{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4684858123378254,
  "eval_steps": 500,
  "global_step": 5597,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008370302167908262,
      "grad_norm": 5.156661510467529,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 1.5667,
      "step": 10
    },
    {
      "epoch": 0.0016740604335816524,
      "grad_norm": 2.557840347290039,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.3469,
      "step": 20
    },
    {
      "epoch": 0.0025110906503724786,
      "grad_norm": 1.6066818237304688,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.2057,
      "step": 30
    },
    {
      "epoch": 0.0033481208671633047,
      "grad_norm": 1.1234897375106812,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 1.0592,
      "step": 40
    },
    {
      "epoch": 0.004185151083954131,
      "grad_norm": 1.0128376483917236,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.9068,
      "step": 50
    },
    {
      "epoch": 0.005022181300744957,
      "grad_norm": 0.7160766124725342,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.8015,
      "step": 60
    },
    {
      "epoch": 0.005859211517535783,
      "grad_norm": 0.47716090083122253,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.8152,
      "step": 70
    },
    {
      "epoch": 0.006696241734326609,
      "grad_norm": 0.5273013114929199,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.7991,
      "step": 80
    },
    {
      "epoch": 0.007533271951117435,
      "grad_norm": 0.4626390039920807,
      "learning_rate": 1.005e-05,
      "loss": 0.7387,
      "step": 90
    },
    {
      "epoch": 0.008370302167908262,
      "grad_norm": 0.5121504664421082,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.7605,
      "step": 100
    },
    {
      "epoch": 0.009207332384699087,
      "grad_norm": 0.4082605540752411,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.7623,
      "step": 110
    },
    {
      "epoch": 0.010044362601489915,
      "grad_norm": 0.5059930682182312,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.764,
      "step": 120
    },
    {
      "epoch": 0.01088139281828074,
      "grad_norm": 0.751139223575592,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.7645,
      "step": 130
    },
    {
      "epoch": 0.011718423035071566,
      "grad_norm": 0.492674320936203,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.781,
      "step": 140
    },
    {
      "epoch": 0.012555453251862391,
      "grad_norm": 0.566781222820282,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.7548,
      "step": 150
    },
    {
      "epoch": 0.013392483468653219,
      "grad_norm": 0.5136772990226746,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.7799,
      "step": 160
    },
    {
      "epoch": 0.014229513685444044,
      "grad_norm": 0.40454524755477905,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.7748,
      "step": 170
    },
    {
      "epoch": 0.01506654390223487,
      "grad_norm": 0.4504382014274597,
      "learning_rate": 2e-05,
      "loss": 0.7757,
      "step": 180
    },
    {
      "epoch": 0.015903574119025696,
      "grad_norm": 0.5842396020889282,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 190
    },
    {
      "epoch": 0.016740604335816523,
      "grad_norm": 0.42499402165412903,
      "learning_rate": 2e-05,
      "loss": 0.7533,
      "step": 200
    },
    {
      "epoch": 0.01757763455260735,
      "grad_norm": 0.4922196567058563,
      "learning_rate": 2e-05,
      "loss": 0.7639,
      "step": 210
    },
    {
      "epoch": 0.018414664769398174,
      "grad_norm": 0.4512643814086914,
      "learning_rate": 2e-05,
      "loss": 0.7663,
      "step": 220
    },
    {
      "epoch": 0.019251694986189002,
      "grad_norm": 0.4370189607143402,
      "learning_rate": 2e-05,
      "loss": 0.765,
      "step": 230
    },
    {
      "epoch": 0.02008872520297983,
      "grad_norm": 0.41970548033714294,
      "learning_rate": 2e-05,
      "loss": 0.745,
      "step": 240
    },
    {
      "epoch": 0.020925755419770653,
      "grad_norm": 0.45108288526535034,
      "learning_rate": 2e-05,
      "loss": 0.7364,
      "step": 250
    },
    {
      "epoch": 0.02176278563656148,
      "grad_norm": 0.42641422152519226,
      "learning_rate": 2e-05,
      "loss": 0.732,
      "step": 260
    },
    {
      "epoch": 0.022599815853352308,
      "grad_norm": 0.3930014371871948,
      "learning_rate": 2e-05,
      "loss": 0.7634,
      "step": 270
    },
    {
      "epoch": 0.02343684607014313,
      "grad_norm": 0.43919554352760315,
      "learning_rate": 2e-05,
      "loss": 0.7621,
      "step": 280
    },
    {
      "epoch": 0.02427387628693396,
      "grad_norm": 0.4371900260448456,
      "learning_rate": 2e-05,
      "loss": 0.7379,
      "step": 290
    },
    {
      "epoch": 0.025110906503724783,
      "grad_norm": 0.4431065618991852,
      "learning_rate": 2e-05,
      "loss": 0.7306,
      "step": 300
    },
    {
      "epoch": 0.02594793672051561,
      "grad_norm": 0.4054403603076935,
      "learning_rate": 2e-05,
      "loss": 0.7532,
      "step": 310
    },
    {
      "epoch": 0.026784966937306438,
      "grad_norm": 0.43971535563468933,
      "learning_rate": 2e-05,
      "loss": 0.7527,
      "step": 320
    },
    {
      "epoch": 0.02762199715409726,
      "grad_norm": 1.0866317749023438,
      "learning_rate": 2e-05,
      "loss": 0.7705,
      "step": 330
    },
    {
      "epoch": 0.02845902737088809,
      "grad_norm": 0.4061005115509033,
      "learning_rate": 2e-05,
      "loss": 0.7739,
      "step": 340
    },
    {
      "epoch": 0.029296057587678916,
      "grad_norm": 0.5896549820899963,
      "learning_rate": 2e-05,
      "loss": 0.7655,
      "step": 350
    },
    {
      "epoch": 0.03013308780446974,
      "grad_norm": 0.44174519181251526,
      "learning_rate": 2e-05,
      "loss": 0.7502,
      "step": 360
    },
    {
      "epoch": 0.030970118021260568,
      "grad_norm": 0.4336935877799988,
      "learning_rate": 2e-05,
      "loss": 0.7119,
      "step": 370
    },
    {
      "epoch": 0.03180714823805139,
      "grad_norm": 0.43263721466064453,
      "learning_rate": 2e-05,
      "loss": 0.7308,
      "step": 380
    },
    {
      "epoch": 0.03264417845484222,
      "grad_norm": 0.4120350480079651,
      "learning_rate": 2e-05,
      "loss": 0.7464,
      "step": 390
    },
    {
      "epoch": 0.033481208671633046,
      "grad_norm": 0.4036485254764557,
      "learning_rate": 2e-05,
      "loss": 0.7601,
      "step": 400
    },
    {
      "epoch": 0.03431823888842387,
      "grad_norm": 0.41021424531936646,
      "learning_rate": 2e-05,
      "loss": 0.7281,
      "step": 410
    },
    {
      "epoch": 0.0351552691052147,
      "grad_norm": 0.4482649564743042,
      "learning_rate": 2e-05,
      "loss": 0.7316,
      "step": 420
    },
    {
      "epoch": 0.035992299322005525,
      "grad_norm": 0.4150967299938202,
      "learning_rate": 2e-05,
      "loss": 0.7365,
      "step": 430
    },
    {
      "epoch": 0.03682932953879635,
      "grad_norm": 0.3876197934150696,
      "learning_rate": 2e-05,
      "loss": 0.7511,
      "step": 440
    },
    {
      "epoch": 0.03766635975558718,
      "grad_norm": 0.40040096640586853,
      "learning_rate": 2e-05,
      "loss": 0.793,
      "step": 450
    },
    {
      "epoch": 0.038503389972378003,
      "grad_norm": 0.4496796727180481,
      "learning_rate": 2e-05,
      "loss": 0.7423,
      "step": 460
    },
    {
      "epoch": 0.03934042018916883,
      "grad_norm": 0.4478641152381897,
      "learning_rate": 2e-05,
      "loss": 0.7485,
      "step": 470
    },
    {
      "epoch": 0.04017745040595966,
      "grad_norm": 0.4376203715801239,
      "learning_rate": 2e-05,
      "loss": 0.7909,
      "step": 480
    },
    {
      "epoch": 0.04101448062275048,
      "grad_norm": 0.42852625250816345,
      "learning_rate": 2e-05,
      "loss": 0.7159,
      "step": 490
    },
    {
      "epoch": 0.041851510839541306,
      "grad_norm": 0.42360687255859375,
      "learning_rate": 2e-05,
      "loss": 0.7401,
      "step": 500
    },
    {
      "epoch": 0.04268854105633214,
      "grad_norm": 0.38901394605636597,
      "learning_rate": 2e-05,
      "loss": 0.7284,
      "step": 510
    },
    {
      "epoch": 0.04352557127312296,
      "grad_norm": 0.4375343322753906,
      "learning_rate": 2e-05,
      "loss": 0.738,
      "step": 520
    },
    {
      "epoch": 0.044362601489913785,
      "grad_norm": 0.44532549381256104,
      "learning_rate": 2e-05,
      "loss": 0.7469,
      "step": 530
    },
    {
      "epoch": 0.045199631706704615,
      "grad_norm": 0.46848392486572266,
      "learning_rate": 2e-05,
      "loss": 0.7366,
      "step": 540
    },
    {
      "epoch": 0.04603666192349544,
      "grad_norm": 0.3851851224899292,
      "learning_rate": 2e-05,
      "loss": 0.7335,
      "step": 550
    },
    {
      "epoch": 0.04687369214028626,
      "grad_norm": 0.40839746594429016,
      "learning_rate": 2e-05,
      "loss": 0.7053,
      "step": 560
    },
    {
