{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9828767123287672,
  "eval_steps": 500,
  "global_step": 11193,
  "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": 2e-05,
      "loss": 0.5237,
      "step": 5090
    },
    {
      "epoch": 0.44783983140147526,
      "grad_norm": 0.4506855309009552,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 5100
    },
    {
      "epoch": 0.44871794871794873,
      "grad_norm": 0.37907153367996216,
      "learning_rate": 2e-05,
      "loss": 0.5664,
      "step": 5110
    },
    {
      "epoch": 0.4495960660344222,
      "grad_norm": 0.37525638937950134,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 5120
    },
    {
      "epoch": 0.4504741833508957,
      "grad_norm": 0.3130715489387512,
      "learning_rate": 2e-05,
      "loss": 0.5587,
      "step": 5130
    },
    {
      "epoch": 0.45135230066736914,
      "grad_norm": 0.3663550317287445,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 5140
    },
    {
      "epoch": 0.4522304179838426,
      "grad_norm": 0.36279919743537903,
      "learning_rate": 2e-05,
      "loss": 0.5782,
      "step": 5150
    },
    {
      "epoch": 0.45310853530031614,
      "grad_norm": 0.34979087114334106,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 5160
    },
    {
      "epoch": 0.4539866526167896,
      "grad_norm": 0.32817918062210083,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 5170
    },
    {
      "epoch": 0.4548647699332631,
      "grad_norm": 0.3272080719470978,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 5180
    },
    {
      "epoch": 0.45574288724973655,
      "grad_norm": 0.3369264602661133,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 5190
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 0.3272131681442261,
      "learning_rate": 2e-05,
      "loss": 0.5787,
      "step": 5200
    },
    {
      "epoch": 0.45749912188268355,
      "grad_norm": 0.35450220108032227,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 5210
    },
    {
      "epoch": 0.458377239199157,
      "grad_norm": 0.3489535450935364,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 5220
    },
    {
      "epoch": 0.4592553565156305,
      "grad_norm": 0.32346659898757935,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 5230
    },
    {
      "epoch": 0.46013347383210396,
      "grad_norm": 0.33930134773254395,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 5240
    },
    {
      "epoch": 0.4610115911485774,
      "grad_norm": 0.36700281500816345,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 5250
    },
    {
      "epoch": 0.46188970846505095,
      "grad_norm": 0.39659515023231506,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 5260
    },
    {
      "epoch": 0.4627678257815244,
      "grad_norm": 0.34309446811676025,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 5270
    },
    {
      "epoch": 0.4636459430979979,
      "grad_norm": 0.38302135467529297,
      "learning_rate": 2e-05,
      "loss": 0.5502,
      "step": 5280
    },
    {
      "epoch": 0.46452406041447136,
      "grad_norm": 0.3356741964817047,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 5290
    },
    {
      "epoch": 0.46540217773094483,
      "grad_norm": 0.33217835426330566,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 5300
    },
    {
      "epoch": 0.46628029504741836,
      "grad_norm": 0.3617849051952362,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 5310
    },
    {
      "epoch": 0.46715841236389183,
      "grad_norm": 0.3203541338443756,
      "learning_rate": 2e-05,
      "loss": 0.5229,
      "step": 5320
    },
    {
      "epoch": 0.4680365296803653,
      "grad_norm": 0.3835137188434601,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 5330
    },
    {
      "epoch": 0.46891464699683877,
      "grad_norm": 0.3798026740550995,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 5340
    },
    {
      "epoch": 0.46979276431331224,
      "grad_norm": 0.38094961643218994,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 5350
    },
    {
      "epoch": 0.47067088162978576,
      "grad_norm": 0.36592814326286316,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 5360
    },
    {
      "epoch": 0.47154899894625923,
      "grad_norm": 0.42350926995277405,
      "learning_rate": 2e-05,
      "loss": 0.5897,
      "step": 5370
    },
    {
      "epoch": 0.4724271162627327,
      "grad_norm": 0.3540163040161133,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 5380
    },
    {
      "epoch": 0.4733052335792062,
      "grad_norm": 0.34340018033981323,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 5390
    },
    {
      "epoch": 0.47418335089567965,
      "grad_norm": 1.1630723476409912,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 5400
    },
    {
      "epoch": 0.47506146821215317,
      "grad_norm": 0.35946592688560486,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 5410
    },
    {
      "epoch": 0.47593958552862664,
      "grad_norm": 0.3362139165401459,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 5420
    },
    {
      "epoch": 0.4768177028451001,
      "grad_norm": 0.3452414572238922,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 5430
    },
    {
      "epoch": 0.4776958201615736,
      "grad_norm": 0.4002755284309387,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 5440
    },
    {
      "epoch": 0.47857393747804705,
      "grad_norm": 0.3435381054878235,
      "learning_rate": 2e-05,
      "loss": 0.5263,
      "step": 5450
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 0.32638534903526306,
      "learning_rate": 2e-05,
      "loss": 0.5321,
      "step": 5460
    },
    {
      "epoch": 0.48033017211099405,
      "grad_norm": 0.33116230368614197,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 5470
    },
    {
      "epoch": 0.4812082894274675,
      "grad_norm": 0.36597350239753723,
      "learning_rate": 2e-05,
      "loss": 0.5462,
      "step": 5480
    },
    {
      "epoch": 0.482086406743941,
      "grad_norm": 0.3249286711215973,
      "learning_rate": 2e-05,
      "loss": 0.5701,
      "step": 5490
    },
    {
      "epoch": 0.48296452406041446,
      "grad_norm": 0.3791310489177704,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 5500
    },
    {
      "epoch": 0.4838426413768879,
      "grad_norm": 0.376712441444397,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 5510
    },
    {
      "epoch": 0.48472075869336145,
      "grad_norm": 0.39217644929885864,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 5520
    },
    {
      "epoch": 0.4855988760098349,
      "grad_norm": 0.3665665090084076,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 5530
    },
    {
      "epoch": 0.4864769933263084,
      "grad_norm": 0.3728981912136078,
      "learning_rate": 2e-05,
      "loss": 0.5617,
      "step": 5540
    },
    {
      "epoch": 0.48735511064278186,
      "grad_norm": 0.3362461030483246,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 5550
    },
    {
      "epoch": 0.48823322795925533,
      "grad_norm": 0.35911497473716736,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 5560
    },
    {
      "epoch": 0.48911134527572886,
      "grad_norm": 0.3467462360858917,
      "learning_rate": 2e-05,
      "loss": 0.5325,
      "step": 5570
    },
    {
      "epoch": 0.48998946259220233,
      "grad_norm": 0.33052322268486023,
      "learning_rate": 2e-05,
      "loss": 0.5304,
      "step": 5580
    },
    {
      "epoch": 0.4908675799086758,
      "grad_norm": 0.3810023367404938,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 5590
    },
    {
      "epoch": 0.49174569722514927,
      "grad_norm": 0.3528052866458893,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 5600
    },
    {
      "epoch": 0.49262381454162274,
      "grad_norm": 0.392827570438385,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 5610
    },
    {
      "epoch": 0.49350193185809627,
      "grad_norm": 0.3751693665981293,
      "learning_rate": 2e-05,
      "loss": 0.5453,
      "step": 5620
    },
    {
      "epoch": 0.49438004917456974,
      "grad_norm": 0.342002809047699,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 5630
    },
    {
      "epoch": 0.4952581664910432,
      "grad_norm": 0.3717895746231079,
      "learning_rate": 2e-05,
      "loss": 0.5713,
      "step": 5640
    },
    {
      "epoch": 0.4961362838075167,
      "grad_norm": 0.35941725969314575,
      "learning_rate": 2e-05,
      "loss": 0.518,
      "step": 5650
    },
    {
      "epoch": 0.49701440112399015,
      "grad_norm": 0.3109247088432312,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 5660
    },
    {
      "epoch": 0.49789251844046367,
      "grad_norm": 0.44551894068717957,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 5670
    },
    {
      "epoch": 0.49877063575693714,
      "grad_norm": 0.4612896144390106,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 5680
    },
    {
      "epoch": 0.4996487530734106,
      "grad_norm": 0.3293391168117523,
      "learning_rate": 2e-05,
      "loss": 0.5447,
      "step": 5690
    },
    {
      "epoch": 0.5005268703898841,
      "grad_norm": 0.33525747060775757,
      "learning_rate": 2e-05,
      "loss": 0.5638,
      "step": 5700
    },
    {
      "epoch": 0.5014049877063576,
      "grad_norm": 0.41552144289016724,
      "learning_rate": 2e-05,
      "loss": 0.5447,
      "step": 5710
    },
    {
      "epoch": 0.502283105022831,
      "grad_norm": 0.4090516269207001,
      "learning_rate": 2e-05,
      "loss": 0.5727,
      "step": 5720
    },
    {
      "epoch": 0.5031612223393045,
      "grad_norm": 0.32900428771972656,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 5730
    },
    {
      "epoch": 0.504039339655778,
      "grad_norm": 0.5638274550437927,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 5740
    },
    {
      "epoch": 0.5049174569722515,
      "grad_norm": 0.40188997983932495,
      "learning_rate": 2e-05,
      "loss": 0.5786,
      "step": 5750
    },
    {
      "epoch": 0.505795574288725,
      "grad_norm": 0.4244239032268524,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 5760
    },
    {
      "epoch": 0.5066736916051985,
      "grad_norm": 0.4956302344799042,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 5770
    },
    {
      "epoch": 0.507551808921672,
      "grad_norm": 0.3654501438140869,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 5780
    },
    {
      "epoch": 0.5084299262381454,
      "grad_norm": 0.34429118037223816,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 5790
    },
    {
      "epoch": 0.5093080435546189,
      "grad_norm": 0.37749597430229187,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 5800
    },
    {
      "epoch": 0.5101861608710924,
      "grad_norm": 0.34172993898391724,
      "learning_rate": 2e-05,
      "loss": 0.5328,
      "step": 5810
    },
    {
      "epoch": 0.5110642781875658,
      "grad_norm": 0.3440622389316559,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 5820
    },
    {
      "epoch": 0.5119423955040393,
      "grad_norm": 0.35116875171661377,
      "learning_rate": 2e-05,
      "loss": 0.5494,
      "step": 5830
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.3572109043598175,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 5840
    },
    {
      "epoch": 0.5136986301369864,
      "grad_norm": 0.414442777633667,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 5850
    },
    {
      "epoch": 0.5145767474534598,
      "grad_norm": 0.3906424045562744,
      "learning_rate": 2e-05,
      "loss": 0.5319,
      "step": 5860
    },
    {
      "epoch": 0.5154548647699333,
