      "learning_rate": 2e-05,
      "loss": 0.585,
      "step": 3600
    },
    {
      "epoch": 0.3170003512469266,
      "grad_norm": 0.36077040433883667,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 3610
    },
    {
      "epoch": 0.3178784685634001,
      "grad_norm": 0.3283092677593231,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 3620
    },
    {
      "epoch": 0.31875658587987354,
      "grad_norm": 0.3335152864456177,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 3630
    },
    {
      "epoch": 0.319634703196347,
      "grad_norm": 0.3362371325492859,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 3640
    },
    {
      "epoch": 0.32051282051282054,
      "grad_norm": 0.3582964241504669,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 3650
    },
    {
      "epoch": 0.321390937829294,
      "grad_norm": 0.3321240544319153,
      "learning_rate": 2e-05,
      "loss": 0.5661,
      "step": 3660
    },
    {
      "epoch": 0.3222690551457675,
      "grad_norm": 0.34263285994529724,
      "learning_rate": 2e-05,
      "loss": 0.5603,
      "step": 3670
    },
    {
      "epoch": 0.32314717246224095,
      "grad_norm": 0.34714293479919434,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 3680
    },
    {
      "epoch": 0.3240252897787144,
      "grad_norm": 0.4024375379085541,
      "learning_rate": 2e-05,
      "loss": 0.5663,
      "step": 3690
    },
    {
      "epoch": 0.3249034070951879,
      "grad_norm": 0.35162466764450073,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 3700
    },
    {
      "epoch": 0.3257815244116614,
      "grad_norm": 0.3222830593585968,
      "learning_rate": 2e-05,
      "loss": 0.536,
      "step": 3710
    },
    {
      "epoch": 0.3266596417281349,
      "grad_norm": 0.3560235798358917,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 3720
    },
    {
      "epoch": 0.32753775904460836,
      "grad_norm": 0.36119699478149414,
      "learning_rate": 2e-05,
      "loss": 0.5689,
      "step": 3730
    },
    {
      "epoch": 0.3284158763610818,
      "grad_norm": 0.3181910514831543,
      "learning_rate": 2e-05,
      "loss": 0.5656,
      "step": 3740
    },
    {
      "epoch": 0.3292939936775553,
      "grad_norm": 0.355177104473114,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 3750
    },
    {
      "epoch": 0.3301721109940288,
      "grad_norm": 0.3285183012485504,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 3760
    },
    {
      "epoch": 0.3310502283105023,
      "grad_norm": 0.3440050780773163,
      "learning_rate": 2e-05,
      "loss": 0.5783,
      "step": 3770
    },
    {
      "epoch": 0.33192834562697576,
      "grad_norm": 0.38265684247016907,
      "learning_rate": 2e-05,
      "loss": 0.5614,
      "step": 3780
    },
    {
      "epoch": 0.33280646294344923,
      "grad_norm": 0.33423545956611633,
      "learning_rate": 2e-05,
      "loss": 0.5393,
      "step": 3790
    },
    {
      "epoch": 0.3336845802599227,
      "grad_norm": 0.31699854135513306,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 3800
    },
    {
      "epoch": 0.33456269757639623,
      "grad_norm": 0.3439178168773651,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 3810
    },
    {
      "epoch": 0.3354408148928697,
      "grad_norm": 0.37247344851493835,
      "learning_rate": 2e-05,
      "loss": 0.5429,
      "step": 3820
    },
    {
      "epoch": 0.33631893220934317,
      "grad_norm": 0.3486614227294922,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 3830
    },
    {
      "epoch": 0.33719704952581664,
      "grad_norm": 0.4223357141017914,
      "learning_rate": 2e-05,
      "loss": 0.5568,
      "step": 3840
    },
    {
      "epoch": 0.3380751668422901,
      "grad_norm": 0.34644559025764465,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 3850
    },
    {
      "epoch": 0.33895328415876363,
      "grad_norm": 0.3471311330795288,
      "learning_rate": 2e-05,
      "loss": 0.5617,
      "step": 3860
    },
    {
      "epoch": 0.3398314014752371,
      "grad_norm": 0.33086150884628296,
      "learning_rate": 2e-05,
      "loss": 0.5801,
      "step": 3870
    },
    {
      "epoch": 0.3407095187917106,
      "grad_norm": 0.3785061240196228,
      "learning_rate": 2e-05,
      "loss": 0.5628,
      "step": 3880
    },
    {
      "epoch": 0.34158763610818405,
      "grad_norm": 0.38897570967674255,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 3890