      "epoch": 0.04771072235707709,
      "grad_norm": 0.4012123644351959,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 570
    },
    {
      "epoch": 0.04854775257386792,
      "grad_norm": 0.3855404555797577,
      "learning_rate": 2e-05,
      "loss": 0.7133,
      "step": 580
    },
    {
      "epoch": 0.04938478279065874,
      "grad_norm": 0.3988204002380371,
      "learning_rate": 2e-05,
      "loss": 0.7541,
      "step": 590
    },
    {
      "epoch": 0.050221813007449566,
      "grad_norm": 0.4450335204601288,
      "learning_rate": 2e-05,
      "loss": 0.7238,
      "step": 600
    },
    {
      "epoch": 0.0510588432242404,
      "grad_norm": 0.4375969469547272,
      "learning_rate": 2e-05,
      "loss": 0.7044,
      "step": 610
    },
    {
      "epoch": 0.05189587344103122,
      "grad_norm": 0.4379328787326813,
      "learning_rate": 2e-05,
      "loss": 0.726,
      "step": 620
    },
    {
      "epoch": 0.052732903657822044,
      "grad_norm": 0.39580798149108887,
      "learning_rate": 2e-05,
      "loss": 0.7463,
      "step": 630
    },
    {
      "epoch": 0.053569933874612875,
      "grad_norm": 0.41373711824417114,
      "learning_rate": 2e-05,
      "loss": 0.7499,
      "step": 640
    },
    {
      "epoch": 0.0544069640914037,
      "grad_norm": 0.4301176369190216,
      "learning_rate": 2e-05,
      "loss": 0.7341,
      "step": 650
    },
    {
      "epoch": 0.05524399430819452,
      "grad_norm": 0.45098692178726196,
      "learning_rate": 2e-05,
      "loss": 0.7318,
      "step": 660
    },
    {
      "epoch": 0.056081024524985354,
      "grad_norm": 0.4196617305278778,
      "learning_rate": 2e-05,
      "loss": 0.734,
      "step": 670
    },
    {
      "epoch": 0.05691805474177618,
      "grad_norm": 0.4272335171699524,
      "learning_rate": 2e-05,
      "loss": 0.7083,
      "step": 680
    },
    {
      "epoch": 0.057755084958567,
      "grad_norm": 0.4190222918987274,
      "learning_rate": 2e-05,
      "loss": 0.7272,
      "step": 690
    },
    {
      "epoch": 0.05859211517535783,
      "grad_norm": 0.42260268330574036,
      "learning_rate": 2e-05,
      "loss": 0.7266,
      "step": 700
    },
    {
      "epoch": 0.059429145392148656,
      "grad_norm": 0.3979990780353546,
      "learning_rate": 2e-05,
      "loss": 0.7265,
      "step": 710
    },
    {
      "epoch": 0.06026617560893948,
      "grad_norm": 0.41610896587371826,
      "learning_rate": 2e-05,
      "loss": 0.7308,
      "step": 720
    },
    {
      "epoch": 0.06110320582573031,
      "grad_norm": 0.44141194224357605,
      "learning_rate": 2e-05,
      "loss": 0.752,
      "step": 730
    },
    {
      "epoch": 0.061940236042521135,
      "grad_norm": 0.40890923142433167,
      "learning_rate": 2e-05,
      "loss": 0.7552,
      "step": 740
    },
    {
      "epoch": 0.06277726625931196,
      "grad_norm": 0.37664929032325745,
      "learning_rate": 2e-05,
      "loss": 0.7363,
      "step": 750
    },
    {
      "epoch": 0.06361429647610278,
      "grad_norm": 0.3718191683292389,
      "learning_rate": 2e-05,
      "loss": 0.7458,
      "step": 760
    },
    {
      "epoch": 0.0644513266928936,
      "grad_norm": 0.4162319004535675,
      "learning_rate": 2e-05,
      "loss": 0.7587,
      "step": 770
    },
    {
      "epoch": 0.06528835690968444,
      "grad_norm": 0.5250889658927917,
      "learning_rate": 2e-05,
      "loss": 0.733,
      "step": 780
    },
    {
      "epoch": 0.06612538712647527,
      "grad_norm": 0.3899112343788147,
      "learning_rate": 2e-05,
      "loss": 0.731,
      "step": 790
    },
    {
      "epoch": 0.06696241734326609,
      "grad_norm": 0.39714515209198,
      "learning_rate": 2e-05,
      "loss": 0.7397,
      "step": 800
    },
    {
      "epoch": 0.06779944756005692,
      "grad_norm": 0.4384140372276306,
      "learning_rate": 2e-05,
      "loss": 0.7334,
      "step": 810
    },
    {
      "epoch": 0.06863647777684774,
      "grad_norm": 0.37715354561805725,
      "learning_rate": 2e-05,
      "loss": 0.7442,
      "step": 820
    },
    {
      "epoch": 0.06947350799363856,
      "grad_norm": 0.4676320552825928,
      "learning_rate": 2e-05,
      "loss": 0.729,
      "step": 830
    },
    {
      "epoch": 0.0703105382104294,
      "grad_norm": 0.4101060926914215,
      "learning_rate": 2e-05,
      "loss": 0.7412,
      "step": 840
    },
    {
      "epoch": 0.07114756842722023,
      "grad_norm": 0.40823832154273987,
      "learning_rate": 2e-05,
      "loss": 0.7404,
      "step": 850
    },
    {
      "epoch": 0.07198459864401105,
      "grad_norm": 0.35889768600463867,
      "learning_rate": 2e-05,
      "loss": 0.7301,
      "step": 860
    },
    {
      "epoch": 0.07282162886080187,
      "grad_norm": 0.35591280460357666,
      "learning_rate": 2e-05,
      "loss": 0.7434,
      "step": 870
    },
    {
      "epoch": 0.0736586590775927,
      "grad_norm": 0.3851705491542816,
      "learning_rate": 2e-05,
      "loss": 0.7002,
      "step": 880
    },
    {
      "epoch": 0.07449568929438352,
      "grad_norm": 0.3983948230743408,
      "learning_rate": 2e-05,
      "loss": 0.7274,
      "step": 890
    },
    {
      "epoch": 0.07533271951117436,
      "grad_norm": 0.3635563254356384,
      "learning_rate": 2e-05,
      "loss": 0.6975,
      "step": 900
    },
    {
      "epoch": 0.07616974972796518,
      "grad_norm": 0.3900563418865204,
      "learning_rate": 2e-05,
      "loss": 0.7467,
      "step": 910
    },
    {
      "epoch": 0.07700677994475601,
      "grad_norm": 0.4121996760368347,
      "learning_rate": 2e-05,
      "loss": 0.7181,
      "step": 920
    },
    {
      "epoch": 0.07784381016154683,
      "grad_norm": 0.35744690895080566,
      "learning_rate": 2e-05,
      "loss": 0.72,
      "step": 930
    },
    {
      "epoch": 0.07868084037833765,
      "grad_norm": 0.44155263900756836,
      "learning_rate": 2e-05,
      "loss": 0.714,
      "step": 940
    },
    {
      "epoch": 0.07951787059512848,
      "grad_norm": 0.415809690952301,
      "learning_rate": 2e-05,
      "loss": 0.7518,
      "step": 950
    },
    {
      "epoch": 0.08035490081191932,
      "grad_norm": 0.3967406153678894,
      "learning_rate": 2e-05,
      "loss": 0.7445,
      "step": 960
    },
    {
      "epoch": 0.08119193102871014,
      "grad_norm": 0.4007624089717865,
      "learning_rate": 2e-05,
      "loss": 0.747,
      "step": 970
    },
    {
      "epoch": 0.08202896124550096,
      "grad_norm": 0.38344693183898926,
      "learning_rate": 2e-05,
      "loss": 0.7295,
      "step": 980
    },
    {
      "epoch": 0.08286599146229179,
      "grad_norm": 0.45661893486976624,
      "learning_rate": 2e-05,
      "loss": 0.7111,
      "step": 990
    },
    {
      "epoch": 0.08370302167908261,
      "grad_norm": 0.4255678951740265,
      "learning_rate": 2e-05,
      "loss": 0.775,
      "step": 1000
    },
    {
      "epoch": 0.08454005189587344,
      "grad_norm": 0.49089524149894714,
      "learning_rate": 2e-05,
      "loss": 0.7387,
      "step": 1010
    },
    {
      "epoch": 0.08537708211266427,
      "grad_norm": 0.3930380642414093,
      "learning_rate": 2e-05,
      "loss": 0.7601,
      "step": 1020
    },
    {
      "epoch": 0.0862141123294551,
      "grad_norm": 0.4231725037097931,
      "learning_rate": 2e-05,
      "loss": 0.6822,
      "step": 1030
    },
    {
      "epoch": 0.08705114254624592,
      "grad_norm": 0.39020201563835144,
      "learning_rate": 2e-05,
      "loss": 0.7325,
      "step": 1040
    },
    {
      "epoch": 0.08788817276303675,
      "grad_norm": 0.429443359375,
      "learning_rate": 2e-05,
      "loss": 0.7194,
      "step": 1050
    },
    {
      "epoch": 0.08872520297982757,
      "grad_norm": 0.38595613837242126,
      "learning_rate": 2e-05,
      "loss": 0.7336,
      "step": 1060
    },
    {
      "epoch": 0.0895622331966184,
      "grad_norm": 0.371380090713501,
      "learning_rate": 2e-05,
      "loss": 0.7325,
      "step": 1070
    },
    {
      "epoch": 0.09039926341340923,
      "grad_norm": 0.396846204996109,
      "learning_rate": 2e-05,
      "loss": 0.7281,
      "step": 1080
    },
    {
      "epoch": 0.09123629363020005,
      "grad_norm": 0.42279285192489624,
      "learning_rate": 2e-05,
      "loss": 0.7383,
      "step": 1090
    },
    {
      "epoch": 0.09207332384699088,
      "grad_norm": 0.41259169578552246,
      "learning_rate": 2e-05,
      "loss": 0.7398,
      "step": 1100
    },
    {
      "epoch": 0.0929103540637817,
      "grad_norm": 0.43002089858055115,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 1110
    },
    {
      "epoch": 0.09374738428057253,
      "grad_norm": 0.4018864929676056,
      "learning_rate": 2e-05,
      "loss": 0.7542,
      "step": 1120
    },
    {
      "epoch": 0.09458441449736335,
      "grad_norm": 0.3879416882991791,
      "learning_rate": 2e-05,
      "loss": 0.724,
      "step": 1130
    },
    {
      "epoch": 0.09542144471415417,
      "grad_norm": 0.40612295269966125,
      "learning_rate": 2e-05,
      "loss": 0.7181,
      "step": 1140