      "grad_norm": 0.3614400029182434,
      "learning_rate": 2e-05,
      "loss": 0.5624,
      "step": 5870
    },
    {
      "epoch": 0.5163329820864068,
      "grad_norm": 0.3355138897895813,
      "learning_rate": 2e-05,
      "loss": 0.5405,
      "step": 5880
    },
    {
      "epoch": 0.5172110994028802,
      "grad_norm": 0.36429378390312195,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 5890
    },
    {
      "epoch": 0.5180892167193537,
      "grad_norm": 0.351931631565094,
      "learning_rate": 2e-05,
      "loss": 0.5372,
      "step": 5900
    },
    {
      "epoch": 0.5189673340358272,
      "grad_norm": 0.5441412925720215,
      "learning_rate": 2e-05,
      "loss": 0.5227,
      "step": 5910
    },
    {
      "epoch": 0.5198454513523006,
      "grad_norm": 0.40622633695602417,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 5920
    },
    {
      "epoch": 0.5207235686687741,
      "grad_norm": 0.37943848967552185,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 5930
    },
    {
      "epoch": 0.5216016859852476,
      "grad_norm": 0.3485409617424011,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 5940
    },
    {
      "epoch": 0.5224798033017211,
      "grad_norm": 0.36290377378463745,
      "learning_rate": 2e-05,
      "loss": 0.5295,
      "step": 5950
    },
    {
      "epoch": 0.5233579206181946,
      "grad_norm": 0.34216833114624023,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 5960
    },
    {
      "epoch": 0.5242360379346681,
      "grad_norm": 0.34164223074913025,
      "learning_rate": 2e-05,
      "loss": 0.5422,
      "step": 5970
    },
    {
      "epoch": 0.5251141552511416,
      "grad_norm": 0.3697287142276764,
      "learning_rate": 2e-05,
      "loss": 0.5551,
      "step": 5980
    },
    {
      "epoch": 0.525992272567615,
      "grad_norm": 0.40399232506752014,
      "learning_rate": 2e-05,
      "loss": 0.5544,
      "step": 5990
    },
    {
      "epoch": 0.5268703898840885,
      "grad_norm": 0.3324736952781677,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 6000
    },
    {
      "epoch": 0.527748507200562,
      "grad_norm": 0.338787317276001,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 6010
    },
    {
      "epoch": 0.5286266245170355,
      "grad_norm": 0.35482701659202576,
      "learning_rate": 2e-05,
      "loss": 0.551,
      "step": 6020
    },
    {
      "epoch": 0.5295047418335089,
      "grad_norm": 0.3528098464012146,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 6030
    },
    {
      "epoch": 0.5303828591499824,
      "grad_norm": 0.35488101840019226,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 6040
    },
    {
      "epoch": 0.5312609764664559,
      "grad_norm": 0.34174397587776184,
      "learning_rate": 2e-05,
      "loss": 0.5623,
      "step": 6050
    },
    {
      "epoch": 0.5321390937829295,
      "grad_norm": 0.37701651453971863,
      "learning_rate": 2e-05,
      "loss": 0.5658,
      "step": 6060
    },
    {
      "epoch": 0.5330172110994029,
      "grad_norm": 0.34780290722846985,
      "learning_rate": 2e-05,
      "loss": 0.518,
      "step": 6070
    },
    {
      "epoch": 0.5338953284158764,
      "grad_norm": 0.37627479434013367,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 6080
    },
    {
      "epoch": 0.5347734457323499,
      "grad_norm": 0.4159623086452484,
      "learning_rate": 2e-05,
      "loss": 0.5327,
      "step": 6090
    },
    {
      "epoch": 0.5356515630488233,
      "grad_norm": 0.35154831409454346,
      "learning_rate": 2e-05,
      "loss": 0.5441,
      "step": 6100
    },
    {
      "epoch": 0.5365296803652968,
      "grad_norm": 0.3735860288143158,
      "learning_rate": 2e-05,
      "loss": 0.5399,
      "step": 6110
    },
    {
      "epoch": 0.5374077976817703,
      "grad_norm": 0.38431957364082336,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 6120
    },
    {
      "epoch": 0.5382859149982437,
      "grad_norm": 0.3541143238544464,
      "learning_rate": 2e-05,
      "loss": 0.5437,
      "step": 6130
    },
    {
      "epoch": 0.5391640323147172,
      "grad_norm": 0.36793428659439087,
      "learning_rate": 2e-05,
      "loss": 0.5696,
      "step": 6140
    },
    {
      "epoch": 0.5400421496311907,
      "grad_norm": 0.3840474784374237,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 6150
    },
    {
      "epoch": 0.5409202669476643,
      "grad_norm": 0.3354026675224304,
      "learning_rate": 2e-05,
      "loss": 0.5057,
      "step": 6160
    },
    {
      "epoch": 0.5417983842641377,
      "grad_norm": 0.4004753530025482,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 6170
    },
    {
      "epoch": 0.5426765015806112,
      "grad_norm": 0.37824204564094543,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 6180
    },
    {
      "epoch": 0.5435546188970847,
      "grad_norm": 0.3929276466369629,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 6190
    },
    {
      "epoch": 0.5444327362135581,
      "grad_norm": 0.33877575397491455,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 6200
    },
    {
      "epoch": 0.5453108535300316,
      "grad_norm": 0.35101282596588135,
      "learning_rate": 2e-05,
      "loss": 0.5342,
      "step": 6210
    },
    {
      "epoch": 0.5461889708465051,
      "grad_norm": 0.32535943388938904,
      "learning_rate": 2e-05,
      "loss": 0.5351,
      "step": 6220
    },
    {
      "epoch": 0.5470670881629786,
      "grad_norm": 0.35272470116615295,
      "learning_rate": 2e-05,
      "loss": 0.5545,
      "step": 6230
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 0.42553919553756714,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 6240
    },
    {
      "epoch": 0.5488233227959255,
      "grad_norm": 0.4717937409877777,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 6250
    },
    {
      "epoch": 0.549701440112399,
      "grad_norm": 0.3492211401462555,
      "learning_rate": 2e-05,
      "loss": 0.5376,
      "step": 6260
    },
    {
      "epoch": 0.5505795574288725,
      "grad_norm": 0.6657759547233582,
      "learning_rate": 2e-05,
      "loss": 0.552,
      "step": 6270
    },
    {
      "epoch": 0.551457674745346,
      "grad_norm": 0.3753378391265869,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 6280
    },
    {
      "epoch": 0.5523357920618195,
      "grad_norm": 0.3669988512992859,
      "learning_rate": 2e-05,
      "loss": 0.5438,
      "step": 6290
    },
    {
      "epoch": 0.553213909378293,
      "grad_norm": 0.3673103153705597,
      "learning_rate": 2e-05,
      "loss": 0.5336,
      "step": 6300
    },
    {
      "epoch": 0.5540920266947664,
      "grad_norm": 0.7225094437599182,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 6310
    },
    {
      "epoch": 0.5549701440112399,
      "grad_norm": 0.35891085863113403,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 6320
    },
    {
      "epoch": 0.5558482613277134,
      "grad_norm": 0.3926536738872528,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 6330
    },
    {
      "epoch": 0.5567263786441868,
      "grad_norm": 0.39494484663009644,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 6340
    },
    {
      "epoch": 0.5576044959606603,
      "grad_norm": 0.3972737789154053,
      "learning_rate": 2e-05,
      "loss": 0.5222,
      "step": 6350
    },
    {
      "epoch": 0.5584826132771338,
      "grad_norm": 0.35632285475730896,
      "learning_rate": 2e-05,
      "loss": 0.5249,
      "step": 6360
    },
    {
      "epoch": 0.5593607305936074,
      "grad_norm": 0.6604139804840088,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 6370
    },
    {
      "epoch": 0.5602388479100808,
      "grad_norm": 0.37049490213394165,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 6380
    },
    {
      "epoch": 0.5611169652265543,
      "grad_norm": 0.34619539976119995,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 6390
    },
    {
      "epoch": 0.5619950825430278,
      "grad_norm": 0.3562653362751007,
      "learning_rate": 2e-05,
      "loss": 0.5486,
      "step": 6400
    },
    {
      "epoch": 0.5628731998595012,
      "grad_norm": 0.3382242023944855,
      "learning_rate": 2e-05,
      "loss": 0.542,
      "step": 6410
    },
    {
      "epoch": 0.5637513171759747,
      "grad_norm": 0.3686565160751343,
      "learning_rate": 2e-05,
      "loss": 0.5496,
      "step": 6420
    },
    {
      "epoch": 0.5646294344924482,
      "grad_norm": 0.35276851058006287,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 6430
    },
    {
      "epoch": 0.5655075518089216,
      "grad_norm": 0.363592267036438,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 6440
    },
    {
      "epoch": 0.5663856691253951,
      "grad_norm": 0.3439541161060333,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 6450
    },
    {
      "epoch": 0.5672637864418686,
      "grad_norm": 0.33681929111480713,
      "learning_rate": 2e-05,
      "loss": 0.5333,
      "step": 6460
    },
    {
      "epoch": 0.5681419037583421,
      "grad_norm": 0.33874133229255676,
      "learning_rate": 2e-05,
      "loss": 0.5104,
      "step": 6470
    },
    {
      "epoch": 0.5690200210748156,
      "grad_norm": 0.3788045346736908,
      "learning_rate": 2e-05,
      "loss": 0.5745,
      "step": 6480
    },
    {
      "epoch": 0.5698981383912891,
      "grad_norm": 0.3881615102291107,
      "learning_rate": 2e-05,
      "loss": 0.5345,
      "step": 6490
    },
    {
      "epoch": 0.5707762557077626,
      "grad_norm": 0.3693525493144989,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 6500
    },
    {
      "epoch": 0.571654373024236,
      "grad_norm": 0.3521127998828888,
      "learning_rate": 2e-05,
      "loss": 0.56,
      "step": 6510
    },
    {
      "epoch": 0.5725324903407095,
      "grad_norm": 0.35816690325737,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 6520
    },
    {
      "epoch": 0.573410607657183,
      "grad_norm": 0.43303847312927246,
      "learning_rate": 2e-05,
      "loss": 0.5417,
      "step": 6530
    },
    {
      "epoch": 0.5742887249736565,
      "grad_norm": 0.37847524881362915,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 6540
    },
    {
      "epoch": 0.5751668422901299,
      "grad_norm": 0.36063164472579956,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 6550
    },
    {
      "epoch": 0.5760449596066034,
      "grad_norm": 0.33830511569976807,
      "learning_rate": 2e-05,
      "loss": 0.542,
      "step": 6560
    },
    {
      "epoch": 0.5769230769230769,
      "grad_norm": 0.37057244777679443,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 6570
    },
    {
      "epoch": 0.5778011942395505,
      "grad_norm": 0.3293398916721344,
      "learning_rate": 2e-05,
      "loss": 0.4984,
      "step": 6580
    },
    {
      "epoch": 0.5786793115560239,
      "grad_norm": 0.32207444310188293,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 6590
    },
    {
      "epoch": 0.5795574288724974,
      "grad_norm": 2.189727783203125,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 6600
    },
    {
      "epoch": 0.5804355461889709,
      "grad_norm": 0.4619254469871521,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 6610
    },
    {
      "epoch": 0.5813136635054443,
      "grad_norm": 0.41954097151756287,
      "learning_rate": 2e-05,
      "loss": 0.5588,
      "step": 6620
    },
    {
      "epoch": 0.5821917808219178,
      "grad_norm": 0.40110310912132263,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 6630
    },
    {
      "epoch": 0.5830698981383913,
      "grad_norm": 0.3591701090335846,
      "learning_rate": 2e-05,
      "loss": 0.5051,
      "step": 6640