    },
    {
      "epoch": 0.3424657534246575,
      "grad_norm": 0.4057118594646454,
      "learning_rate": 2e-05,
      "loss": 0.5637,
      "step": 3900
    },
    {
      "epoch": 0.34334387074113104,
      "grad_norm": 0.3396783769130707,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 3910
    },
    {
      "epoch": 0.3442219880576045,
      "grad_norm": 0.40010005235671997,
      "learning_rate": 2e-05,
      "loss": 0.5412,
      "step": 3920
    },
    {
      "epoch": 0.345100105374078,
      "grad_norm": 0.3718085289001465,
      "learning_rate": 2e-05,
      "loss": 0.5619,
      "step": 3930
    },
    {
      "epoch": 0.34597822269055145,
      "grad_norm": 0.3636755347251892,
      "learning_rate": 2e-05,
      "loss": 0.5774,
      "step": 3940
    },
    {
      "epoch": 0.3468563400070249,
      "grad_norm": 0.34388771653175354,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 3950
    },
    {
      "epoch": 0.34773445732349845,
      "grad_norm": 0.3660244345664978,
      "learning_rate": 2e-05,
      "loss": 0.568,
      "step": 3960
    },
    {
      "epoch": 0.3486125746399719,
      "grad_norm": 0.36620253324508667,
      "learning_rate": 2e-05,
      "loss": 0.571,
      "step": 3970
    },
    {
      "epoch": 0.3494906919564454,
      "grad_norm": 0.3290777802467346,
      "learning_rate": 2e-05,
      "loss": 0.5695,
      "step": 3980
    },
    {
      "epoch": 0.35036880927291886,
      "grad_norm": 0.44158679246902466,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 3990
    },
    {
      "epoch": 0.3512469265893923,
      "grad_norm": 0.3711128532886505,
      "learning_rate": 2e-05,
      "loss": 0.573,
      "step": 4000
    },
    {
      "epoch": 0.3521250439058658,
      "grad_norm": 0.354728639125824,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 4010
    },
    {
      "epoch": 0.3530031612223393,
      "grad_norm": 0.37003782391548157,
      "learning_rate": 2e-05,
      "loss": 0.5713,
      "step": 4020
    },
    {
      "epoch": 0.3538812785388128,
      "grad_norm": 0.3585514724254608,
      "learning_rate": 2e-05,
      "loss": 0.5366,
      "step": 4030
    },
    {
      "epoch": 0.35475939585528626,
      "grad_norm": 0.3750348389148712,
      "learning_rate": 2e-05,
      "loss": 0.5735,
      "step": 4040
    },
    {
      "epoch": 0.35563751317175973,
      "grad_norm": 0.3458133637905121,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 4050
    },
    {
      "epoch": 0.3565156304882332,
      "grad_norm": 0.353854238986969,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 4060
    },
    {
      "epoch": 0.35739374780470673,
      "grad_norm": 0.571679413318634,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 4070
    },
    {
      "epoch": 0.3582718651211802,
      "grad_norm": 0.40862205624580383,
      "learning_rate": 2e-05,
      "loss": 0.5598,
      "step": 4080
    },
    {
      "epoch": 0.35914998243765367,
      "grad_norm": 0.33414268493652344,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 4090
    },
    {
      "epoch": 0.36002809975412714,
      "grad_norm": 0.32345816493034363,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 4100
    },
    {
      "epoch": 0.3609062170706006,
      "grad_norm": 0.37067991495132446,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4110
    },
    {
      "epoch": 0.36178433438707414,
      "grad_norm": 0.32833150029182434,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 4120
    },
    {
      "epoch": 0.3626624517035476,
      "grad_norm": 0.3547900319099426,
      "learning_rate": 2e-05,
      "loss": 0.5622,
      "step": 4130
    },
    {
      "epoch": 0.3635405690200211,
      "grad_norm": 0.3555840849876404,
      "learning_rate": 2e-05,
      "loss": 0.5614,
      "step": 4140
    },
    {
      "epoch": 0.36441868633649455,
      "grad_norm": 0.35306933522224426,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 4150
    },
    {
      "epoch": 0.365296803652968,
      "grad_norm": 0.3732573390007019,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 4160
    },
    {
      "epoch": 0.36617492096944154,
      "grad_norm": 0.33010610938072205,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 4170
    },
    {
      "epoch": 0.367053038285915,
      "grad_norm": 0.34545499086380005,
      "learning_rate": 2e-05,
      "loss": 0.5388,
      "step": 4180
    },
    {
      "epoch": 0.3679311556023885,