    },
    {
      "epoch": 0.09625847493094501,
      "grad_norm": 0.3889906406402588,
      "learning_rate": 2e-05,
      "loss": 0.7206,
      "step": 1150
    },
    {
      "epoch": 0.09709550514773584,
      "grad_norm": 0.3631397783756256,
      "learning_rate": 2e-05,
      "loss": 0.7608,
      "step": 1160
    },
    {
      "epoch": 0.09793253536452666,
      "grad_norm": 0.3583589494228363,
      "learning_rate": 2e-05,
      "loss": 0.7257,
      "step": 1170
    },
    {
      "epoch": 0.09876956558131748,
      "grad_norm": 0.38138115406036377,
      "learning_rate": 2e-05,
      "loss": 0.7296,
      "step": 1180
    },
    {
      "epoch": 0.09960659579810831,
      "grad_norm": 0.381708025932312,
      "learning_rate": 2e-05,
      "loss": 0.7275,
      "step": 1190
    },
    {
      "epoch": 0.10044362601489913,
      "grad_norm": 0.3952488601207733,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 1200
    },
    {
      "epoch": 0.10128065623168997,
      "grad_norm": 0.369489848613739,
      "learning_rate": 2e-05,
      "loss": 0.6827,
      "step": 1210
    },
    {
      "epoch": 0.1021176864484808,
      "grad_norm": 0.3854788541793823,
      "learning_rate": 2e-05,
      "loss": 0.7141,
      "step": 1220
    },
    {
      "epoch": 0.10295471666527162,
      "grad_norm": 0.44547921419143677,
      "learning_rate": 2e-05,
      "loss": 0.7465,
      "step": 1230
    },
    {
      "epoch": 0.10379174688206244,
      "grad_norm": 0.3807377219200134,
      "learning_rate": 2e-05,
      "loss": 0.6965,
      "step": 1240
    },
    {
      "epoch": 0.10462877709885327,
      "grad_norm": 1.0734481811523438,
      "learning_rate": 2e-05,
      "loss": 0.7228,
      "step": 1250
    },
    {
      "epoch": 0.10546580731564409,
      "grad_norm": 0.3832617700099945,
      "learning_rate": 2e-05,
      "loss": 0.76,
      "step": 1260
    },
    {
      "epoch": 0.10630283753243493,
      "grad_norm": 0.4228065013885498,
      "learning_rate": 2e-05,
      "loss": 0.7379,
      "step": 1270
    },
    {
      "epoch": 0.10713986774922575,
      "grad_norm": 0.39900723099708557,
      "learning_rate": 2e-05,
      "loss": 0.6866,
      "step": 1280
    },
    {
      "epoch": 0.10797689796601657,
      "grad_norm": 0.5197460055351257,
      "learning_rate": 2e-05,
      "loss": 0.6915,
      "step": 1290
    },
    {
      "epoch": 0.1088139281828074,
      "grad_norm": 0.6023410558700562,
      "learning_rate": 2e-05,
      "loss": 0.7019,
      "step": 1300
    },
    {
      "epoch": 0.10965095839959822,
      "grad_norm": 0.4066433310508728,
      "learning_rate": 2e-05,
      "loss": 0.7001,
      "step": 1310
    },
    {
      "epoch": 0.11048798861638905,
      "grad_norm": 0.382357656955719,
      "learning_rate": 2e-05,
      "loss": 0.7287,
      "step": 1320
    },
    {
      "epoch": 0.11132501883317988,
      "grad_norm": 0.3749663829803467,
      "learning_rate": 2e-05,
      "loss": 0.7429,
      "step": 1330
    },
    {
      "epoch": 0.11216204904997071,
      "grad_norm": 0.3689495325088501,
      "learning_rate": 2e-05,
      "loss": 0.6842,
      "step": 1340
    },
    {
      "epoch": 0.11299907926676153,
      "grad_norm": 0.41168835759162903,
      "learning_rate": 2e-05,
      "loss": 0.7533,
      "step": 1350
    },
    {
      "epoch": 0.11383610948355236,
      "grad_norm": 0.39960694313049316,
      "learning_rate": 2e-05,
      "loss": 0.7312,
      "step": 1360
    },
    {
      "epoch": 0.11467313970034318,
      "grad_norm": 0.3908539414405823,
      "learning_rate": 2e-05,
      "loss": 0.7085,
      "step": 1370
    },
    {
      "epoch": 0.115510169917134,
      "grad_norm": 0.38182100653648376,
      "learning_rate": 2e-05,
      "loss": 0.7271,
      "step": 1380
    },
    {
      "epoch": 0.11634720013392484,
      "grad_norm": 0.405620276927948,
      "learning_rate": 2e-05,
      "loss": 0.7235,
      "step": 1390
    },
    {
      "epoch": 0.11718423035071567,
      "grad_norm": 0.3966982364654541,
      "learning_rate": 2e-05,
      "loss": 0.7371,
      "step": 1400
    },
    {
      "epoch": 0.11802126056750649,
      "grad_norm": 0.3768328130245209,
      "learning_rate": 2e-05,
      "loss": 0.7172,
      "step": 1410
    },
    {
      "epoch": 0.11885829078429731,
      "grad_norm": 0.4282349944114685,
      "learning_rate": 2e-05,
      "loss": 0.7318,
      "step": 1420
    },
    {
      "epoch": 0.11969532100108814,
      "grad_norm": 0.39045482873916626,
      "learning_rate": 2e-05,
      "loss": 0.704,
      "step": 1430
    },
    {
      "epoch": 0.12053235121787896,
      "grad_norm": 0.3846886456012726,
      "learning_rate": 2e-05,
      "loss": 0.7069,
      "step": 1440
    },
    {
      "epoch": 0.12136938143466978,
      "grad_norm": 0.3845898509025574,
      "learning_rate": 2e-05,
      "loss": 0.703,
      "step": 1450
    },
    {
      "epoch": 0.12220641165146062,
      "grad_norm": 0.4043080508708954,
      "learning_rate": 2e-05,
      "loss": 0.7329,
      "step": 1460
    },
    {
      "epoch": 0.12304344186825145,
      "grad_norm": 0.353148877620697,
      "learning_rate": 2e-05,
      "loss": 0.7039,
      "step": 1470
    },
    {
      "epoch": 0.12388047208504227,
      "grad_norm": 0.3626961410045624,
      "learning_rate": 2e-05,
      "loss": 0.7125,
      "step": 1480
    },
    {
      "epoch": 0.1247175023018331,
      "grad_norm": 0.40537577867507935,
      "learning_rate": 2e-05,
      "loss": 0.6874,
      "step": 1490
    },
    {
      "epoch": 0.12555453251862392,
      "grad_norm": 0.38374269008636475,
      "learning_rate": 2e-05,
      "loss": 0.7166,
      "step": 1500
    },
    {
      "epoch": 0.12639156273541474,
      "grad_norm": 0.3687508702278137,
      "learning_rate": 2e-05,
      "loss": 0.7042,
      "step": 1510
    },
    {
      "epoch": 0.12722859295220557,
      "grad_norm": 0.38297727704048157,
      "learning_rate": 2e-05,
      "loss": 0.7276,
      "step": 1520
    },
    {
      "epoch": 0.1280656231689964,
      "grad_norm": 0.40739986300468445,
      "learning_rate": 2e-05,
      "loss": 0.7352,
      "step": 1530
    },
    {
      "epoch": 0.1289026533857872,
      "grad_norm": 0.5073732733726501,
      "learning_rate": 2e-05,
      "loss": 0.7407,
      "step": 1540
    },
    {
      "epoch": 0.12973968360257807,
      "grad_norm": 0.3655388653278351,
      "learning_rate": 2e-05,
      "loss": 0.707,
      "step": 1550
    },
    {
      "epoch": 0.1305767138193689,
      "grad_norm": 0.38478103280067444,
      "learning_rate": 2e-05,
      "loss": 0.7155,
      "step": 1560
    },
    {
      "epoch": 0.1314137440361597,
      "grad_norm": 0.3790566027164459,
      "learning_rate": 2e-05,
      "loss": 0.7048,
      "step": 1570
    },
    {
      "epoch": 0.13225077425295054,
      "grad_norm": 0.3985552489757538,
      "learning_rate": 2e-05,
      "loss": 0.7469,
      "step": 1580
    },
    {
      "epoch": 0.13308780446974136,
      "grad_norm": 0.3573896586894989,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 1590
    },
    {
      "epoch": 0.13392483468653218,
      "grad_norm": 0.3679613471031189,
      "learning_rate": 2e-05,
      "loss": 0.7063,
      "step": 1600
    },
    {
      "epoch": 0.134761864903323,
      "grad_norm": 0.4195422828197479,
      "learning_rate": 2e-05,
      "loss": 0.7214,
      "step": 1610
    },
    {
      "epoch": 0.13559889512011383,
      "grad_norm": 0.4262971878051758,
      "learning_rate": 2e-05,
      "loss": 0.7214,
      "step": 1620
    },
    {
      "epoch": 0.13643592533690466,
      "grad_norm": 0.3953056335449219,
      "learning_rate": 2e-05,
      "loss": 0.7457,
      "step": 1630
    },
    {
      "epoch": 0.13727295555369548,
      "grad_norm": 0.3896598517894745,
      "learning_rate": 2e-05,
      "loss": 0.68,
      "step": 1640
    },
    {
      "epoch": 0.1381099857704863,
      "grad_norm": 0.39134523272514343,
      "learning_rate": 2e-05,
      "loss": 0.7233,
      "step": 1650
    },
    {
      "epoch": 0.13894701598727713,
      "grad_norm": 0.38543522357940674,
      "learning_rate": 2e-05,
      "loss": 0.7078,
      "step": 1660
    },
    {
      "epoch": 0.13978404620406798,
      "grad_norm": 0.37859565019607544,
      "learning_rate": 2e-05,
      "loss": 0.7141,
      "step": 1670
    },
    {
      "epoch": 0.1406210764208588,
      "grad_norm": 0.3654524087905884,
      "learning_rate": 2e-05,
      "loss": 0.6817,
      "step": 1680
    },
    {
      "epoch": 0.14145810663764963,
      "grad_norm": 0.3827599585056305,
      "learning_rate": 2e-05,
      "loss": 0.6995,
      "step": 1690
    },
    {
      "epoch": 0.14229513685444045,
      "grad_norm": 0.3548571467399597,
      "learning_rate": 2e-05,
      "loss": 0.7106,
      "step": 1700
    },
    {
      "epoch": 0.14313216707123128,
      "grad_norm": 0.4650750756263733,
      "learning_rate": 2e-05,
      "loss": 0.7498,
      "step": 1710
    },
    {
      "epoch": 0.1439691972880221,
      "grad_norm": 0.38478001952171326,
      "learning_rate": 2e-05,