    },
    {
      "epoch": 0.5839480154548647,
      "grad_norm": 0.35716038942337036,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 6650
    },
    {
      "epoch": 0.5848261327713382,
      "grad_norm": 0.3700909912586212,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 6660
    },
    {
      "epoch": 0.5857042500878117,
      "grad_norm": 0.33744364976882935,
      "learning_rate": 2e-05,
      "loss": 0.5347,
      "step": 6670
    },
    {
      "epoch": 0.5865823674042853,
      "grad_norm": 0.3300972878932953,
      "learning_rate": 2e-05,
      "loss": 0.5369,
      "step": 6680
    },
    {
      "epoch": 0.5874604847207587,
      "grad_norm": 0.3552870750427246,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 6690
    },
    {
      "epoch": 0.5883386020372322,
      "grad_norm": 0.3834725618362427,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 6700
    },
    {
      "epoch": 0.5892167193537057,
      "grad_norm": 0.3839147984981537,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 6710
    },
    {
      "epoch": 0.5900948366701791,
      "grad_norm": 0.59331214427948,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 6720
    },
    {
      "epoch": 0.5909729539866526,
      "grad_norm": 0.3954465389251709,
      "learning_rate": 2e-05,
      "loss": 0.5152,
      "step": 6730
    },
    {
      "epoch": 0.5918510713031261,
      "grad_norm": 0.34026336669921875,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 6740
    },
    {
      "epoch": 0.5927291886195996,
      "grad_norm": 0.30854731798171997,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 6750
    },
    {
      "epoch": 0.593607305936073,
      "grad_norm": 0.40077298879623413,
      "learning_rate": 2e-05,
      "loss": 0.5345,
      "step": 6760
    },
    {
      "epoch": 0.5944854232525465,
      "grad_norm": 0.38599157333374023,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 6770
    },
    {
      "epoch": 0.59536354056902,
      "grad_norm": 0.3399379253387451,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 6780
    },
    {
      "epoch": 0.5962416578854935,
      "grad_norm": 0.4242315888404846,
      "learning_rate": 2e-05,
      "loss": 0.5659,
      "step": 6790
    },
    {
      "epoch": 0.597119775201967,
      "grad_norm": 0.41016197204589844,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 6800
    },
    {
      "epoch": 0.5979978925184405,
      "grad_norm": 0.3675384521484375,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 6810
    },
    {
      "epoch": 0.598876009834914,
      "grad_norm": 0.36674362421035767,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 6820
    },
    {
      "epoch": 0.5997541271513874,
      "grad_norm": 0.36403900384902954,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 6830
    },
    {
      "epoch": 0.6006322444678609,
      "grad_norm": 0.3233964741230011,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 6840
    },
    {
      "epoch": 0.6015103617843344,
      "grad_norm": 0.35906732082366943,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 6850
    },
    {
      "epoch": 0.6023884791008078,
      "grad_norm": 0.44579043984413147,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 6860
    },
    {
      "epoch": 0.6032665964172813,
      "grad_norm": 0.3734249174594879,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 6870
    },
    {
      "epoch": 0.6041447137337548,
      "grad_norm": 0.3817939758300781,
      "learning_rate": 2e-05,
      "loss": 0.5386,
      "step": 6880
    },
    {
      "epoch": 0.6050228310502284,
      "grad_norm": 0.3592665195465088,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 6890
    },
    {
      "epoch": 0.6059009483667018,
      "grad_norm": 0.3357742726802826,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 6900
    },
    {
      "epoch": 0.6067790656831753,
      "grad_norm": 0.3839417099952698,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 6910
    },
    {
      "epoch": 0.6076571829996488,
      "grad_norm": 0.38926008343696594,
      "learning_rate": 2e-05,
      "loss": 0.5447,
      "step": 6920
    },
    {
      "epoch": 0.6085353003161222,
      "grad_norm": 0.35524076223373413,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 6930
    },
    {
      "epoch": 0.6094134176325957,
      "grad_norm": 0.3686339259147644,
      "learning_rate": 2e-05,
      "loss": 0.5394,
      "step": 6940
    },
    {
      "epoch": 0.6102915349490692,
      "grad_norm": 0.35624366998672485,
      "learning_rate": 2e-05,
      "loss": 0.5563,
      "step": 6950
    },
    {
      "epoch": 0.6111696522655427,
      "grad_norm": 0.37592020630836487,
      "learning_rate": 2e-05,
      "loss": 0.5638,
      "step": 6960
    },
    {
      "epoch": 0.6120477695820161,
      "grad_norm": 0.37664344906806946,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 6970
    },
    {
      "epoch": 0.6129258868984896,
      "grad_norm": 0.37481987476348877,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 6980
    },
    {
      "epoch": 0.6138040042149632,
      "grad_norm": 0.3589646518230438,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 6990
    },
    {
      "epoch": 0.6146821215314366,
      "grad_norm": 0.3746212124824524,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 7000
    },
    {
      "epoch": 0.6155602388479101,
      "grad_norm": 0.4300224483013153,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 7010
    },
    {
      "epoch": 0.6164383561643836,
      "grad_norm": 0.361026793718338,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 7020
    },
    {
      "epoch": 0.617316473480857,
      "grad_norm": 0.35648638010025024,
      "learning_rate": 2e-05,
      "loss": 0.5726,
      "step": 7030
    },
    {
      "epoch": 0.6181945907973305,
      "grad_norm": 0.3696216642856598,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 7040
    },
    {
      "epoch": 0.619072708113804,
      "grad_norm": 0.3665052056312561,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 7050
    },
    {
      "epoch": 0.6199508254302775,
      "grad_norm": 0.38564133644104004,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 7060
    },
    {
      "epoch": 0.6208289427467509,
      "grad_norm": 0.38364851474761963,
      "learning_rate": 2e-05,
      "loss": 0.5638,
      "step": 7070
    },
    {
      "epoch": 0.6217070600632244,
      "grad_norm": 0.3789573907852173,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 7080
    },
    {
      "epoch": 0.6225851773796979,
      "grad_norm": 0.350759893655777,
      "learning_rate": 2e-05,
      "loss": 0.5541,
      "step": 7090
    },
    {
      "epoch": 0.6234632946961715,
      "grad_norm": 0.41135862469673157,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 7100
    },
    {
      "epoch": 0.6243414120126449,
      "grad_norm": 0.5514920353889465,
      "learning_rate": 2e-05,
      "loss": 0.5163,
      "step": 7110
    },
    {
      "epoch": 0.6252195293291184,
      "grad_norm": 0.33282777667045593,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 7120
    },
    {
      "epoch": 0.6260976466455919,
      "grad_norm": 0.36902472376823425,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 7130
    },
    {
      "epoch": 0.6269757639620653,
      "grad_norm": 0.385030597448349,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 7140
    },
    {
      "epoch": 0.6278538812785388,
      "grad_norm": 0.37406888604164124,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 7150
    },
    {
      "epoch": 0.6287319985950123,
      "grad_norm": 0.3616451621055603,
      "learning_rate": 2e-05,
      "loss": 0.5447,
      "step": 7160
    },
    {
      "epoch": 0.6296101159114857,
      "grad_norm": 0.38639265298843384,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 7170
    },
    {
      "epoch": 0.6304882332279592,
      "grad_norm": 0.3478398025035858,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 7180
    },
    {
      "epoch": 0.6313663505444327,
      "grad_norm": 0.38517823815345764,
      "learning_rate": 2e-05,
      "loss": 0.5302,
      "step": 7190
    },
    {
      "epoch": 0.6322444678609063,
      "grad_norm": 0.4034563899040222,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 7200
    },
    {
      "epoch": 0.6331225851773797,
      "grad_norm": 0.35593393445014954,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 7210
    },
    {
      "epoch": 0.6340007024938532,
      "grad_norm": 0.3834037780761719,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 7220
    },
    {
      "epoch": 0.6348788198103267,
      "grad_norm": 0.38702449202537537,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 7230
    },
    {
      "epoch": 0.6357569371268001,
      "grad_norm": 0.43966740369796753,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 7240
    },
    {
      "epoch": 0.6366350544432736,
      "grad_norm": 0.375143438577652,
      "learning_rate": 2e-05,
      "loss": 0.5287,
      "step": 7250
    },
    {
      "epoch": 0.6375131717597471,
      "grad_norm": 0.4144490957260132,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 7260
    },
    {
      "epoch": 0.6383912890762206,
      "grad_norm": 0.41663193702697754,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 7270
    },
    {
      "epoch": 0.639269406392694,
      "grad_norm": 0.3983076512813568,
      "learning_rate": 2e-05,
      "loss": 0.5341,
      "step": 7280
    },
    {
      "epoch": 0.6401475237091675,
      "grad_norm": 0.41534072160720825,
      "learning_rate": 2e-05,
      "loss": 0.5517,
      "step": 7290
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 0.42874225974082947,
      "learning_rate": 2e-05,
      "loss": 0.5273,
      "step": 7300
    },
    {
      "epoch": 0.6419037583421145,
      "grad_norm": 0.3712378442287445,
      "learning_rate": 2e-05,
      "loss": 0.5168,
      "step": 7310
    },
    {
      "epoch": 0.642781875658588,
      "grad_norm": 0.38884034752845764,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 7320
    },
    {
      "epoch": 0.6436599929750615,
      "grad_norm": 0.3321040868759155,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 7330
    },
    {
      "epoch": 0.644538110291535,
      "grad_norm": 0.36604633927345276,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 7340
    },
    {
      "epoch": 0.6454162276080084,
      "grad_norm": 0.349410742521286,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 7350
    },
    {
      "epoch": 0.6462943449244819,
      "grad_norm": 0.3755151629447937,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 7360
    },
    {
      "epoch": 0.6471724622409554,
      "grad_norm": 0.38417160511016846,
      "learning_rate": 2e-05,
      "loss": 0.5279,
      "step": 7370
    },
    {
      "epoch": 0.6480505795574288,
      "grad_norm": 0.391490638256073,
      "learning_rate": 2e-05,
      "loss": 0.5445,
      "step": 7380
    },
    {
      "epoch": 0.6489286968739023,
      "grad_norm": 0.387766033411026,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 7390
    },
    {
      "epoch": 0.6498068141903758,
      "grad_norm": 0.3682886064052582,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 7400
    },
    {
      "epoch": 0.6506849315068494,
      "grad_norm": 0.3653990626335144,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 7410
    },
    {
      "epoch": 0.6515630488233228,
      "grad_norm": 0.38434356451034546,
      "learning_rate": 2e-05,
      "loss": 0.5615,
      "step": 7420
    },
    {
      "epoch": 0.6524411661397963,
      "grad_norm": 0.3517230451107025,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 7430