      "grad_norm": 0.3661598265171051,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 4190
    },
    {
      "epoch": 0.36880927291886195,
      "grad_norm": 0.47115686535835266,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 4200
    },
    {
      "epoch": 0.3696873902353354,
      "grad_norm": 0.3498151898384094,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 4210
    },
    {
      "epoch": 0.37056550755180895,
      "grad_norm": 0.3819350004196167,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 4220
    },
    {
      "epoch": 0.3714436248682824,
      "grad_norm": 0.38068294525146484,
      "learning_rate": 2e-05,
      "loss": 0.6169,
      "step": 4230
    },
    {
      "epoch": 0.3723217421847559,
      "grad_norm": 0.3401268720626831,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 4240
    },
    {
      "epoch": 0.37319985950122936,
      "grad_norm": 0.3547501266002655,
      "learning_rate": 2e-05,
      "loss": 0.6074,
      "step": 4250
    },
    {
      "epoch": 0.37407797681770283,
      "grad_norm": 0.34808698296546936,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 4260
    },
    {
      "epoch": 0.3749560941341763,
      "grad_norm": 0.3525558114051819,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 4270
    },
    {
      "epoch": 0.3758342114506498,
      "grad_norm": 0.3462215065956116,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 4280
    },
    {
      "epoch": 0.3767123287671233,
      "grad_norm": 0.3861493170261383,
      "learning_rate": 2e-05,
      "loss": 0.5461,
      "step": 4290
    },
    {
      "epoch": 0.37759044608359676,
      "grad_norm": 0.34702154994010925,
      "learning_rate": 2e-05,
      "loss": 0.5548,
      "step": 4300
    },
    {
      "epoch": 0.37846856340007023,
      "grad_norm": 0.384822279214859,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 4310
    },
    {
      "epoch": 0.3793466807165437,
      "grad_norm": 0.38018009066581726,
      "learning_rate": 2e-05,
      "loss": 0.5326,
      "step": 4320
    },
    {
      "epoch": 0.38022479803301723,
      "grad_norm": 0.3570401966571808,
      "learning_rate": 2e-05,
      "loss": 0.5697,
      "step": 4330
    },
    {
      "epoch": 0.3811029153494907,
      "grad_norm": 0.3081170618534088,
      "learning_rate": 2e-05,
      "loss": 0.5485,
      "step": 4340
    },
    {
      "epoch": 0.38198103266596417,
      "grad_norm": 0.3312861919403076,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 4350
    },
    {
      "epoch": 0.38285914998243764,
      "grad_norm": 0.34447282552719116,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 4360
    },
    {
      "epoch": 0.3837372672989111,
      "grad_norm": 0.355824738740921,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 4370
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.3751704692840576,
      "learning_rate": 2e-05,
      "loss": 0.5455,
      "step": 4380
    },
    {
      "epoch": 0.3854935019318581,
      "grad_norm": 0.3565281629562378,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 4390
    },
    {
      "epoch": 0.3863716192483316,
      "grad_norm": 0.3726107180118561,
      "learning_rate": 2e-05,
      "loss": 0.5668,
      "step": 4400
    },
    {
      "epoch": 0.38724973656480505,
      "grad_norm": 0.3569621443748474,
      "learning_rate": 2e-05,
      "loss": 0.5582,
      "step": 4410
    },
    {
      "epoch": 0.3881278538812785,
      "grad_norm": 0.3850657641887665,
      "learning_rate": 2e-05,
      "loss": 0.551,
      "step": 4420
    },
    {
      "epoch": 0.38900597119775204,
      "grad_norm": 0.3001329302787781,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 4430
    },
    {
      "epoch": 0.3898840885142255,
      "grad_norm": 0.3708983063697815,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 4440
    },
    {
      "epoch": 0.390762205830699,
      "grad_norm": 0.3833453953266144,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 4450
    },
    {
      "epoch": 0.39164032314717245,
      "grad_norm": 0.3339364528656006,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 4460
    },
    {
      "epoch": 0.3925184404636459,
      "grad_norm": 0.3600437641143799,
      "learning_rate": 2e-05,
      "loss": 0.558,
      "step": 4470
    },
    {
      "epoch": 0.39339655778011945,
      "grad_norm": 0.3660196363925934,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 4480
    },
    {