      "loss": 0.7699,
      "step": 1720
    },
    {
      "epoch": 0.14480622750481292,
      "grad_norm": 0.37154585123062134,
      "learning_rate": 2e-05,
      "loss": 0.6941,
      "step": 1730
    },
    {
      "epoch": 0.14564325772160375,
      "grad_norm": 0.38709065318107605,
      "learning_rate": 2e-05,
      "loss": 0.7337,
      "step": 1740
    },
    {
      "epoch": 0.14648028793839457,
      "grad_norm": 1.7314486503601074,
      "learning_rate": 2e-05,
      "loss": 0.7005,
      "step": 1750
    },
    {
      "epoch": 0.1473173181551854,
      "grad_norm": 0.3551817834377289,
      "learning_rate": 2e-05,
      "loss": 0.7132,
      "step": 1760
    },
    {
      "epoch": 0.14815434837197622,
      "grad_norm": 0.3846317529678345,
      "learning_rate": 2e-05,
      "loss": 0.6762,
      "step": 1770
    },
    {
      "epoch": 0.14899137858876704,
      "grad_norm": 0.37913134694099426,
      "learning_rate": 2e-05,
      "loss": 0.7101,
      "step": 1780
    },
    {
      "epoch": 0.1498284088055579,
      "grad_norm": 0.3794521391391754,
      "learning_rate": 2e-05,
      "loss": 0.7327,
      "step": 1790
    },
    {
      "epoch": 0.15066543902234872,
      "grad_norm": 0.3727302551269531,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 1800
    },
    {
      "epoch": 0.15150246923913954,
      "grad_norm": 0.39975666999816895,
      "learning_rate": 2e-05,
      "loss": 0.7283,
      "step": 1810
    },
    {
      "epoch": 0.15233949945593037,
      "grad_norm": 0.37394288182258606,
      "learning_rate": 2e-05,
      "loss": 0.6815,
      "step": 1820
    },
    {
      "epoch": 0.1531765296727212,
      "grad_norm": 0.38191404938697815,
      "learning_rate": 2e-05,
      "loss": 0.723,
      "step": 1830
    },
    {
      "epoch": 0.15401355988951201,
      "grad_norm": 0.3914385437965393,
      "learning_rate": 2e-05,
      "loss": 0.6959,
      "step": 1840
    },
    {
      "epoch": 0.15485059010630284,
      "grad_norm": 0.36960726976394653,
      "learning_rate": 2e-05,
      "loss": 0.6981,
      "step": 1850
    },
    {
      "epoch": 0.15568762032309366,
      "grad_norm": 0.4533153474330902,
      "learning_rate": 2e-05,
      "loss": 0.7042,
      "step": 1860
    },
    {
      "epoch": 0.15652465053988449,
      "grad_norm": 0.390516459941864,
      "learning_rate": 2e-05,
      "loss": 0.7285,
      "step": 1870
    },
    {
      "epoch": 0.1573616807566753,
      "grad_norm": 0.36687996983528137,
      "learning_rate": 2e-05,
      "loss": 0.7039,
      "step": 1880
    },
    {
      "epoch": 0.15819871097346613,
      "grad_norm": 0.3998259902000427,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 1890
    },
    {
      "epoch": 0.15903574119025696,
      "grad_norm": 0.3630969226360321,
      "learning_rate": 2e-05,
      "loss": 0.7156,
      "step": 1900
    },
    {
      "epoch": 0.15987277140704778,
      "grad_norm": 0.49144166707992554,
      "learning_rate": 2e-05,
      "loss": 0.7492,
      "step": 1910
    },
    {
      "epoch": 0.16070980162383863,
      "grad_norm": 0.3666890263557434,
      "learning_rate": 2e-05,
      "loss": 0.6957,
      "step": 1920
    },
    {
      "epoch": 0.16154683184062946,
      "grad_norm": 0.36240652203559875,
      "learning_rate": 2e-05,
      "loss": 0.7208,
      "step": 1930
    },
    {
      "epoch": 0.16238386205742028,
      "grad_norm": 0.37068548798561096,
      "learning_rate": 2e-05,
      "loss": 0.7593,
      "step": 1940
    },
    {
      "epoch": 0.1632208922742111,
      "grad_norm": 0.38422319293022156,
      "learning_rate": 2e-05,
      "loss": 0.718,
      "step": 1950
    },
    {
      "epoch": 0.16405792249100193,
      "grad_norm": 0.4526020884513855,
      "learning_rate": 2e-05,
      "loss": 0.7152,
      "step": 1960
    },
    {
      "epoch": 0.16489495270779275,
      "grad_norm": 0.42693546414375305,
      "learning_rate": 2e-05,
      "loss": 0.7031,
      "step": 1970
    },
    {
      "epoch": 0.16573198292458358,
      "grad_norm": 0.352283239364624,
      "learning_rate": 2e-05,
      "loss": 0.6788,
      "step": 1980
    },
    {
      "epoch": 0.1665690131413744,
      "grad_norm": 0.4254308342933655,
      "learning_rate": 2e-05,
      "loss": 0.7004,
      "step": 1990
    },
    {
      "epoch": 0.16740604335816522,
      "grad_norm": 0.37109827995300293,
      "learning_rate": 2e-05,
      "loss": 0.7183,
      "step": 2000
    },
    {
      "epoch": 0.16824307357495605,
      "grad_norm": 0.3963578939437866,
      "learning_rate": 2e-05,
      "loss": 0.7152,
      "step": 2010
    },
    {
      "epoch": 0.16908010379174687,
      "grad_norm": 0.4323606491088867,
      "learning_rate": 2e-05,
      "loss": 0.7275,
      "step": 2020
    },
    {
      "epoch": 0.1699171340085377,
      "grad_norm": 0.38154488801956177,
      "learning_rate": 2e-05,
      "loss": 0.7043,
      "step": 2030
    },
    {
      "epoch": 0.17075416422532855,
      "grad_norm": 0.37403401732444763,
      "learning_rate": 2e-05,
      "loss": 0.6961,
      "step": 2040
    },
    {
      "epoch": 0.17159119444211937,
      "grad_norm": 0.4297429919242859,
      "learning_rate": 2e-05,
      "loss": 0.7279,
      "step": 2050
    },
    {
      "epoch": 0.1724282246589102,
      "grad_norm": 0.3764204978942871,
      "learning_rate": 2e-05,
      "loss": 0.7169,
      "step": 2060
    },
    {
      "epoch": 0.17326525487570102,
      "grad_norm": 0.4536464512348175,
      "learning_rate": 2e-05,
      "loss": 0.7076,
      "step": 2070
    },
    {
      "epoch": 0.17410228509249184,
      "grad_norm": 0.3891889750957489,
      "learning_rate": 2e-05,
      "loss": 0.6937,
      "step": 2080
    },
    {
      "epoch": 0.17493931530928267,
      "grad_norm": 0.3979426920413971,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 2090
    },
    {
      "epoch": 0.1757763455260735,
      "grad_norm": 0.33533552289009094,
      "learning_rate": 2e-05,
      "loss": 0.6911,
      "step": 2100
    },
    {
      "epoch": 0.17661337574286431,
      "grad_norm": 0.4086405038833618,
      "learning_rate": 2e-05,
      "loss": 0.6961,
      "step": 2110
    },
    {
      "epoch": 0.17745040595965514,
      "grad_norm": 0.32550907135009766,
      "learning_rate": 2e-05,
      "loss": 0.7189,
      "step": 2120
    },
    {
      "epoch": 0.17828743617644596,
      "grad_norm": 0.41129428148269653,
      "learning_rate": 2e-05,
      "loss": 0.7132,
      "step": 2130
    },
    {
      "epoch": 0.1791244663932368,
      "grad_norm": 0.38379132747650146,
      "learning_rate": 2e-05,
      "loss": 0.7083,
      "step": 2140
    },
    {
      "epoch": 0.1799614966100276,
      "grad_norm": 0.33447548747062683,
      "learning_rate": 2e-05,
      "loss": 0.6957,
      "step": 2150
    },
    {
      "epoch": 0.18079852682681846,
      "grad_norm": 0.36413493752479553,
      "learning_rate": 2e-05,
      "loss": 0.6845,
      "step": 2160
    },
    {
      "epoch": 0.18163555704360929,
      "grad_norm": 0.3704940676689148,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 2170
    },
    {
      "epoch": 0.1824725872604001,
      "grad_norm": 0.37158524990081787,
      "learning_rate": 2e-05,
      "loss": 0.6945,
      "step": 2180
    },
    {
      "epoch": 0.18330961747719093,
      "grad_norm": 0.37925466895103455,
      "learning_rate": 2e-05,
      "loss": 0.7088,
      "step": 2190
    },
    {
      "epoch": 0.18414664769398176,
      "grad_norm": 0.3649779260158539,
      "learning_rate": 2e-05,
      "loss": 0.7342,
      "step": 2200
    },
    {
      "epoch": 0.18498367791077258,
      "grad_norm": 0.38125574588775635,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 2210
    },
    {
      "epoch": 0.1858207081275634,
      "grad_norm": 0.4088246524333954,
      "learning_rate": 2e-05,
      "loss": 0.7006,
      "step": 2220
    },
    {
      "epoch": 0.18665773834435423,
      "grad_norm": 0.4514538645744324,
      "learning_rate": 2e-05,
      "loss": 0.7126,
      "step": 2230
    },
    {
      "epoch": 0.18749476856114505,
      "grad_norm": 0.3832446336746216,
      "learning_rate": 2e-05,
      "loss": 0.6912,
      "step": 2240
    },
    {
      "epoch": 0.18833179877793588,
      "grad_norm": 0.37790173292160034,
      "learning_rate": 2e-05,
      "loss": 0.6738,
      "step": 2250
    },
    {
      "epoch": 0.1891688289947267,
      "grad_norm": 0.3716677129268646,
      "learning_rate": 2e-05,
      "loss": 0.7122,
      "step": 2260
    },
    {
      "epoch": 0.19000585921151752,
      "grad_norm": 0.3565468490123749,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 2270
    },
    {
      "epoch": 0.19084288942830835,
      "grad_norm": 0.37897953391075134,
      "learning_rate": 2e-05,
      "loss": 0.6941,
      "step": 2280
    },
    {
      "epoch": 0.1916799196450992,
      "grad_norm": 0.3777736723423004,
      "learning_rate": 2e-05,
      "loss": 0.7166,
      "step": 2290
    },
    {
      "epoch": 0.19251694986189002,
      "grad_norm": 0.3585146367549896,