    },
    {
      "epoch": 0.6533192834562698,
      "grad_norm": 0.4123498201370239,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 7440
    },
    {
      "epoch": 0.6541974007727432,
      "grad_norm": 0.35225310921669006,
      "learning_rate": 2e-05,
      "loss": 0.5416,
      "step": 7450
    },
    {
      "epoch": 0.6550755180892167,
      "grad_norm": 0.4089106619358063,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 7460
    },
    {
      "epoch": 0.6559536354056902,
      "grad_norm": 0.35886386036872864,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 7470
    },
    {
      "epoch": 0.6568317527221637,
      "grad_norm": 0.4046400189399719,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 7480
    },
    {
      "epoch": 0.6577098700386371,
      "grad_norm": 0.3327883780002594,
      "learning_rate": 2e-05,
      "loss": 0.5319,
      "step": 7490
    },
    {
      "epoch": 0.6585879873551106,
      "grad_norm": 0.3243897557258606,
      "learning_rate": 2e-05,
      "loss": 0.4992,
      "step": 7500
    },
    {
      "epoch": 0.6594661046715842,
      "grad_norm": 0.32098570466041565,
      "learning_rate": 2e-05,
      "loss": 0.5112,
      "step": 7510
    },
    {
      "epoch": 0.6603442219880576,
      "grad_norm": 0.32692790031433105,
      "learning_rate": 2e-05,
      "loss": 0.5486,
      "step": 7520
    },
    {
      "epoch": 0.6612223393045311,
      "grad_norm": 0.36064329743385315,
      "learning_rate": 2e-05,
      "loss": 0.5237,
      "step": 7530
    },
    {
      "epoch": 0.6621004566210046,
      "grad_norm": 0.3691558241844177,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 7540
    },
    {
      "epoch": 0.662978573937478,
      "grad_norm": 0.37237343192100525,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 7550
    },
    {
      "epoch": 0.6638566912539515,
      "grad_norm": 0.3124850392341614,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 7560
    },
    {
      "epoch": 0.664734808570425,
      "grad_norm": 0.5980959534645081,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 7570
    },
    {
      "epoch": 0.6656129258868985,
      "grad_norm": 0.3454231023788452,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 7580
    },
    {
      "epoch": 0.6664910432033719,
      "grad_norm": 0.339607298374176,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 7590
    },
    {
      "epoch": 0.6673691605198454,
      "grad_norm": 0.3856877088546753,
      "learning_rate": 2e-05,
      "loss": 0.5033,
      "step": 7600
    },
    {
      "epoch": 0.668247277836319,
      "grad_norm": 0.38363566994667053,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 7610
    },
    {
      "epoch": 0.6691253951527925,
      "grad_norm": 0.37652063369750977,
      "learning_rate": 2e-05,
      "loss": 0.5333,
      "step": 7620
    },
    {
      "epoch": 0.6700035124692659,
      "grad_norm": 0.38799041509628296,
      "learning_rate": 2e-05,
      "loss": 0.5797,
      "step": 7630
    },
    {
      "epoch": 0.6708816297857394,
      "grad_norm": 0.3674986660480499,
      "learning_rate": 2e-05,
      "loss": 0.5438,
      "step": 7640
    },
    {
      "epoch": 0.6717597471022129,
      "grad_norm": 0.36374545097351074,
      "learning_rate": 2e-05,
      "loss": 0.5178,
      "step": 7650
    },
    {
      "epoch": 0.6726378644186863,
      "grad_norm": 0.5509585738182068,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 7660
    },
    {
      "epoch": 0.6735159817351598,
      "grad_norm": 0.3675175905227661,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 7670
    },
    {
      "epoch": 0.6743940990516333,
      "grad_norm": 0.3642314672470093,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 7680
    },
    {
      "epoch": 0.6752722163681067,
      "grad_norm": 0.35739898681640625,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 7690
    },
    {
      "epoch": 0.6761503336845802,
      "grad_norm": 0.3372306227684021,
      "learning_rate": 2e-05,
      "loss": 0.5331,
      "step": 7700
    },
    {
      "epoch": 0.6770284510010537,
      "grad_norm": 0.41274043917655945,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 7710
    },
    {
      "epoch": 0.6779065683175273,
      "grad_norm": 0.4053538143634796,
      "learning_rate": 2e-05,
      "loss": 0.5136,
      "step": 7720
    },
    {
      "epoch": 0.6787846856340007,
      "grad_norm": 0.34812358021736145,
      "learning_rate": 2e-05,
      "loss": 0.5568,
      "step": 7730
    },
    {
      "epoch": 0.6796628029504742,
      "grad_norm": 0.4623987078666687,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 7740
    },
    {
      "epoch": 0.6805409202669477,
      "grad_norm": 0.3724430799484253,
      "learning_rate": 2e-05,
      "loss": 0.5089,
      "step": 7750
    },
    {
      "epoch": 0.6814190375834211,
      "grad_norm": 0.3585948050022125,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 7760
    },
    {
      "epoch": 0.6822971548998946,
      "grad_norm": 0.3574715852737427,
      "learning_rate": 2e-05,
      "loss": 0.5698,
      "step": 7770
    },
    {
      "epoch": 0.6831752722163681,
      "grad_norm": 0.3673342764377594,
      "learning_rate": 2e-05,
      "loss": 0.5444,
      "step": 7780
    },
    {
      "epoch": 0.6840533895328416,
      "grad_norm": 0.3601725399494171,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 7790
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 0.3924255073070526,
      "learning_rate": 2e-05,
      "loss": 0.5276,
      "step": 7800
    },
    {
      "epoch": 0.6858096241657885,
      "grad_norm": 0.3959064185619354,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 7810
    },
    {
      "epoch": 0.6866877414822621,
      "grad_norm": 0.36403688788414,
      "learning_rate": 2e-05,
      "loss": 0.5263,
      "step": 7820
    },
    {
      "epoch": 0.6875658587987356,
      "grad_norm": 0.3518686294555664,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 7830
    },
    {
      "epoch": 0.688443976115209,
      "grad_norm": 0.35768091678619385,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 7840
    },
    {
      "epoch": 0.6893220934316825,
      "grad_norm": 0.4977511465549469,
      "learning_rate": 2e-05,
      "loss": 0.5097,
      "step": 7850
    },
    {
      "epoch": 0.690200210748156,
      "grad_norm": 0.3299480974674225,
      "learning_rate": 2e-05,
      "loss": 0.5102,
      "step": 7860
    },
    {
      "epoch": 0.6910783280646294,
      "grad_norm": 0.3348446190357208,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 7870
    },
    {
      "epoch": 0.6919564453811029,
      "grad_norm": 0.40021848678588867,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 7880
    },
    {
      "epoch": 0.6928345626975764,
      "grad_norm": 0.35872453451156616,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 7890
    },
    {
      "epoch": 0.6937126800140498,
      "grad_norm": 0.36417943239212036,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 7900
    },
    {
      "epoch": 0.6945907973305233,
      "grad_norm": 0.4112679958343506,
      "learning_rate": 2e-05,
      "loss": 0.5735,
      "step": 7910
    },
    {
      "epoch": 0.6954689146469969,
      "grad_norm": 0.3450177013874054,
      "learning_rate": 2e-05,
      "loss": 0.5216,
      "step": 7920
    },
    {
      "epoch": 0.6963470319634704,
      "grad_norm": 0.4286362826824188,
      "learning_rate": 2e-05,
      "loss": 0.5179,
      "step": 7930
    },
    {
      "epoch": 0.6972251492799438,
      "grad_norm": 0.39560723304748535,
      "learning_rate": 2e-05,
      "loss": 0.5324,
      "step": 7940
    },
    {
      "epoch": 0.6981032665964173,
      "grad_norm": 0.3739476799964905,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 7950
    },
    {
      "epoch": 0.6989813839128908,
      "grad_norm": 0.3491816818714142,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 7960
    },
    {
      "epoch": 0.6998595012293642,
      "grad_norm": 0.406505286693573,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 7970
    },
    {
      "epoch": 0.7007376185458377,
      "grad_norm": 0.39122891426086426,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 7980
    },
    {
      "epoch": 0.7016157358623112,
      "grad_norm": 0.34855541586875916,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 7990
    },
    {
      "epoch": 0.7024938531787847,
      "grad_norm": 0.42656245827674866,
      "learning_rate": 2e-05,
      "loss": 0.5135,
      "step": 8000
    },
    {
      "epoch": 0.7033719704952581,
      "grad_norm": 0.4182247817516327,
      "learning_rate": 2e-05,
      "loss": 0.4973,
      "step": 8010
    },
    {
      "epoch": 0.7042500878117316,
      "grad_norm": 0.3775343894958496,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 8020
    },
    {
      "epoch": 0.7051282051282052,
      "grad_norm": 0.34143245220184326,
      "learning_rate": 2e-05,
      "loss": 0.521,
      "step": 8030
    },
    {
      "epoch": 0.7060063224446786,
      "grad_norm": 0.8663638234138489,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 8040
    },
    {
      "epoch": 0.7068844397611521,
      "grad_norm": 0.35475587844848633,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 8050
    },
    {
      "epoch": 0.7077625570776256,
      "grad_norm": 0.35852864384651184,
      "learning_rate": 2e-05,
      "loss": 0.5175,
      "step": 8060
    },
    {
      "epoch": 0.7086406743940991,
      "grad_norm": 0.37462738156318665,
      "learning_rate": 2e-05,
      "loss": 0.5661,
      "step": 8070
    },
    {
      "epoch": 0.7095187917105725,
      "grad_norm": 0.4020882844924927,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 8080
    },
    {
      "epoch": 0.710396909027046,
      "grad_norm": 0.37707847356796265,
      "learning_rate": 2e-05,
      "loss": 0.5537,
      "step": 8090
    },
    {
      "epoch": 0.7112750263435195,
      "grad_norm": 0.36018750071525574,
      "learning_rate": 2e-05,
      "loss": 0.5182,
      "step": 8100
    },
    {
      "epoch": 0.7121531436599929,
      "grad_norm": 0.3923760652542114,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 8110
    },
    {
      "epoch": 0.7130312609764664,
      "grad_norm": 0.3870335817337036,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 8120
    },
    {
      "epoch": 0.71390937829294,
      "grad_norm": 0.3400082588195801,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 8130
    },
    {
      "epoch": 0.7147874956094135,
      "grad_norm": 0.3602992594242096,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 8140
    },
    {
      "epoch": 0.7156656129258869,
      "grad_norm": 0.4131246507167816,
      "learning_rate": 2e-05,
      "loss": 0.537,
      "step": 8150
    },
    {
      "epoch": 0.7165437302423604,
      "grad_norm": 0.3459698259830475,
      "learning_rate": 2e-05,
      "loss": 0.5205,
      "step": 8160
    },
    {
      "epoch": 0.7174218475588339,
      "grad_norm": 0.3666693866252899,
      "learning_rate": 2e-05,
      "loss": 0.5422,
      "step": 8170
    },
    {
      "epoch": 0.7182999648753073,
      "grad_norm": 0.3607214689254761,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 8180
    },
    {
      "epoch": 0.7191780821917808,
      "grad_norm": 0.3537225127220154,
      "learning_rate": 2e-05,
      "loss": 0.5351,
      "step": 8190
    },
    {
      "epoch": 0.7200561995082543,
      "grad_norm": 0.38986700773239136,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 8200
    },
    {
      "epoch": 0.7209343168247278,
      "grad_norm": 0.3699057698249817,
      "learning_rate": 2e-05,
      "loss": 0.5342,