      "epoch": 0.3942746750965929,
      "grad_norm": 0.41776615381240845,
      "learning_rate": 2e-05,
      "loss": 0.5558,
      "step": 4490
    },
    {
      "epoch": 0.3951527924130664,
      "grad_norm": 0.32263505458831787,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 4500
    },
    {
      "epoch": 0.39603090972953986,
      "grad_norm": 0.36971616744995117,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 4510
    },
    {
      "epoch": 0.39690902704601333,
      "grad_norm": 0.361156702041626,
      "learning_rate": 2e-05,
      "loss": 0.5782,
      "step": 4520
    },
    {
      "epoch": 0.39778714436248686,
      "grad_norm": 0.3905126452445984,
      "learning_rate": 2e-05,
      "loss": 0.5847,
      "step": 4530
    },
    {
      "epoch": 0.3986652616789603,
      "grad_norm": 0.4000948667526245,
      "learning_rate": 2e-05,
      "loss": 0.5708,
      "step": 4540
    },
    {
      "epoch": 0.3995433789954338,
      "grad_norm": 0.34038838744163513,
      "learning_rate": 2e-05,
      "loss": 0.5488,
      "step": 4550
    },
    {
      "epoch": 0.40042149631190727,
      "grad_norm": 0.3478802442550659,
      "learning_rate": 2e-05,
      "loss": 0.5483,
      "step": 4560
    },
    {
      "epoch": 0.40129961362838074,
      "grad_norm": 0.3798333704471588,
      "learning_rate": 2e-05,
      "loss": 0.5787,
      "step": 4570
    },
    {
      "epoch": 0.4021777309448542,
      "grad_norm": 0.3496931195259094,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 4580
    },
    {
      "epoch": 0.40305584826132773,
      "grad_norm": 0.34191030263900757,
      "learning_rate": 2e-05,
      "loss": 0.5455,
      "step": 4590
    },
    {
      "epoch": 0.4039339655778012,
      "grad_norm": 0.3460274934768677,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 4600
    },
    {
      "epoch": 0.40481208289427467,
      "grad_norm": 0.34307581186294556,
      "learning_rate": 2e-05,
      "loss": 0.5548,
      "step": 4610
    },
    {
      "epoch": 0.40569020021074814,
      "grad_norm": 0.3330712914466858,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 4620
    },
    {
      "epoch": 0.4065683175272216,
      "grad_norm": 0.4285266697406769,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 4630
    },
    {
      "epoch": 0.40744643484369514,
      "grad_norm": 0.3933718800544739,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 4640
    },
    {
      "epoch": 0.4083245521601686,
      "grad_norm": 0.39737027883529663,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 4650
    },
    {
      "epoch": 0.4092026694766421,
      "grad_norm": 0.3589046001434326,
      "learning_rate": 2e-05,
      "loss": 0.5666,
      "step": 4660
    },
    {
      "epoch": 0.41008078679311555,
      "grad_norm": 0.3344109058380127,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4670
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 0.3640746474266052,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 4680
    },
    {
      "epoch": 0.41183702142606254,
      "grad_norm": 0.3989454209804535,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 4690
    },
    {
      "epoch": 0.412715138742536,
      "grad_norm": 0.38026702404022217,
      "learning_rate": 2e-05,
      "loss": 0.5692,
      "step": 4700
    },
    {
      "epoch": 0.4135932560590095,
      "grad_norm": 0.352771133184433,
      "learning_rate": 2e-05,
      "loss": 0.5502,
      "step": 4710
    },
    {
      "epoch": 0.41447137337548295,
      "grad_norm": 0.42258620262145996,
      "learning_rate": 2e-05,
      "loss": 0.5496,
      "step": 4720
    },
    {
      "epoch": 0.4153494906919564,
      "grad_norm": 0.3527478277683258,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 4730
    },
    {
      "epoch": 0.41622760800842995,
      "grad_norm": 0.3126488924026489,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 4740
    },
    {
      "epoch": 0.4171057253249034,
      "grad_norm": 0.3601631820201874,
      "learning_rate": 2e-05,
      "loss": 0.5727,
      "step": 4750
    },
    {
      "epoch": 0.4179838426413769,
      "grad_norm": 0.3719983994960785,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 4760
    },
    {
      "epoch": 0.41886195995785036,
      "grad_norm": 0.37243515253067017,
      "learning_rate": 2e-05,
      "loss": 0.5655,
      "step": 4770
    },
    {
      "epoch": 0.41974007727432383,