      "learning_rate": 2e-05,
      "loss": 0.6928,
      "step": 2300
    },
    {
      "epoch": 0.19335398007868085,
      "grad_norm": 0.3698466122150421,
      "learning_rate": 2e-05,
      "loss": 0.6937,
      "step": 2310
    },
    {
      "epoch": 0.19419101029547167,
      "grad_norm": 0.3647162914276123,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 2320
    },
    {
      "epoch": 0.1950280405122625,
      "grad_norm": 0.4572893977165222,
      "learning_rate": 2e-05,
      "loss": 0.6978,
      "step": 2330
    },
    {
      "epoch": 0.19586507072905332,
      "grad_norm": 0.42052674293518066,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 2340
    },
    {
      "epoch": 0.19670210094584414,
      "grad_norm": 0.4102165102958679,
      "learning_rate": 2e-05,
      "loss": 0.7097,
      "step": 2350
    },
    {
      "epoch": 0.19753913116263497,
      "grad_norm": 0.3792290687561035,
      "learning_rate": 2e-05,
      "loss": 0.6917,
      "step": 2360
    },
    {
      "epoch": 0.1983761613794258,
      "grad_norm": 0.41515079140663147,
      "learning_rate": 2e-05,
      "loss": 0.7033,
      "step": 2370
    },
    {
      "epoch": 0.19921319159621662,
      "grad_norm": 0.44105610251426697,
      "learning_rate": 2e-05,
      "loss": 0.7093,
      "step": 2380
    },
    {
      "epoch": 0.20005022181300744,
      "grad_norm": 0.3615840971469879,
      "learning_rate": 2e-05,
      "loss": 0.7512,
      "step": 2390
    },
    {
      "epoch": 0.20088725202979826,
      "grad_norm": 0.3563784062862396,
      "learning_rate": 2e-05,
      "loss": 0.7046,
      "step": 2400
    },
    {
      "epoch": 0.20172428224658911,
      "grad_norm": 0.3593883216381073,
      "learning_rate": 2e-05,
      "loss": 0.7143,
      "step": 2410
    },
    {
      "epoch": 0.20256131246337994,
      "grad_norm": 0.375445157289505,
      "learning_rate": 2e-05,
      "loss": 0.678,
      "step": 2420
    },
    {
      "epoch": 0.20339834268017076,
      "grad_norm": 0.3783409595489502,
      "learning_rate": 2e-05,
      "loss": 0.6926,
      "step": 2430
    },
    {
      "epoch": 0.2042353728969616,
      "grad_norm": 0.3709818720817566,
      "learning_rate": 2e-05,
      "loss": 0.6974,
      "step": 2440
    },
    {
      "epoch": 0.2050724031137524,
      "grad_norm": 0.3591563403606415,
      "learning_rate": 2e-05,
      "loss": 0.7022,
      "step": 2450
    },
    {
      "epoch": 0.20590943333054323,
      "grad_norm": 0.42944005131721497,
      "learning_rate": 2e-05,
      "loss": 0.7151,
      "step": 2460
    },
    {
      "epoch": 0.20674646354733406,
      "grad_norm": 0.36380648612976074,
      "learning_rate": 2e-05,
      "loss": 0.6938,
      "step": 2470
    },
    {
      "epoch": 0.20758349376412488,
      "grad_norm": 0.3823877274990082,
      "learning_rate": 2e-05,
      "loss": 0.6853,
      "step": 2480
    },
    {
      "epoch": 0.2084205239809157,
      "grad_norm": 0.40340110659599304,
      "learning_rate": 2e-05,
      "loss": 0.6983,
      "step": 2490
    },
    {
      "epoch": 0.20925755419770653,
      "grad_norm": 0.3656036853790283,
      "learning_rate": 2e-05,
      "loss": 0.6927,
      "step": 2500
    },
    {
      "epoch": 0.21009458441449735,
      "grad_norm": 0.3953785002231598,
      "learning_rate": 2e-05,
      "loss": 0.6779,
      "step": 2510
    },
    {
      "epoch": 0.21093161463128818,
      "grad_norm": 0.40045875310897827,
      "learning_rate": 2e-05,
      "loss": 0.6812,
      "step": 2520
    },
    {
      "epoch": 0.211768644848079,
      "grad_norm": 0.3604848086833954,
      "learning_rate": 2e-05,
      "loss": 0.6908,
      "step": 2530
    },
    {
      "epoch": 0.21260567506486985,
      "grad_norm": 0.40517356991767883,
      "learning_rate": 2e-05,
      "loss": 0.7074,
      "step": 2540
    },
    {
      "epoch": 0.21344270528166068,
      "grad_norm": 0.3920334279537201,
      "learning_rate": 2e-05,
      "loss": 0.7021,
      "step": 2550
    },
    {
      "epoch": 0.2142797354984515,
      "grad_norm": 0.35391831398010254,
      "learning_rate": 2e-05,
      "loss": 0.704,
      "step": 2560
    },
    {
      "epoch": 0.21511676571524233,
      "grad_norm": 0.33125635981559753,
      "learning_rate": 2e-05,
      "loss": 0.6751,
      "step": 2570
    },
    {
      "epoch": 0.21595379593203315,
      "grad_norm": 0.3804548382759094,
      "learning_rate": 2e-05,
      "loss": 0.6929,
      "step": 2580
    },
    {
      "epoch": 0.21679082614882397,
      "grad_norm": 0.41652804613113403,
      "learning_rate": 2e-05,
      "loss": 0.7224,
      "step": 2590
    },
    {
      "epoch": 0.2176278563656148,
      "grad_norm": 0.37819674611091614,
      "learning_rate": 2e-05,
      "loss": 0.6942,
      "step": 2600
    },
    {
      "epoch": 0.21846488658240562,
      "grad_norm": 0.3700343668460846,
      "learning_rate": 2e-05,
      "loss": 0.6922,
      "step": 2610
    },
    {
      "epoch": 0.21930191679919644,
      "grad_norm": 0.39303967356681824,
      "learning_rate": 2e-05,
      "loss": 0.6836,
      "step": 2620
    },
    {
      "epoch": 0.22013894701598727,
      "grad_norm": 0.3704970180988312,
      "learning_rate": 2e-05,
      "loss": 0.6656,
      "step": 2630
    },
    {
      "epoch": 0.2209759772327781,
      "grad_norm": 0.35388532280921936,
      "learning_rate": 2e-05,
      "loss": 0.7203,
      "step": 2640
    },
    {
      "epoch": 0.22181300744956892,
      "grad_norm": 0.3718905746936798,
      "learning_rate": 2e-05,
      "loss": 0.6942,
      "step": 2650
    },
    {
      "epoch": 0.22265003766635977,
      "grad_norm": 0.36480191349983215,
      "learning_rate": 2e-05,
      "loss": 0.7213,
      "step": 2660
    },
    {
      "epoch": 0.2234870678831506,
      "grad_norm": 0.5288359522819519,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 2670
    },
    {
      "epoch": 0.22432409809994142,
      "grad_norm": 0.3652160167694092,
      "learning_rate": 2e-05,
      "loss": 0.708,
      "step": 2680
    },
    {
      "epoch": 0.22516112831673224,
      "grad_norm": 0.36390116810798645,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 2690
    },
    {
      "epoch": 0.22599815853352306,
      "grad_norm": 0.3840753734111786,
      "learning_rate": 2e-05,
      "loss": 0.6633,
      "step": 2700
    },
    {
      "epoch": 0.2268351887503139,
      "grad_norm": 0.3530333638191223,
      "learning_rate": 2e-05,
      "loss": 0.7124,
      "step": 2710
    },
    {
      "epoch": 0.2276722189671047,
      "grad_norm": 0.342545747756958,
      "learning_rate": 2e-05,
      "loss": 0.697,
      "step": 2720
    },
    {
      "epoch": 0.22850924918389554,
      "grad_norm": 0.5206401348114014,
      "learning_rate": 2e-05,
      "loss": 0.6976,
      "step": 2730
    },
    {
      "epoch": 0.22934627940068636,
      "grad_norm": 0.3752775192260742,
      "learning_rate": 2e-05,
      "loss": 0.7146,
      "step": 2740
    },
    {
      "epoch": 0.23018330961747718,
      "grad_norm": 0.3544803857803345,
      "learning_rate": 2e-05,
      "loss": 0.7199,
      "step": 2750
    },
    {
      "epoch": 0.231020339834268,
      "grad_norm": 0.47715863585472107,
      "learning_rate": 2e-05,
      "loss": 0.7116,
      "step": 2760
    },
    {
      "epoch": 0.23185737005105883,
      "grad_norm": 0.3936082720756531,
      "learning_rate": 2e-05,
      "loss": 0.7104,
      "step": 2770
    },
    {
      "epoch": 0.23269440026784968,
      "grad_norm": 0.3845888376235962,
      "learning_rate": 2e-05,
      "loss": 0.7153,
      "step": 2780
    },
    {
      "epoch": 0.2335314304846405,
      "grad_norm": 0.3471842110157013,
      "learning_rate": 2e-05,
      "loss": 0.6582,
      "step": 2790
    },
    {
      "epoch": 0.23436846070143133,
      "grad_norm": 0.3849339485168457,
      "learning_rate": 2e-05,
      "loss": 0.7134,
      "step": 2800
    },
    {
      "epoch": 0.23520549091822215,
      "grad_norm": 0.4041845500469208,
      "learning_rate": 2e-05,
      "loss": 0.6917,
      "step": 2810
    },
    {
      "epoch": 0.23604252113501298,
      "grad_norm": 0.347912460565567,
      "learning_rate": 2e-05,
      "loss": 0.7065,
      "step": 2820
    },
    {
      "epoch": 0.2368795513518038,
      "grad_norm": 0.383291631937027,
      "learning_rate": 2e-05,
      "loss": 0.7237,
      "step": 2830
    },
    {
      "epoch": 0.23771658156859463,
      "grad_norm": 0.3993476927280426,
      "learning_rate": 2e-05,
      "loss": 0.7263,
      "step": 2840
    },
    {
      "epoch": 0.23855361178538545,
      "grad_norm": 0.364463210105896,
      "learning_rate": 2e-05,
      "loss": 0.6825,
      "step": 2850
    },
    {
      "epoch": 0.23939064200217627,
      "grad_norm": 0.3854297697544098,
      "learning_rate": 2e-05,
      "loss": 0.6828,
      "step": 2860
    },
    {
      "epoch": 0.2402276722189671,
      "grad_norm": 0.4553110599517822,
      "learning_rate": 2e-05,
      "loss": 0.6841,
      "step": 2870
    },
    {
      "epoch": 0.24106470243575792,
      "grad_norm": 0.3404019773006439,