      "step": 8210
    },
    {
      "epoch": 0.7218124341412012,
      "grad_norm": 0.36077243089675903,
      "learning_rate": 2e-05,
      "loss": 0.5073,
      "step": 8220
    },
    {
      "epoch": 0.7226905514576747,
      "grad_norm": 0.3669591248035431,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 8230
    },
    {
      "epoch": 0.7235686687741483,
      "grad_norm": 0.4265690743923187,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 8240
    },
    {
      "epoch": 0.7244467860906217,
      "grad_norm": 0.393914133310318,
      "learning_rate": 2e-05,
      "loss": 0.5057,
      "step": 8250
    },
    {
      "epoch": 0.7253249034070952,
      "grad_norm": 0.3423199951648712,
      "learning_rate": 2e-05,
      "loss": 0.5202,
      "step": 8260
    },
    {
      "epoch": 0.7262030207235687,
      "grad_norm": 0.3599262535572052,
      "learning_rate": 2e-05,
      "loss": 0.503,
      "step": 8270
    },
    {
      "epoch": 0.7270811380400422,
      "grad_norm": 0.37274062633514404,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 8280
    },
    {
      "epoch": 0.7279592553565156,
      "grad_norm": 0.3968351185321808,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 8290
    },
    {
      "epoch": 0.7288373726729891,
      "grad_norm": 0.37620899081230164,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 8300
    },
    {
      "epoch": 0.7297154899894626,
      "grad_norm": 0.4297877252101898,
      "learning_rate": 2e-05,
      "loss": 0.5648,
      "step": 8310
    },
    {
      "epoch": 0.730593607305936,
      "grad_norm": 0.41446828842163086,
      "learning_rate": 2e-05,
      "loss": 0.5254,
      "step": 8320
    },
    {
      "epoch": 0.7314717246224095,
      "grad_norm": 0.38081762194633484,
      "learning_rate": 2e-05,
      "loss": 0.5279,
      "step": 8330
    },
    {
      "epoch": 0.7323498419388831,
      "grad_norm": 0.35616862773895264,
      "learning_rate": 2e-05,
      "loss": 0.5509,
      "step": 8340
    },
    {
      "epoch": 0.7332279592553566,
      "grad_norm": 0.34098437428474426,
      "learning_rate": 2e-05,
      "loss": 0.5192,
      "step": 8350
    },
    {
      "epoch": 0.73410607657183,
      "grad_norm": 0.34920236468315125,
      "learning_rate": 2e-05,
      "loss": 0.5425,
      "step": 8360
    },
    {
      "epoch": 0.7349841938883035,
      "grad_norm": 0.35303348302841187,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 8370
    },
    {
      "epoch": 0.735862311204777,
      "grad_norm": 0.3416445255279541,
      "learning_rate": 2e-05,
      "loss": 0.4947,
      "step": 8380
    },
    {
      "epoch": 0.7367404285212504,
      "grad_norm": 0.3269459307193756,
      "learning_rate": 2e-05,
      "loss": 0.4869,
      "step": 8390
    },
    {
      "epoch": 0.7376185458377239,
      "grad_norm": 0.39124423265457153,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 8400
    },
    {
      "epoch": 0.7384966631541974,
      "grad_norm": 0.3930005431175232,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 8410
    },
    {
      "epoch": 0.7393747804706708,
      "grad_norm": 0.3368057310581207,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 8420
    },
    {
      "epoch": 0.7402528977871443,
      "grad_norm": 0.3374296724796295,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 8430
    },
    {
      "epoch": 0.7411310151036179,
      "grad_norm": 0.3659954071044922,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 8440
    },
    {
      "epoch": 0.7420091324200914,
      "grad_norm": 0.35467082262039185,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 8450
    },
    {
      "epoch": 0.7428872497365648,
      "grad_norm": 0.39995360374450684,
      "learning_rate": 2e-05,
      "loss": 0.5422,
      "step": 8460
    },
    {
      "epoch": 0.7437653670530383,
      "grad_norm": 0.3873901963233948,
      "learning_rate": 2e-05,
      "loss": 0.5121,
      "step": 8470
    },
    {
      "epoch": 0.7446434843695118,
      "grad_norm": 0.3225319981575012,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 8480
    },
    {
      "epoch": 0.7455216016859852,
      "grad_norm": 0.36217013001441956,
      "learning_rate": 2e-05,
      "loss": 0.5371,
      "step": 8490
    },
    {
      "epoch": 0.7463997190024587,
      "grad_norm": 0.3635914921760559,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 8500
    },
    {
      "epoch": 0.7472778363189322,
      "grad_norm": 0.37029337882995605,
      "learning_rate": 2e-05,
      "loss": 0.5312,
      "step": 8510
    },
    {
      "epoch": 0.7481559536354057,
      "grad_norm": 0.7079707980155945,
      "learning_rate": 2e-05,
      "loss": 0.5107,
      "step": 8520
    },
    {
      "epoch": 0.7490340709518791,
      "grad_norm": 0.5115028619766235,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 8530
    },
    {
      "epoch": 0.7499121882683526,
      "grad_norm": 0.3803026080131531,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 8540
    },
    {
      "epoch": 0.7507903055848262,
      "grad_norm": 0.3842320740222931,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 8550
    },
    {
      "epoch": 0.7516684229012996,
      "grad_norm": 0.3200344741344452,
      "learning_rate": 2e-05,
      "loss": 0.5157,
      "step": 8560
    },
    {
      "epoch": 0.7525465402177731,
      "grad_norm": 0.33753013610839844,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 8570
    },
    {
      "epoch": 0.7534246575342466,
      "grad_norm": 0.36847206950187683,
      "learning_rate": 2e-05,
      "loss": 0.5518,
      "step": 8580
    },
    {
      "epoch": 0.7543027748507201,
      "grad_norm": 0.38106676936149597,
      "learning_rate": 2e-05,
      "loss": 0.5328,
      "step": 8590
    },
    {
      "epoch": 0.7551808921671935,
      "grad_norm": 0.40302711725234985,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 8600
    },
    {
      "epoch": 0.756059009483667,
      "grad_norm": 0.36126527190208435,
      "learning_rate": 2e-05,
      "loss": 0.4959,
      "step": 8610
    },
    {
      "epoch": 0.7569371268001405,
      "grad_norm": 0.35232260823249817,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 8620
    },
    {
      "epoch": 0.7578152441166139,
      "grad_norm": 1.0423580408096313,
      "learning_rate": 2e-05,
      "loss": 0.534,
      "step": 8630
    },
    {
      "epoch": 0.7586933614330874,
      "grad_norm": 0.35034653544425964,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 8640
    },
    {
      "epoch": 0.759571478749561,
      "grad_norm": 0.4030856192111969,
      "learning_rate": 2e-05,
      "loss": 0.5594,
      "step": 8650
    },
    {
      "epoch": 0.7604495960660345,
      "grad_norm": 0.39116695523262024,
      "learning_rate": 2e-05,
      "loss": 0.5612,
      "step": 8660
    },
    {
      "epoch": 0.7613277133825079,
      "grad_norm": 0.3975934684276581,
      "learning_rate": 2e-05,
      "loss": 0.5318,
      "step": 8670
    },
    {
      "epoch": 0.7622058306989814,
      "grad_norm": 0.3384407162666321,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 8680
    },
    {
      "epoch": 0.7630839480154549,
      "grad_norm": 0.37688133120536804,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 8690
    },
    {
      "epoch": 0.7639620653319283,
      "grad_norm": 0.36061954498291016,
      "learning_rate": 2e-05,
      "loss": 0.5137,
      "step": 8700
    },
    {
      "epoch": 0.7648401826484018,
      "grad_norm": 0.3372800946235657,
      "learning_rate": 2e-05,
      "loss": 0.5085,
      "step": 8710
    },
    {
      "epoch": 0.7657182999648753,
      "grad_norm": 0.42839062213897705,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 8720
    },
    {
      "epoch": 0.7665964172813488,
      "grad_norm": 0.7200695276260376,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 8730
    },
    {
      "epoch": 0.7674745345978222,
      "grad_norm": 0.35950419306755066,
      "learning_rate": 2e-05,
      "loss": 0.5298,
      "step": 8740
    },
    {
      "epoch": 0.7683526519142958,
      "grad_norm": 0.38468137383461,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 8750
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.3701854348182678,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 8760
    },
    {
      "epoch": 0.7701088865472427,
      "grad_norm": 0.41889986395835876,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 8770
    },
    {
      "epoch": 0.7709870038637162,
      "grad_norm": 0.35237425565719604,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 8780
    },
    {
      "epoch": 0.7718651211801897,
      "grad_norm": 0.3622305989265442,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 8790
    },
    {
      "epoch": 0.7727432384966632,
      "grad_norm": 0.45641666650772095,
      "learning_rate": 2e-05,
      "loss": 0.4918,
      "step": 8800
    },
    {
      "epoch": 0.7736213558131366,
      "grad_norm": 0.3457285463809967,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 8810
    },
    {
      "epoch": 0.7744994731296101,
      "grad_norm": 0.377014696598053,
      "learning_rate": 2e-05,
      "loss": 0.5314,
      "step": 8820
    },
    {
      "epoch": 0.7753775904460836,
      "grad_norm": 0.5919713973999023,
      "learning_rate": 2e-05,
      "loss": 0.5149,
      "step": 8830
    },
    {
      "epoch": 0.776255707762557,
      "grad_norm": 0.3160865604877472,
      "learning_rate": 2e-05,
      "loss": 0.5132,
      "step": 8840
    },
    {
      "epoch": 0.7771338250790305,
      "grad_norm": 0.3805709481239319,
      "learning_rate": 2e-05,
      "loss": 0.5399,
      "step": 8850
    },
    {
      "epoch": 0.7780119423955041,
      "grad_norm": 0.367819219827652,
      "learning_rate": 2e-05,
      "loss": 0.5248,
      "step": 8860
    },
    {
      "epoch": 0.7788900597119776,
      "grad_norm": 0.3556886911392212,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 8870
    },
    {
      "epoch": 0.779768177028451,
      "grad_norm": 0.3660673201084137,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 8880
    },
    {
      "epoch": 0.7806462943449245,
      "grad_norm": 0.3697036802768707,
      "learning_rate": 2e-05,
      "loss": 0.5163,
      "step": 8890
    },
    {
      "epoch": 0.781524411661398,
      "grad_norm": 0.36866995692253113,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 8900
    },
    {
      "epoch": 0.7824025289778714,
      "grad_norm": 0.3309222459793091,
      "learning_rate": 2e-05,
      "loss": 0.5219,
      "step": 8910
    },
    {
      "epoch": 0.7832806462943449,
      "grad_norm": 0.3505431115627289,
      "learning_rate": 2e-05,
      "loss": 0.5369,
      "step": 8920
    },
    {
      "epoch": 0.7841587636108184,
      "grad_norm": 0.3914420008659363,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 8930
    },
    {
      "epoch": 0.7850368809272918,
      "grad_norm": 0.426370769739151,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 8940
    },
    {
      "epoch": 0.7859149982437653,
      "grad_norm": 0.4388853907585144,
      "learning_rate": 2e-05,
      "loss": 0.4916,
      "step": 8950
    },
    {
      "epoch": 0.7867931155602389,
      "grad_norm": 0.40870317816734314,
      "learning_rate": 2e-05,
      "loss": 0.5178,
      "step": 8960
    },
    {
      "epoch": 0.7876712328767124,
      "grad_norm": 0.3455701172351837,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 8970
    },
    {
      "epoch": 0.7885493501931858,
      "grad_norm": 0.3844262659549713,
      "learning_rate": 2e-05,
      "loss": 0.5429,
      "step": 8980
    },
    {
      "epoch": 0.7894274675096593,