      "grad_norm": 0.33687853813171387,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 4780
    },
    {
      "epoch": 0.42061819459079736,
      "grad_norm": 0.3543761372566223,
      "learning_rate": 2e-05,
      "loss": 0.5631,
      "step": 4790
    },
    {
      "epoch": 0.4214963119072708,
      "grad_norm": 0.35337427258491516,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 4800
    },
    {
      "epoch": 0.4223744292237443,
      "grad_norm": 0.36009758710861206,
      "learning_rate": 2e-05,
      "loss": 0.5705,
      "step": 4810
    },
    {
      "epoch": 0.42325254654021777,
      "grad_norm": 0.3840942978858948,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 4820
    },
    {
      "epoch": 0.42413066385669124,
      "grad_norm": 0.38357433676719666,
      "learning_rate": 2e-05,
      "loss": 0.5341,
      "step": 4830
    },
    {
      "epoch": 0.42500878117316476,
      "grad_norm": 0.3523583710193634,
      "learning_rate": 2e-05,
      "loss": 0.5698,
      "step": 4840
    },
    {
      "epoch": 0.42588689848963823,
      "grad_norm": 0.3704957067966461,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 4850
    },
    {
      "epoch": 0.4267650158061117,
      "grad_norm": 0.44775083661079407,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 4860
    },
    {
      "epoch": 0.4276431331225852,
      "grad_norm": 0.3357139229774475,
      "learning_rate": 2e-05,
      "loss": 0.5699,
      "step": 4870
    },
    {
      "epoch": 0.42852125043905864,
      "grad_norm": 0.3484988212585449,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 4880
    },
    {
      "epoch": 0.4293993677555321,
      "grad_norm": 0.37846001982688904,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 4890
    },
    {
      "epoch": 0.43027748507200564,
      "grad_norm": 0.357652485370636,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 4900
    },
    {
      "epoch": 0.4311556023884791,
      "grad_norm": 0.3637485206127167,
      "learning_rate": 2e-05,
      "loss": 0.5516,
      "step": 4910
    },
    {
      "epoch": 0.4320337197049526,
      "grad_norm": 0.39857861399650574,
      "learning_rate": 2e-05,
      "loss": 0.5693,
      "step": 4920
    },
    {
      "epoch": 0.43291183702142605,
      "grad_norm": 0.3457987904548645,
      "learning_rate": 2e-05,
      "loss": 0.5615,
      "step": 4930
    },
    {
      "epoch": 0.4337899543378995,
      "grad_norm": 0.4940570294857025,
      "learning_rate": 2e-05,
      "loss": 0.5709,
      "step": 4940
    },
    {
      "epoch": 0.43466807165437304,
      "grad_norm": 0.3299383521080017,
      "learning_rate": 2e-05,
      "loss": 0.5578,
      "step": 4950
    },
    {
      "epoch": 0.4355461889708465,
      "grad_norm": 0.3799186050891876,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 4960
    },
    {
      "epoch": 0.43642430628732,
      "grad_norm": 0.4419582784175873,
      "learning_rate": 2e-05,
      "loss": 0.5417,
      "step": 4970
    },
    {
      "epoch": 0.43730242360379346,
      "grad_norm": 0.4167982041835785,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 4980
    },
    {
      "epoch": 0.4381805409202669,
      "grad_norm": 0.329133003950119,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 4990
    },
    {
      "epoch": 0.43905865823674045,
      "grad_norm": 0.3540366291999817,
      "learning_rate": 2e-05,
      "loss": 0.5416,
      "step": 5000
    },
    {
      "epoch": 0.4399367755532139,
      "grad_norm": 0.3827556073665619,
      "learning_rate": 2e-05,
      "loss": 0.5591,
      "step": 5010
    },
    {
      "epoch": 0.4408148928696874,
      "grad_norm": 0.37153664231300354,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 5020
    },
    {
      "epoch": 0.44169301018616086,
      "grad_norm": 0.44997745752334595,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 5030
    },
    {
      "epoch": 0.44257112750263433,
      "grad_norm": 0.426015704870224,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 5040
    },
    {
      "epoch": 0.44344924481910786,
      "grad_norm": 0.38029444217681885,
      "learning_rate": 2e-05,
      "loss": 0.5685,
      "step": 5050
    },
    {
      "epoch": 0.4443273621355813,
      "grad_norm": 0.37514182925224304,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 5060
    },
    {
      "epoch": 0.4452054794520548,
      "grad_norm": 0.3483792841434479,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 5070
    },
    {