      "learning_rate": 2e-05,
      "loss": 0.6824,
      "step": 2880
    },
    {
      "epoch": 0.24190173265254875,
      "grad_norm": 0.3552730679512024,
      "learning_rate": 2e-05,
      "loss": 0.705,
      "step": 2890
    },
    {
      "epoch": 0.24273876286933957,
      "grad_norm": 0.6009185910224915,
      "learning_rate": 2e-05,
      "loss": 0.6872,
      "step": 2900
    },
    {
      "epoch": 0.24357579308613042,
      "grad_norm": 0.3935545086860657,
      "learning_rate": 2e-05,
      "loss": 0.6606,
      "step": 2910
    },
    {
      "epoch": 0.24441282330292124,
      "grad_norm": 0.3839946389198303,
      "learning_rate": 2e-05,
      "loss": 0.6758,
      "step": 2920
    },
    {
      "epoch": 0.24524985351971207,
      "grad_norm": 0.49954575300216675,
      "learning_rate": 2e-05,
      "loss": 0.6747,
      "step": 2930
    },
    {
      "epoch": 0.2460868837365029,
      "grad_norm": 0.3401426672935486,
      "learning_rate": 2e-05,
      "loss": 0.6798,
      "step": 2940
    },
    {
      "epoch": 0.24692391395329372,
      "grad_norm": 0.38917040824890137,
      "learning_rate": 2e-05,
      "loss": 0.6955,
      "step": 2950
    },
    {
      "epoch": 0.24776094417008454,
      "grad_norm": 0.3529467284679413,
      "learning_rate": 2e-05,
      "loss": 0.6733,
      "step": 2960
    },
    {
      "epoch": 0.24859797438687536,
      "grad_norm": 0.3835350275039673,
      "learning_rate": 2e-05,
      "loss": 0.7163,
      "step": 2970
    },
    {
      "epoch": 0.2494350046036662,
      "grad_norm": 0.34143391251564026,
      "learning_rate": 2e-05,
      "loss": 0.7019,
      "step": 2980
    },
    {
      "epoch": 0.250272034820457,
      "grad_norm": 0.39724600315093994,
      "learning_rate": 2e-05,
      "loss": 0.6812,
      "step": 2990
    },
    {
      "epoch": 0.25110906503724784,
      "grad_norm": 0.3517789840698242,
      "learning_rate": 2e-05,
      "loss": 0.6866,
      "step": 3000
    },
    {
      "epoch": 0.25194609525403866,
      "grad_norm": 0.457095742225647,
      "learning_rate": 2e-05,
      "loss": 0.6892,
      "step": 3010
    },
    {
      "epoch": 0.2527831254708295,
      "grad_norm": 0.3822527825832367,
      "learning_rate": 2e-05,
      "loss": 0.7063,
      "step": 3020
    },
    {
      "epoch": 0.2536201556876203,
      "grad_norm": 0.39527907967567444,
      "learning_rate": 2e-05,
      "loss": 0.6757,
      "step": 3030
    },
    {
      "epoch": 0.25445718590441113,
      "grad_norm": 0.36703014373779297,
      "learning_rate": 2e-05,
      "loss": 0.6991,
      "step": 3040
    },
    {
      "epoch": 0.25529421612120196,
      "grad_norm": 0.5913490056991577,
      "learning_rate": 2e-05,
      "loss": 0.6808,
      "step": 3050
    },
    {
      "epoch": 0.2561312463379928,
      "grad_norm": 0.6630072593688965,
      "learning_rate": 2e-05,
      "loss": 0.7075,
      "step": 3060
    },
    {
      "epoch": 0.2569682765547836,
      "grad_norm": 0.3538056015968323,
      "learning_rate": 2e-05,
      "loss": 0.7061,
      "step": 3070
    },
    {
      "epoch": 0.2578053067715744,
      "grad_norm": 0.4272959530353546,
      "learning_rate": 2e-05,
      "loss": 0.6843,
      "step": 3080
    },
    {
      "epoch": 0.2586423369883653,
      "grad_norm": 0.396391898393631,
      "learning_rate": 2e-05,
      "loss": 0.6751,
      "step": 3090
    },
    {
      "epoch": 0.25947936720515613,
      "grad_norm": 0.3552750051021576,
      "learning_rate": 2e-05,
      "loss": 0.7147,
      "step": 3100
    },
    {
      "epoch": 0.26031639742194695,
      "grad_norm": 0.3866838216781616,
      "learning_rate": 2e-05,
      "loss": 0.736,
      "step": 3110
    },
    {
      "epoch": 0.2611534276387378,
      "grad_norm": 0.37459060549736023,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 3120
    },
    {
      "epoch": 0.2619904578555286,
      "grad_norm": 0.3690750300884247,
      "learning_rate": 2e-05,
      "loss": 0.6834,
      "step": 3130
    },
    {
      "epoch": 0.2628274880723194,
      "grad_norm": 0.4790518879890442,
      "learning_rate": 2e-05,
      "loss": 0.6947,
      "step": 3140
    },
    {
      "epoch": 0.26366451828911025,
      "grad_norm": 0.3825061023235321,
      "learning_rate": 2e-05,
      "loss": 0.7086,
      "step": 3150
    },
    {
      "epoch": 0.2645015485059011,
      "grad_norm": 0.43931442499160767,
      "learning_rate": 2e-05,
      "loss": 0.7131,
      "step": 3160
    },
    {
      "epoch": 0.2653385787226919,
      "grad_norm": 0.36091071367263794,
      "learning_rate": 2e-05,
      "loss": 0.6678,
      "step": 3170
    },
    {
      "epoch": 0.2661756089394827,
      "grad_norm": 0.35754019021987915,
      "learning_rate": 2e-05,
      "loss": 0.6757,
      "step": 3180
    },
    {
      "epoch": 0.26701263915627355,
      "grad_norm": 0.34211859107017517,
      "learning_rate": 2e-05,
      "loss": 0.6844,
      "step": 3190
    },
    {
      "epoch": 0.26784966937306437,
      "grad_norm": 0.3811708390712738,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 3200
    },
    {
      "epoch": 0.2686866995898552,
      "grad_norm": 0.37679523229599,
      "learning_rate": 2e-05,
      "loss": 0.6768,
      "step": 3210
    },
    {
      "epoch": 0.269523729806646,
      "grad_norm": 0.3931836187839508,
      "learning_rate": 2e-05,
      "loss": 0.7104,
      "step": 3220
    },
    {
      "epoch": 0.27036076002343684,
      "grad_norm": 0.4057767689228058,
      "learning_rate": 2e-05,
      "loss": 0.6791,
      "step": 3230
    },
    {
      "epoch": 0.27119779024022767,
      "grad_norm": 0.33529049158096313,
      "learning_rate": 2e-05,
      "loss": 0.6784,
      "step": 3240
    },
    {
      "epoch": 0.2720348204570185,
      "grad_norm": 0.34509772062301636,
      "learning_rate": 2e-05,
      "loss": 0.7474,
      "step": 3250
    },
    {
      "epoch": 0.2728718506738093,
      "grad_norm": 0.3847754895687103,
      "learning_rate": 2e-05,
      "loss": 0.6765,
      "step": 3260
    },
    {
      "epoch": 0.27370888089060014,
      "grad_norm": 0.38386350870132446,
      "learning_rate": 2e-05,
      "loss": 0.6782,
      "step": 3270
    },
    {
      "epoch": 0.27454591110739096,
      "grad_norm": 0.4805894196033478,
      "learning_rate": 2e-05,
      "loss": 0.6864,
      "step": 3280
    },
    {
      "epoch": 0.2753829413241818,
      "grad_norm": 0.46428972482681274,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 3290
    },
    {
      "epoch": 0.2762199715409726,
      "grad_norm": 0.3913302719593048,
      "learning_rate": 2e-05,
      "loss": 0.7239,
      "step": 3300
    },
    {
      "epoch": 0.27705700175776343,
      "grad_norm": 0.3995335102081299,
      "learning_rate": 2e-05,
      "loss": 0.6828,
      "step": 3310
    },
    {
      "epoch": 0.27789403197455426,
      "grad_norm": 0.36607876420021057,
      "learning_rate": 2e-05,
      "loss": 0.6621,
      "step": 3320
    },
    {
      "epoch": 0.2787310621913451,
      "grad_norm": 0.38472291827201843,
      "learning_rate": 2e-05,
      "loss": 0.6712,
      "step": 3330
    },
    {
      "epoch": 0.27956809240813596,
      "grad_norm": 0.34564754366874695,
      "learning_rate": 2e-05,
      "loss": 0.6863,
      "step": 3340
    },
    {
      "epoch": 0.2804051226249268,
      "grad_norm": 0.3804911673069,
      "learning_rate": 2e-05,
      "loss": 0.7107,
      "step": 3350
    },
    {
      "epoch": 0.2812421528417176,
      "grad_norm": 0.3177351653575897,
      "learning_rate": 2e-05,
      "loss": 0.6595,
      "step": 3360
    },
    {
      "epoch": 0.28207918305850843,
      "grad_norm": 0.3490181863307953,
      "learning_rate": 2e-05,
      "loss": 0.6671,
      "step": 3370
    },
    {
      "epoch": 0.28291621327529926,
      "grad_norm": 0.5247802734375,
      "learning_rate": 2e-05,
      "loss": 0.6614,
      "step": 3380
    },
    {
      "epoch": 0.2837532434920901,
      "grad_norm": 0.36323365569114685,
      "learning_rate": 2e-05,
      "loss": 0.684,
      "step": 3390
    },
    {
      "epoch": 0.2845902737088809,
      "grad_norm": 0.359318345785141,
      "learning_rate": 2e-05,
      "loss": 0.661,
      "step": 3400
    },
    {
      "epoch": 0.2854273039256717,
      "grad_norm": 0.3711358308792114,
      "learning_rate": 2e-05,
      "loss": 0.6748,
      "step": 3410
    },
    {
      "epoch": 0.28626433414246255,
      "grad_norm": 0.36873385310173035,
      "learning_rate": 2e-05,
      "loss": 0.6655,
      "step": 3420
    },
    {
      "epoch": 0.2871013643592534,
      "grad_norm": 0.398560494184494,
      "learning_rate": 2e-05,
      "loss": 0.6624,
      "step": 3430
    },
    {
      "epoch": 0.2879383945760442,
      "grad_norm": 0.48586469888687134,
      "learning_rate": 2e-05,
      "loss": 0.6993,
      "step": 3440
    },
    {
      "epoch": 0.288775424792835,
      "grad_norm": 0.3617458641529083,
      "learning_rate": 2e-05,
      "loss": 0.6928,
      "step": 3450
    },
    {
      "epoch": 0.28961245500962585,