      "grad_norm": 0.41728633642196655,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 8990
    },
    {
      "epoch": 0.7903055848261328,
      "grad_norm": 0.3692452311515808,
      "learning_rate": 2e-05,
      "loss": 0.5041,
      "step": 9000
    },
    {
      "epoch": 0.7911837021426062,
      "grad_norm": 0.41731610894203186,
      "learning_rate": 2e-05,
      "loss": 0.5719,
      "step": 9010
    },
    {
      "epoch": 0.7920618194590797,
      "grad_norm": 0.36330726742744446,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 9020
    },
    {
      "epoch": 0.7929399367755532,
      "grad_norm": 0.34662947058677673,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 9030
    },
    {
      "epoch": 0.7938180540920267,
      "grad_norm": 0.3440372049808502,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 9040
    },
    {
      "epoch": 0.7946961714085001,
      "grad_norm": 0.43932822346687317,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 9050
    },
    {
      "epoch": 0.7955742887249737,
      "grad_norm": 0.3770134747028351,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 9060
    },
    {
      "epoch": 0.7964524060414472,
      "grad_norm": 0.4530024528503418,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 9070
    },
    {
      "epoch": 0.7973305233579207,
      "grad_norm": 0.43836209177970886,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 9080
    },
    {
      "epoch": 0.7982086406743941,
      "grad_norm": 0.3687434196472168,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 9090
    },
    {
      "epoch": 0.7990867579908676,
      "grad_norm": 0.37632307410240173,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 9100
    },
    {
      "epoch": 0.7999648753073411,
      "grad_norm": 0.457126647233963,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 9110
    },
    {
      "epoch": 0.8008429926238145,
      "grad_norm": 0.3946188986301422,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 9120
    },
    {
      "epoch": 0.801721109940288,
      "grad_norm": 0.376436710357666,
      "learning_rate": 2e-05,
      "loss": 0.55,
      "step": 9130
    },
    {
      "epoch": 0.8025992272567615,
      "grad_norm": 0.39046794176101685,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 9140
    },
    {
      "epoch": 0.8034773445732349,
      "grad_norm": 0.3780176341533661,
      "learning_rate": 2e-05,
      "loss": 0.5489,
      "step": 9150
    },
    {
      "epoch": 0.8043554618897084,
      "grad_norm": 0.41745725274086,
      "learning_rate": 2e-05,
      "loss": 0.5032,
      "step": 9160
    },
    {
      "epoch": 0.805233579206182,
      "grad_norm": 0.3657802641391754,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 9170
    },
    {
      "epoch": 0.8061116965226555,
      "grad_norm": 0.3446328341960907,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 9180
    },
    {
      "epoch": 0.8069898138391289,
      "grad_norm": 0.3532489538192749,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 9190
    },
    {
      "epoch": 0.8078679311556024,
      "grad_norm": 0.35104814171791077,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 9200
    },
    {
      "epoch": 0.8087460484720759,
      "grad_norm": 0.3849884569644928,
      "learning_rate": 2e-05,
      "loss": 0.5529,
      "step": 9210
    },
    {
      "epoch": 0.8096241657885493,
      "grad_norm": 0.3549913763999939,
      "learning_rate": 2e-05,
      "loss": 0.5329,
      "step": 9220
    },
    {
      "epoch": 0.8105022831050228,
      "grad_norm": 0.35253211855888367,
      "learning_rate": 2e-05,
      "loss": 0.5372,
      "step": 9230
    },
    {
      "epoch": 0.8113804004214963,
      "grad_norm": 0.4000281095504761,
      "learning_rate": 2e-05,
      "loss": 0.5235,
      "step": 9240
    },
    {
      "epoch": 0.8122585177379698,
      "grad_norm": 0.4137227535247803,
      "learning_rate": 2e-05,
      "loss": 0.5195,
      "step": 9250
    },
    {
      "epoch": 0.8131366350544432,
      "grad_norm": 0.3557901978492737,
      "learning_rate": 2e-05,
      "loss": 0.5317,
      "step": 9260
    },
    {
      "epoch": 0.8140147523709168,
      "grad_norm": 0.3564394414424896,
      "learning_rate": 2e-05,
      "loss": 0.5117,
      "step": 9270
    },
    {
      "epoch": 0.8148928696873903,
      "grad_norm": 0.4149726629257202,
      "learning_rate": 2e-05,
      "loss": 0.4963,
      "step": 9280
    },
    {
      "epoch": 0.8157709870038637,
      "grad_norm": 0.38313156366348267,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 9290
    },
    {
      "epoch": 0.8166491043203372,
      "grad_norm": 0.3937954008579254,
      "learning_rate": 2e-05,
      "loss": 0.5209,
      "step": 9300
    },
    {
      "epoch": 0.8175272216368107,
      "grad_norm": 0.46990418434143066,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 9310
    },
    {
      "epoch": 0.8184053389532842,
      "grad_norm": 0.35929715633392334,
      "learning_rate": 2e-05,
      "loss": 0.4997,
      "step": 9320
    },
    {
      "epoch": 0.8192834562697576,
      "grad_norm": 0.3432956635951996,
      "learning_rate": 2e-05,
      "loss": 0.5235,
      "step": 9330
    },
    {
      "epoch": 0.8201615735862311,
      "grad_norm": 0.3798321783542633,
      "learning_rate": 2e-05,
      "loss": 0.5202,
      "step": 9340
    },
    {
      "epoch": 0.8210396909027046,
      "grad_norm": 0.39609459042549133,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 9350
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 0.3473357558250427,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 9360
    },
    {
      "epoch": 0.8227959255356516,
      "grad_norm": 0.42205357551574707,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 9370
    },
    {
      "epoch": 0.8236740428521251,
      "grad_norm": 0.3444143533706665,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 9380
    },
    {
      "epoch": 0.8245521601685986,
      "grad_norm": 0.4150013029575348,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 9390
    },
    {
      "epoch": 0.825430277485072,
      "grad_norm": 0.4412094056606293,
      "learning_rate": 2e-05,
      "loss": 0.5167,
      "step": 9400
    },
    {
      "epoch": 0.8263083948015455,
      "grad_norm": 0.3375869691371918,
      "learning_rate": 2e-05,
      "loss": 0.512,
      "step": 9410
    },
    {
      "epoch": 0.827186512118019,
      "grad_norm": 0.8202062845230103,
      "learning_rate": 2e-05,
      "loss": 0.5134,
      "step": 9420
    },
    {
      "epoch": 0.8280646294344924,
      "grad_norm": 0.39543843269348145,
      "learning_rate": 2e-05,
      "loss": 0.5109,
      "step": 9430
    },
    {
      "epoch": 0.8289427467509659,
      "grad_norm": 0.4139493405818939,
      "learning_rate": 2e-05,
      "loss": 0.51,
      "step": 9440
    },
    {
      "epoch": 0.8298208640674394,
      "grad_norm": 0.40602636337280273,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 9450
    },
    {
      "epoch": 0.8306989813839128,
      "grad_norm": 0.4320943355560303,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 9460
    },
    {
      "epoch": 0.8315770987003863,
      "grad_norm": 0.37309151887893677,
      "learning_rate": 2e-05,
      "loss": 0.4971,
      "step": 9470
    },
    {
      "epoch": 0.8324552160168599,
      "grad_norm": 0.3717033565044403,
      "learning_rate": 2e-05,
      "loss": 0.5394,
      "step": 9480
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4066571295261383,
      "learning_rate": 2e-05,
      "loss": 0.5201,
      "step": 9490
    },
    {
      "epoch": 0.8342114506498068,
      "grad_norm": 0.3845977485179901,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 9500
    },
    {
      "epoch": 0.8350895679662803,
      "grad_norm": 0.40641430020332336,
      "learning_rate": 2e-05,
      "loss": 0.5398,
      "step": 9510
    },
    {
      "epoch": 0.8359676852827538,
      "grad_norm": 0.5320346355438232,
      "learning_rate": 2e-05,
      "loss": 0.5086,
      "step": 9520
    },
    {
      "epoch": 0.8368458025992273,
      "grad_norm": 0.4034406542778015,
      "learning_rate": 2e-05,
      "loss": 0.5119,
      "step": 9530
    },
    {
      "epoch": 0.8377239199157007,
      "grad_norm": 0.38039225339889526,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 9540
    },
    {
      "epoch": 0.8386020372321742,
      "grad_norm": 0.42131057381629944,
      "learning_rate": 2e-05,
      "loss": 0.4817,
      "step": 9550
    },
    {
      "epoch": 0.8394801545486477,
      "grad_norm": 0.40473049879074097,
      "learning_rate": 2e-05,
      "loss": 0.5358,
      "step": 9560
    },
    {
      "epoch": 0.8403582718651211,
      "grad_norm": 0.35858047008514404,
      "learning_rate": 2e-05,
      "loss": 0.533,
      "step": 9570
    },
    {
      "epoch": 0.8412363891815947,
      "grad_norm": 0.42400094866752625,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 9580
    },
    {
      "epoch": 0.8421145064980682,
      "grad_norm": 0.3788333237171173,
      "learning_rate": 2e-05,
      "loss": 0.5448,
      "step": 9590
    },
    {
      "epoch": 0.8429926238145417,
      "grad_norm": 0.3528345227241516,
      "learning_rate": 2e-05,
      "loss": 0.503,
      "step": 9600
    },
    {
      "epoch": 0.8438707411310151,
      "grad_norm": 0.39206498861312866,
      "learning_rate": 2e-05,
      "loss": 0.4944,
      "step": 9610
    },
    {
      "epoch": 0.8447488584474886,
      "grad_norm": 0.37929803133010864,
      "learning_rate": 2e-05,
      "loss": 0.5236,
      "step": 9620
    },
    {
      "epoch": 0.8456269757639621,
      "grad_norm": 0.7173373103141785,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 9630
    },
    {
      "epoch": 0.8465050930804355,
      "grad_norm": 0.4085571765899658,
      "learning_rate": 2e-05,
      "loss": 0.5037,
      "step": 9640
    },
    {
      "epoch": 0.847383210396909,
      "grad_norm": 0.3408905565738678,
      "learning_rate": 2e-05,
      "loss": 0.5321,
      "step": 9650
    },
    {
      "epoch": 0.8482613277133825,
      "grad_norm": 0.4172150492668152,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 9660
    },
    {
      "epoch": 0.8491394450298559,
      "grad_norm": 0.452698290348053,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 9670
    },
    {
      "epoch": 0.8500175623463295,
      "grad_norm": 0.3671005666255951,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 9680
    },
    {
      "epoch": 0.850895679662803,
      "grad_norm": 0.3805868625640869,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 9690
    },
    {
      "epoch": 0.8517737969792765,
      "grad_norm": 0.3417166471481323,
      "learning_rate": 2e-05,
      "loss": 0.5334,
      "step": 9700
    },
    {
      "epoch": 0.8526519142957499,
      "grad_norm": 0.38528764247894287,
      "learning_rate": 2e-05,
      "loss": 0.5272,
      "step": 9710
    },
    {
      "epoch": 0.8535300316122234,
      "grad_norm": 0.3779175877571106,
      "learning_rate": 2e-05,
      "loss": 0.5249,
      "step": 9720
    },
    {
      "epoch": 0.8544081489286969,
      "grad_norm": 0.3780665695667267,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 9730
    },
    {
      "epoch": 0.8552862662451703,
      "grad_norm": 0.3794831931591034,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 9740
    },
    {
      "epoch": 0.8561643835616438,
      "grad_norm": 0.3387886583805084,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 9750
    },
    {
      "epoch": 0.8570425008781173,
      "grad_norm": 0.44111213088035583,
      "learning_rate": 2e-05,
      "loss": 0.5186,
      "step": 9760
    },
    {
      "epoch": 0.8579206181945908,