      "epoch": 0.44608359676852827,
      "grad_norm": 0.363000750541687,
      "learning_rate": 2e-05,
      "loss": 0.5378,
      "step": 5080
    },
    {
      "epoch": 0.44696171408500174,
      "grad_norm": 0.3529207110404968,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.5237,
      "step": 5090
    },
    {
      "epoch": 0.44783983140147526,
      "grad_norm": 0.3317299485206604,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.5515,
      "step": 5100
    },
    {
      "epoch": 0.44871794871794873,
      "grad_norm": 0.3519271910190582,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5664,
      "step": 5110
    },
    {
      "epoch": 0.4495960660344222,
      "grad_norm": 0.37423521280288696,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.5433,
      "step": 5120
    },
    {
      "epoch": 0.4504741833508957,
      "grad_norm": 0.4020208716392517,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.5587,
      "step": 5130
    },
    {
      "epoch": 0.45135230066736914,
      "grad_norm": 0.36146169900894165,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5384,
      "step": 5140
    },
    {
      "epoch": 0.4522304179838426,
      "grad_norm": 0.3452405333518982,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.5781,
      "step": 5150
    },
    {
      "epoch": 0.45310853530031614,
      "grad_norm": 0.35994046926498413,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.5431,
      "step": 5160
    },
    {
      "epoch": 0.4539866526167896,
      "grad_norm": 0.3277939558029175,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.5481,
      "step": 5170
    },
    {
      "epoch": 0.4548647699332631,
      "grad_norm": 0.3314633071422577,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5425,
      "step": 5180
    },
    {
      "epoch": 0.45574288724973655,
      "grad_norm": 0.3389431834220886,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.5416,
      "step": 5190
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 0.3273237347602844,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.5785,
      "step": 5200
    },
    {
      "epoch": 0.45749912188268355,
      "grad_norm": 0.34634122252464294,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5672,
      "step": 5210
    },
    {
      "epoch": 0.458377239199157,
      "grad_norm": 0.38971054553985596,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.5397,
      "step": 5220
    },
    {
      "epoch": 0.4592553565156305,
      "grad_norm": 0.32956525683403015,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5483,
      "step": 5230
    },
    {
      "epoch": 0.46013347383210396,
      "grad_norm": 0.39701682329177856,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.5515,
      "step": 5240
    },
    {
      "epoch": 0.4610115911485774,
      "grad_norm": 0.3715463876724243,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.5415,
      "step": 5250
    },
    {
      "epoch": 0.46188970846505095,
      "grad_norm": 0.47152459621429443,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.5342,
      "step": 5260
    },
    {
      "epoch": 0.4627678257815244,
      "grad_norm": 0.35315677523612976,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.5272,
      "step": 5270
    },
    {
      "epoch": 0.4636459430979979,
      "grad_norm": 0.38288676738739014,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.5485,
      "step": 5280
    },
    {
      "epoch": 0.46452406041447136,
      "grad_norm": 0.3380102813243866,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.5161,
      "step": 5290
    },
    {
      "epoch": 0.46540217773094483,
      "grad_norm": 0.3496308922767639,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.5454,
      "step": 5300
    },
    {
      "epoch": 0.46628029504741836,
      "grad_norm": 0.36380675435066223,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.5359,
      "step": 5310
    },
    {
      "epoch": 0.46715841236389183,
      "grad_norm": 0.320933997631073,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.5204,
      "step": 5320
    },
    {
      "epoch": 0.4680365296803653,
      "grad_norm": 0.3674697279930115,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.5406,
      "step": 5330
    },
    {
      "epoch": 0.46891464699683877,
      "grad_norm": 0.38482779264450073,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5506,
      "step": 5340
    },
    {
      "epoch": 0.46979276431331224,