      "grad_norm": 0.3452693521976471,
      "learning_rate": 2e-05,
      "loss": 0.6965,
      "step": 3460
    },
    {
      "epoch": 0.29044948522641667,
      "grad_norm": 0.39020127058029175,
      "learning_rate": 2e-05,
      "loss": 0.691,
      "step": 3470
    },
    {
      "epoch": 0.2912865154432075,
      "grad_norm": 0.38490548729896545,
      "learning_rate": 2e-05,
      "loss": 0.6824,
      "step": 3480
    },
    {
      "epoch": 0.2921235456599983,
      "grad_norm": 0.3624000549316406,
      "learning_rate": 2e-05,
      "loss": 0.6849,
      "step": 3490
    },
    {
      "epoch": 0.29296057587678914,
      "grad_norm": 0.40137097239494324,
      "learning_rate": 2e-05,
      "loss": 0.6779,
      "step": 3500
    },
    {
      "epoch": 0.29379760609357997,
      "grad_norm": 0.3723010718822479,
      "learning_rate": 2e-05,
      "loss": 0.6639,
      "step": 3510
    },
    {
      "epoch": 0.2946346363103708,
      "grad_norm": 0.33410966396331787,
      "learning_rate": 2e-05,
      "loss": 0.6849,
      "step": 3520
    },
    {
      "epoch": 0.2954716665271616,
      "grad_norm": 0.4087796211242676,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 3530
    },
    {
      "epoch": 0.29630869674395244,
      "grad_norm": 0.3803013265132904,
      "learning_rate": 2e-05,
      "loss": 0.6705,
      "step": 3540
    },
    {
      "epoch": 0.29714572696074326,
      "grad_norm": 0.38683268427848816,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 3550
    },
    {
      "epoch": 0.2979827571775341,
      "grad_norm": 0.3987625241279602,
      "learning_rate": 2e-05,
      "loss": 0.693,
      "step": 3560
    },
    {
      "epoch": 0.2988197873943249,
      "grad_norm": 0.3848333954811096,
      "learning_rate": 2e-05,
      "loss": 0.691,
      "step": 3570
    },
    {
      "epoch": 0.2996568176111158,
      "grad_norm": 0.3905976116657257,
      "learning_rate": 2e-05,
      "loss": 0.6745,
      "step": 3580
    },
    {
      "epoch": 0.3004938478279066,
      "grad_norm": 0.36275070905685425,
      "learning_rate": 2e-05,
      "loss": 0.6693,
      "step": 3590
    },
    {
      "epoch": 0.30133087804469744,
      "grad_norm": 0.3495936691761017,
      "learning_rate": 2e-05,
      "loss": 0.6845,
      "step": 3600
    },
    {
      "epoch": 0.30216790826148826,
      "grad_norm": 0.36721131205558777,
      "learning_rate": 2e-05,
      "loss": 0.6807,
      "step": 3610
    },
    {
      "epoch": 0.3030049384782791,
      "grad_norm": 0.42015427350997925,
      "learning_rate": 2e-05,
      "loss": 0.6934,
      "step": 3620
    },
    {
      "epoch": 0.3038419686950699,
      "grad_norm": 0.41565388441085815,
      "learning_rate": 2e-05,
      "loss": 0.6817,
      "step": 3630
    },
    {
      "epoch": 0.30467899891186073,
      "grad_norm": 0.3864046633243561,
      "learning_rate": 2e-05,
      "loss": 0.6972,
      "step": 3640
    },
    {
      "epoch": 0.30551602912865156,
      "grad_norm": 0.3387700021266937,
      "learning_rate": 2e-05,
      "loss": 0.6753,
      "step": 3650
    },
    {
      "epoch": 0.3063530593454424,
      "grad_norm": 0.37594422698020935,
      "learning_rate": 2e-05,
      "loss": 0.6808,
      "step": 3660
    },
    {
      "epoch": 0.3071900895622332,
      "grad_norm": 0.3680058419704437,
      "learning_rate": 2e-05,
      "loss": 0.6924,
      "step": 3670
    },
    {
      "epoch": 0.30802711977902403,
      "grad_norm": 0.37362930178642273,
      "learning_rate": 2e-05,
      "loss": 0.6598,
      "step": 3680
    },
    {
      "epoch": 0.30886414999581485,
      "grad_norm": 0.3793889880180359,
      "learning_rate": 2e-05,
      "loss": 0.6754,
      "step": 3690
    },
    {
      "epoch": 0.3097011802126057,
      "grad_norm": 0.35503870248794556,
      "learning_rate": 2e-05,
      "loss": 0.6834,
      "step": 3700
    },
    {
      "epoch": 0.3105382104293965,
      "grad_norm": 0.3719038665294647,
      "learning_rate": 2e-05,
      "loss": 0.6721,
      "step": 3710
    },
    {
      "epoch": 0.3113752406461873,
      "grad_norm": 0.38428905606269836,
      "learning_rate": 2e-05,
      "loss": 0.6499,
      "step": 3720
    },
    {
      "epoch": 0.31221227086297815,
      "grad_norm": 0.374455988407135,
      "learning_rate": 2e-05,
      "loss": 0.657,
      "step": 3730
    },
    {
      "epoch": 0.31304930107976897,
      "grad_norm": 0.37055298686027527,
      "learning_rate": 2e-05,
      "loss": 0.691,
      "step": 3740
    },
    {
      "epoch": 0.3138863312965598,
      "grad_norm": 0.37087371945381165,
      "learning_rate": 2e-05,
      "loss": 0.6859,
      "step": 3750
    },
    {
      "epoch": 0.3147233615133506,
      "grad_norm": 0.44824519753456116,
      "learning_rate": 2e-05,
      "loss": 0.689,
      "step": 3760
    },
    {
      "epoch": 0.31556039173014144,
      "grad_norm": 0.40546661615371704,
      "learning_rate": 2e-05,
      "loss": 0.6715,
      "step": 3770
    },
    {
      "epoch": 0.31639742194693227,
      "grad_norm": 0.39258283376693726,
      "learning_rate": 2e-05,
      "loss": 0.6877,
      "step": 3780
    },
    {
      "epoch": 0.3172344521637231,
      "grad_norm": 0.36289212107658386,
      "learning_rate": 2e-05,
      "loss": 0.6668,
      "step": 3790
    },
    {
      "epoch": 0.3180714823805139,
      "grad_norm": 0.40301036834716797,
      "learning_rate": 2e-05,
      "loss": 0.6785,
      "step": 3800
    },
    {
      "epoch": 0.31890851259730474,
      "grad_norm": 0.3613869249820709,
      "learning_rate": 2e-05,
      "loss": 0.6792,
      "step": 3810
    },
    {
      "epoch": 0.31974554281409556,
      "grad_norm": 0.3650999069213867,
      "learning_rate": 2e-05,
      "loss": 0.6548,
      "step": 3820
    },
    {
      "epoch": 0.32058257303088644,
      "grad_norm": 0.38625913858413696,
      "learning_rate": 2e-05,
      "loss": 0.63,
      "step": 3830
    },
    {
      "epoch": 0.32141960324767727,
      "grad_norm": 0.35693198442459106,
      "learning_rate": 2e-05,
      "loss": 0.6974,
      "step": 3840
    },
    {
      "epoch": 0.3222566334644681,
      "grad_norm": 0.3826311230659485,
      "learning_rate": 2e-05,
      "loss": 0.6842,
      "step": 3850
    },
    {
      "epoch": 0.3230936636812589,
      "grad_norm": 0.3499303162097931,
      "learning_rate": 2e-05,
      "loss": 0.667,
      "step": 3860
    },
    {
      "epoch": 0.32393069389804974,
      "grad_norm": 0.33624884486198425,
      "learning_rate": 2e-05,
      "loss": 0.6643,
      "step": 3870
    },
    {
      "epoch": 0.32476772411484056,
      "grad_norm": 0.4605383574962616,
      "learning_rate": 2e-05,
      "loss": 0.6631,
      "step": 3880
    },
    {
      "epoch": 0.3256047543316314,
      "grad_norm": 0.3703138828277588,
      "learning_rate": 2e-05,
      "loss": 0.7092,
      "step": 3890
    },
    {
      "epoch": 0.3264417845484222,
      "grad_norm": 0.3547705411911011,
      "learning_rate": 2e-05,
      "loss": 0.674,
      "step": 3900
    },
    {
      "epoch": 0.32727881476521303,
      "grad_norm": 0.38592010736465454,
      "learning_rate": 2e-05,
      "loss": 0.729,
      "step": 3910
    },
    {
      "epoch": 0.32811584498200386,
      "grad_norm": 0.38469675183296204,
      "learning_rate": 2e-05,
      "loss": 0.6379,
      "step": 3920
    },
    {
      "epoch": 0.3289528751987947,
      "grad_norm": 0.38590508699417114,
      "learning_rate": 2e-05,
      "loss": 0.6671,
      "step": 3930
    },
    {
      "epoch": 0.3297899054155855,
      "grad_norm": 0.4092082381248474,
      "learning_rate": 2e-05,
      "loss": 0.7011,
      "step": 3940
    },
    {
      "epoch": 0.33062693563237633,
      "grad_norm": 0.3793179988861084,
      "learning_rate": 2e-05,
      "loss": 0.6973,
      "step": 3950
    },
    {
      "epoch": 0.33146396584916715,
      "grad_norm": 0.38254857063293457,
      "learning_rate": 2e-05,
      "loss": 0.6909,
      "step": 3960
    },
    {
      "epoch": 0.332300996065958,
      "grad_norm": 0.3839905261993408,
      "learning_rate": 2e-05,
      "loss": 0.6891,
      "step": 3970
    },
    {
      "epoch": 0.3331380262827488,
      "grad_norm": 0.38743963837623596,
      "learning_rate": 2e-05,
      "loss": 0.7079,
      "step": 3980
    },
    {
      "epoch": 0.3339750564995396,
      "grad_norm": 0.381570965051651,
      "learning_rate": 2e-05,
      "loss": 0.7051,
      "step": 3990
    },
    {
      "epoch": 0.33481208671633045,
      "grad_norm": 0.3812105059623718,
      "learning_rate": 2e-05,
      "loss": 0.6781,
      "step": 4000
    },
    {
      "epoch": 0.33564911693312127,
      "grad_norm": 0.3621882200241089,
      "learning_rate": 2e-05,
      "loss": 0.6923,
      "step": 4010
    },
    {
      "epoch": 0.3364861471499121,
      "grad_norm": 0.35323214530944824,
      "learning_rate": 2e-05,
      "loss": 0.6851,
      "step": 4020
    },
    {
      "epoch": 0.3373231773667029,
      "grad_norm": 0.35751551389694214,
      "learning_rate": 2e-05,
      "loss": 0.6655,
      "step": 4030
    },
    {