      "grad_norm": 0.36196234822273254,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 9770
    },
    {
      "epoch": 0.8587987355110642,
      "grad_norm": 0.39516180753707886,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 9780
    },
    {
      "epoch": 0.8596768528275378,
      "grad_norm": 0.43974196910858154,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 9790
    },
    {
      "epoch": 0.8605549701440113,
      "grad_norm": 0.3747071325778961,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 9800
    },
    {
      "epoch": 0.8614330874604847,
      "grad_norm": 0.3854893147945404,
      "learning_rate": 2e-05,
      "loss": 0.5201,
      "step": 9810
    },
    {
      "epoch": 0.8623112047769582,
      "grad_norm": 0.36453890800476074,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 9820
    },
    {
      "epoch": 0.8631893220934317,
      "grad_norm": 0.38408753275871277,
      "learning_rate": 2e-05,
      "loss": 0.5158,
      "step": 9830
    },
    {
      "epoch": 0.8640674394099052,
      "grad_norm": 0.38889285922050476,
      "learning_rate": 2e-05,
      "loss": 0.5357,
      "step": 9840
    },
    {
      "epoch": 0.8649455567263786,
      "grad_norm": 0.36039865016937256,
      "learning_rate": 2e-05,
      "loss": 0.4874,
      "step": 9850
    },
    {
      "epoch": 0.8658236740428521,
      "grad_norm": 0.38093459606170654,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 9860
    },
    {
      "epoch": 0.8667017913593256,
      "grad_norm": 0.3656720817089081,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 9870
    },
    {
      "epoch": 0.867579908675799,
      "grad_norm": 0.3956576883792877,
      "learning_rate": 2e-05,
      "loss": 0.5203,
      "step": 9880
    },
    {
      "epoch": 0.8684580259922726,
      "grad_norm": 0.3962959051132202,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 9890
    },
    {
      "epoch": 0.8693361433087461,
      "grad_norm": 0.3339545726776123,
      "learning_rate": 2e-05,
      "loss": 0.5172,
      "step": 9900
    },
    {
      "epoch": 0.8702142606252196,
      "grad_norm": 0.3764873147010803,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 9910
    },
    {
      "epoch": 0.871092377941693,
      "grad_norm": 0.39194461703300476,
      "learning_rate": 2e-05,
      "loss": 0.5335,
      "step": 9920
    },
    {
      "epoch": 0.8719704952581665,
      "grad_norm": 0.4330295920372009,
      "learning_rate": 2e-05,
      "loss": 0.5226,
      "step": 9930
    },
    {
      "epoch": 0.87284861257464,
      "grad_norm": 0.3872968554496765,
      "learning_rate": 2e-05,
      "loss": 0.5326,
      "step": 9940
    },
    {
      "epoch": 0.8737267298911134,
      "grad_norm": 0.35313788056373596,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 9950
    },
    {
      "epoch": 0.8746048472075869,
      "grad_norm": 0.4037555158138275,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 9960
    },
    {
      "epoch": 0.8754829645240604,
      "grad_norm": 0.37577012181282043,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 9970
    },
    {
      "epoch": 0.8763610818405339,
      "grad_norm": 0.41566163301467896,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 9980
    },
    {
      "epoch": 0.8772391991570073,
      "grad_norm": 0.35795408487319946,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 9990
    },
    {
      "epoch": 0.8781173164734809,
      "grad_norm": 0.3899826109409332,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 10000
    },
    {
      "epoch": 0.8789954337899544,
      "grad_norm": 0.3735426664352417,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 10010
    },
    {
      "epoch": 0.8798735511064278,
      "grad_norm": 0.41033244132995605,
      "learning_rate": 2e-05,
      "loss": 0.5248,
      "step": 10020
    },
    {
      "epoch": 0.8807516684229013,
      "grad_norm": 0.48665952682495117,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 10030
    },
    {
      "epoch": 0.8816297857393748,
      "grad_norm": 0.36845892667770386,
      "learning_rate": 2e-05,
      "loss": 0.5284,
      "step": 10040
    },
    {
      "epoch": 0.8825079030558483,
      "grad_norm": 0.37565088272094727,
      "learning_rate": 2e-05,
      "loss": 0.5224,
      "step": 10050
    },
    {
      "epoch": 0.8833860203723217,
      "grad_norm": 0.3514876961708069,
      "learning_rate": 2e-05,
      "loss": 0.5358,
      "step": 10060
    },
    {
      "epoch": 0.8842641376887952,
      "grad_norm": 0.4170683026313782,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 10070
    },
    {
      "epoch": 0.8851422550052687,
      "grad_norm": 0.40892186760902405,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 10080
    },
    {
      "epoch": 0.8860203723217421,
      "grad_norm": 0.386258065700531,
      "learning_rate": 2e-05,
      "loss": 0.5235,
      "step": 10090
    },
    {
      "epoch": 0.8868984896382157,
      "grad_norm": 0.42378196120262146,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 10100
    },
    {
      "epoch": 0.8877766069546892,
      "grad_norm": 0.4582432508468628,
      "learning_rate": 2e-05,
      "loss": 0.5341,
      "step": 10110
    },
    {
      "epoch": 0.8886547242711627,
      "grad_norm": 0.3912559151649475,
      "learning_rate": 2e-05,
      "loss": 0.5161,
      "step": 10120
    },
    {
      "epoch": 0.8895328415876361,
      "grad_norm": 0.42153626680374146,
      "learning_rate": 2e-05,
      "loss": 0.5322,
      "step": 10130
    },
    {
      "epoch": 0.8904109589041096,
      "grad_norm": 0.45925837755203247,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 10140
    },
    {
      "epoch": 0.8912890762205831,
      "grad_norm": 0.34821271896362305,
      "learning_rate": 2e-05,
      "loss": 0.521,
      "step": 10150
    },
    {
      "epoch": 0.8921671935370565,
      "grad_norm": 0.396720290184021,
      "learning_rate": 2e-05,
      "loss": 0.518,
      "step": 10160
    },
    {
      "epoch": 0.89304531085353,
      "grad_norm": 0.42553746700286865,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 10170
    },
    {
      "epoch": 0.8939234281700035,
      "grad_norm": 0.3408048450946808,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.4801,
      "step": 10180
    },
    {
      "epoch": 0.894801545486477,
      "grad_norm": 0.4914748966693878,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.4915,
      "step": 10190
    },
    {
      "epoch": 0.8956796628029505,
      "grad_norm": 0.3801792562007904,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.5229,
      "step": 10200
    },
    {
      "epoch": 0.896557780119424,
      "grad_norm": 0.33922654390335083,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.5382,
      "step": 10210
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 0.3711055815219879,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.5369,
      "step": 10220
    },
    {
      "epoch": 0.8983140147523709,
      "grad_norm": 0.39101558923721313,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.5171,
      "step": 10230
    },
    {
      "epoch": 0.8991921320688444,
      "grad_norm": 0.6736558079719543,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.5147,
      "step": 10240
    },
    {
      "epoch": 0.9000702493853179,
      "grad_norm": 0.381369948387146,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.523,
      "step": 10250
    },
    {
      "epoch": 0.9009483667017913,
      "grad_norm": 0.37278738617897034,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.5066,
      "step": 10260
    },
    {
      "epoch": 0.9018264840182648,
      "grad_norm": 0.44853389263153076,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.5472,
      "step": 10270
    },
    {
      "epoch": 0.9027046013347383,
      "grad_norm": 0.39217785000801086,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.5333,
      "step": 10280
    },
    {
      "epoch": 0.9035827186512118,
      "grad_norm": 0.3403521478176117,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.5386,
      "step": 10290
    },
    {
      "epoch": 0.9044608359676852,
      "grad_norm": 0.35794126987457275,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.5061,
      "step": 10300
    },
    {
      "epoch": 0.9053389532841588,
      "grad_norm": 0.36915165185928345,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5375,
      "step": 10310
    },
    {
      "epoch": 0.9062170706006323,
      "grad_norm": 0.3945119380950928,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.5312,
      "step": 10320
    },
    {
      "epoch": 0.9070951879171057,
      "grad_norm": 0.33585160970687866,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.5246,
      "step": 10330
    },
    {
      "epoch": 0.9079733052335792,
      "grad_norm": 0.3573198914527893,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.5077,
      "step": 10340
    },
    {
      "epoch": 0.9088514225500527,
      "grad_norm": 0.41619598865509033,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.5386,
      "step": 10350
    },
    {
      "epoch": 0.9097295398665262,
      "grad_norm": 0.33636951446533203,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.5181,
      "step": 10360
    },
    {
      "epoch": 0.9106076571829996,
      "grad_norm": 0.42571765184402466,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.5304,
      "step": 10370
    },
    {
      "epoch": 0.9114857744994731,
      "grad_norm": 0.36110204458236694,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.4949,
      "step": 10380
    },
    {
      "epoch": 0.9123638918159466,
      "grad_norm": 0.37305185198783875,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.5118,
      "step": 10390
    },
    {
      "epoch": 0.91324200913242,
      "grad_norm": 0.5075505375862122,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.5031,
      "step": 10400
    },
    {
      "epoch": 0.9141201264488936,
      "grad_norm": 0.4220711588859558,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.5186,
      "step": 10410
    },
    {
      "epoch": 0.9149982437653671,
      "grad_norm": 0.4159349501132965,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.5483,
      "step": 10420
    },
    {
      "epoch": 0.9158763610818406,
      "grad_norm": 0.39402544498443604,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.5201,
      "step": 10430
    },
    {
      "epoch": 0.916754478398314,
      "grad_norm": 0.3722699284553528,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.5141,
      "step": 10440
    },
    {
      "epoch": 0.9176325957147875,
      "grad_norm": 0.43839120864868164,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.5462,
      "step": 10450
    },
    {
      "epoch": 0.918510713031261,
      "grad_norm": 0.3853551149368286,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.5353,
      "step": 10460
    },
    {
      "epoch": 0.9193888303477344,
      "grad_norm": 0.40310370922088623,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.5049,
      "step": 10470
    },
    {
      "epoch": 0.9202669476642079,
      "grad_norm": 0.3759850561618805,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.5052,
      "step": 10480
    },
    {
      "epoch": 0.9211450649806814,
      "grad_norm": 0.3700179159641266,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.5018,
      "step": 10490
    },
    {
      "epoch": 0.9220231822971549,
      "grad_norm": 0.3945019841194153,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.5364,
      "step": 10500
    },
    {
      "epoch": 0.9229012996136284,
      "grad_norm": 0.37300020456314087,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5097,