      "grad_norm": 0.3878147602081299,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.5438,
      "step": 5350
    },
    {
      "epoch": 0.47067088162978576,
      "grad_norm": 0.37005263566970825,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.5596,
      "step": 5360
    },
    {
      "epoch": 0.47154899894625923,
      "grad_norm": 0.4110938012599945,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5858,
      "step": 5370
    },
    {
      "epoch": 0.4724271162627327,
      "grad_norm": 0.3732818365097046,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.5383,
      "step": 5380
    },
    {
      "epoch": 0.4733052335792062,
      "grad_norm": 0.35991671681404114,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.5476,
      "step": 5390
    },
    {
      "epoch": 0.47418335089567965,
      "grad_norm": 0.3809298574924469,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.5531,
      "step": 5400
    },
    {
      "epoch": 0.47506146821215317,
      "grad_norm": 0.367240846157074,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.537,
      "step": 5410
    },
    {
      "epoch": 0.47593958552862664,
      "grad_norm": 0.3301045298576355,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.5186,
      "step": 5420
    },
    {
      "epoch": 0.4768177028451001,
      "grad_norm": 0.3584490418434143,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.539,
      "step": 5430
    },
    {
      "epoch": 0.4776958201615736,
      "grad_norm": 0.3851182460784912,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.543,
      "step": 5440
    },
    {
      "epoch": 0.47857393747804705,
      "grad_norm": 0.34949174523353577,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.5201,
      "step": 5450
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 0.3128294050693512,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.5266,
      "step": 5460
    },
    {
      "epoch": 0.48033017211099405,
      "grad_norm": 0.32458004355430603,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.5222,
      "step": 5470
    },
    {
      "epoch": 0.4812082894274675,
      "grad_norm": 0.3747503459453583,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.541,
      "step": 5480
    },
    {
      "epoch": 0.482086406743941,
      "grad_norm": 0.3244788944721222,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.5643,
      "step": 5490
    },
    {
      "epoch": 0.48296452406041446,
      "grad_norm": 0.3767109215259552,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.5436,
      "step": 5500
    },
    {
      "epoch": 0.4838426413768879,
      "grad_norm": 0.40771064162254333,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.5572,
      "step": 5510
    },
    {
      "epoch": 0.48472075869336145,
      "grad_norm": 0.3545834422111511,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.5216,
      "step": 5520
    },
    {
      "epoch": 0.4855988760098349,
      "grad_norm": 0.334891676902771,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.5498,
      "step": 5530
    },
    {
      "epoch": 0.4864769933263084,
      "grad_norm": 0.3897126019001007,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.5564,
      "step": 5540
    },
    {
      "epoch": 0.48735511064278186,
      "grad_norm": 0.3418819308280945,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.5528,
      "step": 5550
    },
    {
      "epoch": 0.48823322795925533,
      "grad_norm": 0.3467920422554016,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.5349,
      "step": 5560
    },
    {
      "epoch": 0.48911134527572886,
      "grad_norm": 0.33173272013664246,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.5265,
      "step": 5570
    },
    {
      "epoch": 0.48998946259220233,
      "grad_norm": 0.3358282446861267,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.5245,
      "step": 5580
    },
    {
      "epoch": 0.4908675799086758,
      "grad_norm": 0.4318811297416687,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.5411,
      "step": 5590
    },
    {
      "epoch": 0.49148226203020723,
      "step": 5597,
      "total_flos": 2.4985159616803373e+19,
      "train_loss": 0.049488754597906345,
      "train_runtime": 5707.0961,
      "train_samples_per_second": 11.769,
      "train_steps_per_second": 0.981
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4985159616803373e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}