      "step": 10510
    },
    {
      "epoch": 0.9237794169301019,
      "grad_norm": 0.40198683738708496,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.5202,
      "step": 10520
    },
    {
      "epoch": 0.9246575342465754,
      "grad_norm": 0.3739705979824066,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.5042,
      "step": 10530
    },
    {
      "epoch": 0.9255356515630488,
      "grad_norm": 0.3539210557937622,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.4933,
      "step": 10540
    },
    {
      "epoch": 0.9264137688795223,
      "grad_norm": 0.40182894468307495,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.4873,
      "step": 10550
    },
    {
      "epoch": 0.9272918861959958,
      "grad_norm": 0.39215341210365295,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.5264,
      "step": 10560
    },
    {
      "epoch": 0.9281700035124693,
      "grad_norm": 0.4018980860710144,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.5197,
      "step": 10570
    },
    {
      "epoch": 0.9290481208289427,
      "grad_norm": 0.3528560996055603,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.5329,
      "step": 10580
    },
    {
      "epoch": 0.9299262381454162,
      "grad_norm": 0.372596800327301,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.5361,
      "step": 10590
    },
    {
      "epoch": 0.9308043554618897,
      "grad_norm": 0.361334890127182,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.5152,
      "step": 10600
    },
    {
      "epoch": 0.9316824727783631,
      "grad_norm": 0.41073355078697205,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.492,
      "step": 10610
    },
    {
      "epoch": 0.9325605900948367,
      "grad_norm": 0.3389752209186554,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.5027,
      "step": 10620
    },
    {
      "epoch": 0.9334387074113102,
      "grad_norm": 0.37932464480400085,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.5179,
      "step": 10630
    },
    {
      "epoch": 0.9343168247277837,
      "grad_norm": 0.35319021344184875,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.5006,
      "step": 10640
    },
    {
      "epoch": 0.9351949420442571,
      "grad_norm": 0.35317912697792053,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.4948,
      "step": 10650
    },
    {
      "epoch": 0.9360730593607306,
      "grad_norm": 0.3635043203830719,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.5182,
      "step": 10660
    },
    {
      "epoch": 0.9369511766772041,
      "grad_norm": 0.37823760509490967,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.519,
      "step": 10670
    },
    {
      "epoch": 0.9378292939936775,
      "grad_norm": 0.38922634720802307,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.4973,
      "step": 10680
    },
    {
      "epoch": 0.938707411310151,
      "grad_norm": 0.34728217124938965,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.5104,
      "step": 10690
    },
    {
      "epoch": 0.9395855286266245,
      "grad_norm": 0.43056586384773254,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.5169,
      "step": 10700
    },
    {
      "epoch": 0.940463645943098,
      "grad_norm": 0.38191765546798706,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.4998,
      "step": 10710
    },
    {
      "epoch": 0.9413417632595715,
      "grad_norm": 0.36074593663215637,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.5177,
      "step": 10720
    },
    {
      "epoch": 0.942219880576045,
      "grad_norm": 0.36359384655952454,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.5078,
      "step": 10730
    },
    {
      "epoch": 0.9430979978925185,
      "grad_norm": 0.40432849526405334,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.496,
      "step": 10740
    },
    {
      "epoch": 0.9439761152089919,
      "grad_norm": 0.3420577347278595,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.5259,
      "step": 10750
    },
    {
      "epoch": 0.9448542325254654,
      "grad_norm": 0.3968747556209564,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.5083,
      "step": 10760
    },
    {
      "epoch": 0.9457323498419389,
      "grad_norm": 0.354543000459671,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.5218,
      "step": 10770
    },
    {
      "epoch": 0.9466104671584123,
      "grad_norm": 0.3537101447582245,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4822,
      "step": 10780
    },
    {
      "epoch": 0.9474885844748858,
      "grad_norm": 0.37943634390830994,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.5377,
      "step": 10790
    },
    {
      "epoch": 0.9483667017913593,
      "grad_norm": 0.3877718448638916,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.5086,
      "step": 10800
    },
    {
      "epoch": 0.9492448191078328,
      "grad_norm": 0.3664681613445282,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.4856,
      "step": 10810
    },
    {
      "epoch": 0.9501229364243063,
      "grad_norm": 0.3573661744594574,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.4867,
      "step": 10820
    },
    {
      "epoch": 0.9510010537407798,
      "grad_norm": 0.3850177526473999,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.4949,
      "step": 10830
    },
    {
      "epoch": 0.9518791710572533,
      "grad_norm": 0.35647669434547424,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.5118,
      "step": 10840
    },
    {
      "epoch": 0.9527572883737268,
      "grad_norm": 0.3525557219982147,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.4956,
      "step": 10850
    },
    {
      "epoch": 0.9536354056902002,
      "grad_norm": 0.32639312744140625,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.5045,
      "step": 10860
    },
    {
      "epoch": 0.9545135230066737,
      "grad_norm": 0.39365288615226746,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.4952,
      "step": 10870
    },
    {
      "epoch": 0.9553916403231472,
      "grad_norm": 0.36014804244041443,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.503,
      "step": 10880
    },
    {
      "epoch": 0.9562697576396206,
      "grad_norm": 0.34816285967826843,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.5113,
      "step": 10890
    },
    {
      "epoch": 0.9571478749560941,
      "grad_norm": 0.3804636299610138,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.5174,
      "step": 10900
    },
    {
      "epoch": 0.9580259922725676,
      "grad_norm": 0.3751678764820099,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.4838,
      "step": 10910
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 0.36495593190193176,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.4987,
      "step": 10920
    },
    {
      "epoch": 0.9597822269055146,
      "grad_norm": 0.37349212169647217,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.4981,
      "step": 10930
    },
    {
      "epoch": 0.9606603442219881,
      "grad_norm": 0.33641213178634644,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.5175,
      "step": 10940
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 0.4052804112434387,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.5313,
      "step": 10950
    },
    {
      "epoch": 0.962416578854935,
      "grad_norm": 0.3545972406864166,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.5126,
      "step": 10960
    },
    {
      "epoch": 0.9632946961714085,
      "grad_norm": 0.3503860831260681,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.5121,
      "step": 10970
    },
    {
      "epoch": 0.964172813487882,
      "grad_norm": 0.37088459730148315,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.512,
      "step": 10980
    },
    {
      "epoch": 0.9650509308043554,
      "grad_norm": 0.4455522298812866,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.5166,
      "step": 10990
    },
    {
      "epoch": 0.9659290481208289,
      "grad_norm": 0.3878462016582489,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.5446,
      "step": 11000
    },
    {
      "epoch": 0.9668071654373024,
      "grad_norm": 0.4048650860786438,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.5061,
      "step": 11010
    },
    {
      "epoch": 0.9676852827537759,
      "grad_norm": 0.3744140565395355,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.5332,
      "step": 11020
    },
    {
      "epoch": 0.9685634000702494,
      "grad_norm": 0.3554249107837677,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.5054,
      "step": 11030
    },
    {
      "epoch": 0.9694415173867229,
      "grad_norm": 0.4837280809879303,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.5139,
      "step": 11040
    },
    {
      "epoch": 0.9703196347031964,
      "grad_norm": 0.3678087294101715,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.5222,
      "step": 11050
    },
    {
      "epoch": 0.9711977520196698,
      "grad_norm": 0.3621269166469574,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.516,
      "step": 11060
    },
    {
      "epoch": 0.9720758693361433,
      "grad_norm": 0.35130324959754944,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.5196,
      "step": 11070
    },
    {
      "epoch": 0.9729539866526168,
      "grad_norm": 0.36483705043792725,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.5417,
      "step": 11080
    },
    {
      "epoch": 0.9738321039690903,
      "grad_norm": 0.37802159786224365,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.5235,
      "step": 11090
    },
    {
      "epoch": 0.9747102212855637,
      "grad_norm": 0.5161332488059998,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.5261,
      "step": 11100
    },
    {
      "epoch": 0.9755883386020372,
      "grad_norm": 0.4065404236316681,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.5031,
      "step": 11110
    },
    {
      "epoch": 0.9764664559185107,
      "grad_norm": 0.35460755228996277,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.4963,
      "step": 11120
    },
    {
      "epoch": 0.9773445732349842,
      "grad_norm": 0.36679762601852417,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.5069,
      "step": 11130
    },
    {
      "epoch": 0.9782226905514577,
      "grad_norm": 0.34028077125549316,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.5063,
      "step": 11140
    },
    {
      "epoch": 0.9791008078679312,
      "grad_norm": 0.3446286618709564,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.4968,
      "step": 11150
    },
    {
      "epoch": 0.9799789251844047,
      "grad_norm": 0.35653531551361084,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.5028,
      "step": 11160
    },
    {
      "epoch": 0.9808570425008781,
      "grad_norm": 0.3821895718574524,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.5111,
      "step": 11170
    },
    {
      "epoch": 0.9817351598173516,
      "grad_norm": 0.3923240303993225,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.4959,
      "step": 11180
    },
    {
      "epoch": 0.9826132771338251,
      "grad_norm": 0.35956841707229614,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.5141,
      "step": 11190
    },
    {
      "epoch": 0.9828767123287672,
      "step": 11193,
      "total_flos": 4.9965855215861105e+19,
      "train_loss": 0.04665133536240056,
      "train_runtime": 11396.3148,
      "train_samples_per_second": 11.786,
      "train_steps_per_second": 0.982
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "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": 4.9965855215861105e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}