      "epoch": 0.33816020758349374,
      "grad_norm": 0.3949991464614868,
      "learning_rate": 2e-05,
      "loss": 0.6621,
      "step": 4040
    },
    {
      "epoch": 0.33899723780028457,
      "grad_norm": 0.3553832173347473,
      "learning_rate": 2e-05,
      "loss": 0.7026,
      "step": 4050
    },
    {
      "epoch": 0.3398342680170754,
      "grad_norm": 0.39129525423049927,
      "learning_rate": 2e-05,
      "loss": 0.6727,
      "step": 4060
    },
    {
      "epoch": 0.3406712982338662,
      "grad_norm": 0.39109840989112854,
      "learning_rate": 2e-05,
      "loss": 0.692,
      "step": 4070
    },
    {
      "epoch": 0.3415083284506571,
      "grad_norm": 0.384739488363266,
      "learning_rate": 2e-05,
      "loss": 0.6578,
      "step": 4080
    },
    {
      "epoch": 0.3423453586674479,
      "grad_norm": 0.34750863909721375,
      "learning_rate": 2e-05,
      "loss": 0.6972,
      "step": 4090
    },
    {
      "epoch": 0.34318238888423874,
      "grad_norm": 0.3817720115184784,
      "learning_rate": 2e-05,
      "loss": 0.6563,
      "step": 4100
    },
    {
      "epoch": 0.34401941910102957,
      "grad_norm": 0.38964515924453735,
      "learning_rate": 2e-05,
      "loss": 0.6591,
      "step": 4110
    },
    {
      "epoch": 0.3448564493178204,
      "grad_norm": 0.38531047105789185,
      "learning_rate": 2e-05,
      "loss": 0.6475,
      "step": 4120
    },
    {
      "epoch": 0.3456934795346112,
      "grad_norm": 0.4181869626045227,
      "learning_rate": 2e-05,
      "loss": 0.6746,
      "step": 4130
    },
    {
      "epoch": 0.34653050975140204,
      "grad_norm": 0.36208441853523254,
      "learning_rate": 2e-05,
      "loss": 0.6825,
      "step": 4140
    },
    {
      "epoch": 0.34736753996819286,
      "grad_norm": 0.3654119074344635,
      "learning_rate": 2e-05,
      "loss": 0.6887,
      "step": 4150
    },
    {
      "epoch": 0.3482045701849837,
      "grad_norm": 0.39477288722991943,
      "learning_rate": 2e-05,
      "loss": 0.6977,
      "step": 4160
    },
    {
      "epoch": 0.3490416004017745,
      "grad_norm": 0.36870384216308594,
      "learning_rate": 2e-05,
      "loss": 0.6725,
      "step": 4170
    },
    {
      "epoch": 0.34987863061856533,
      "grad_norm": 0.35877320170402527,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 4180
    },
    {
      "epoch": 0.35071566083535616,
      "grad_norm": 0.5562852621078491,
      "learning_rate": 2e-05,
      "loss": 0.6439,
      "step": 4190
    },
    {
      "epoch": 0.351552691052147,
      "grad_norm": 0.40305373072624207,
      "learning_rate": 2e-05,
      "loss": 0.703,
      "step": 4200
    },
    {
      "epoch": 0.3523897212689378,
      "grad_norm": 0.41178417205810547,
      "learning_rate": 2e-05,
      "loss": 0.6618,
      "step": 4210
    },
    {
      "epoch": 0.35322675148572863,
      "grad_norm": 0.3854200541973114,
      "learning_rate": 2e-05,
      "loss": 0.6662,
      "step": 4220
    },
    {
      "epoch": 0.35406378170251945,
      "grad_norm": 0.3953651487827301,
      "learning_rate": 2e-05,
      "loss": 0.7133,
      "step": 4230
    },
    {
      "epoch": 0.3549008119193103,
      "grad_norm": 0.3878295123577118,
      "learning_rate": 2e-05,
      "loss": 0.701,
      "step": 4240
    },
    {
      "epoch": 0.3557378421361011,
      "grad_norm": 0.36245018243789673,
      "learning_rate": 2e-05,
      "loss": 0.6104,
      "step": 4250
    },
    {
      "epoch": 0.3565748723528919,
      "grad_norm": 0.3872743844985962,
      "learning_rate": 2e-05,
      "loss": 0.6878,
      "step": 4260
    },
    {
      "epoch": 0.35741190256968275,
      "grad_norm": 0.4376787543296814,
      "learning_rate": 2e-05,
      "loss": 0.6741,
      "step": 4270
    },
    {
      "epoch": 0.3582489327864736,
      "grad_norm": 0.5118808150291443,
      "learning_rate": 2e-05,
      "loss": 0.7126,
      "step": 4280
    },
    {
      "epoch": 0.3590859630032644,
      "grad_norm": 0.35995811223983765,
      "learning_rate": 2e-05,
      "loss": 0.6814,
      "step": 4290
    },
    {
      "epoch": 0.3599229932200552,
      "grad_norm": 0.32932382822036743,
      "learning_rate": 2e-05,
      "loss": 0.6686,
      "step": 4300
    },
    {
      "epoch": 0.36076002343684604,
      "grad_norm": 0.3927007019519806,
      "learning_rate": 2e-05,
      "loss": 0.6526,
      "step": 4310
    },
    {
      "epoch": 0.3615970536536369,
      "grad_norm": 0.3740251958370209,
      "learning_rate": 2e-05,
      "loss": 0.6658,
      "step": 4320
    },
    {
      "epoch": 0.36243408387042775,
      "grad_norm": 0.3808770477771759,
      "learning_rate": 2e-05,
      "loss": 0.6923,
      "step": 4330
    },
    {
      "epoch": 0.36327111408721857,
      "grad_norm": 0.3733738660812378,
      "learning_rate": 2e-05,
      "loss": 0.6758,
      "step": 4340
    },
    {
      "epoch": 0.3641081443040094,
      "grad_norm": 0.36265361309051514,
      "learning_rate": 2e-05,
      "loss": 0.6714,
      "step": 4350
    },
    {
      "epoch": 0.3649451745208002,
      "grad_norm": 0.37637802958488464,
      "learning_rate": 2e-05,
      "loss": 0.7047,
      "step": 4360
    },
    {
      "epoch": 0.36578220473759104,
      "grad_norm": 0.3615052402019501,
      "learning_rate": 2e-05,
      "loss": 0.6887,
      "step": 4370
    },
    {
      "epoch": 0.36661923495438187,
      "grad_norm": 0.3507406413555145,
      "learning_rate": 2e-05,
      "loss": 0.6817,
      "step": 4380
    },
    {
      "epoch": 0.3674562651711727,
      "grad_norm": 0.3859863579273224,
      "learning_rate": 2e-05,
      "loss": 0.6908,
      "step": 4390
    },
    {
      "epoch": 0.3682932953879635,
      "grad_norm": 0.5150337815284729,
      "learning_rate": 2e-05,
      "loss": 0.6829,
      "step": 4400
    },
    {
      "epoch": 0.36913032560475434,
      "grad_norm": 0.3702976107597351,
      "learning_rate": 2e-05,
      "loss": 0.6628,
      "step": 4410
    },
    {
      "epoch": 0.36996735582154516,
      "grad_norm": 0.36254793405532837,
      "learning_rate": 2e-05,
      "loss": 0.6735,
      "step": 4420
    },
    {
      "epoch": 0.370804386038336,
      "grad_norm": 0.3495381474494934,
      "learning_rate": 2e-05,
      "loss": 0.6815,
      "step": 4430
    },
    {
      "epoch": 0.3716414162551268,
      "grad_norm": 0.3943999707698822,
      "learning_rate": 2e-05,
      "loss": 0.6797,
      "step": 4440
    },
    {
      "epoch": 0.37247844647191763,
      "grad_norm": 0.3715941905975342,
      "learning_rate": 2e-05,
      "loss": 0.6615,
      "step": 4450
    },
    {
      "epoch": 0.37331547668870846,
      "grad_norm": 0.47644540667533875,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 4460
    },
    {
      "epoch": 0.3741525069054993,
      "grad_norm": 0.40302789211273193,
      "learning_rate": 2e-05,
      "loss": 0.6556,
      "step": 4470
    },
    {
      "epoch": 0.3749895371222901,
      "grad_norm": 0.366240531206131,
      "learning_rate": 2e-05,
      "loss": 0.6787,
      "step": 4480
    },
    {
      "epoch": 0.37582656733908093,
      "grad_norm": 0.3775447607040405,
      "learning_rate": 2e-05,
      "loss": 0.6518,
      "step": 4490
    },
    {
      "epoch": 0.37666359755587175,
      "grad_norm": 0.3555198609828949,
      "learning_rate": 2e-05,
      "loss": 0.6844,
      "step": 4500
    },
    {
      "epoch": 0.3775006277726626,
      "grad_norm": 0.40677034854888916,
      "learning_rate": 2e-05,
      "loss": 0.6899,
      "step": 4510
    },
    {
      "epoch": 0.3783376579894534,
      "grad_norm": 0.33568084239959717,
      "learning_rate": 2e-05,
      "loss": 0.6329,
      "step": 4520
    },
    {
      "epoch": 0.3791746882062442,
      "grad_norm": 0.4298158586025238,
      "learning_rate": 2e-05,
      "loss": 0.6856,
      "step": 4530
    },
    {
      "epoch": 0.38001171842303505,
      "grad_norm": 0.3619917333126068,
      "learning_rate": 2e-05,
      "loss": 0.655,
      "step": 4540
    },
    {
      "epoch": 0.3808487486398259,
      "grad_norm": 0.34173473715782166,
      "learning_rate": 2e-05,
      "loss": 0.6386,
      "step": 4550
    },
    {
      "epoch": 0.3816857788566167,
      "grad_norm": 0.41530799865722656,
      "learning_rate": 2e-05,
      "loss": 0.6436,
      "step": 4560
    },
    {
      "epoch": 0.3825228090734076,
      "grad_norm": 0.3594314455986023,
      "learning_rate": 2e-05,
      "loss": 0.6765,
      "step": 4570
    },
    {
      "epoch": 0.3833598392901984,
      "grad_norm": 0.3838607370853424,
      "learning_rate": 2e-05,
      "loss": 0.656,
      "step": 4580
    },
    {
      "epoch": 0.3841968695069892,
      "grad_norm": 0.3307185769081116,
      "learning_rate": 2e-05,
      "loss": 0.6976,
      "step": 4590
    },
    {
      "epoch": 0.38503389972378005,
      "grad_norm": 0.4034730792045593,
      "learning_rate": 2e-05,
      "loss": 0.6824,
      "step": 4600
    },
    {
      "epoch": 0.3858709299405709,
      "grad_norm": 0.38577455282211304,
      "learning_rate": 2e-05,
      "loss": 0.6516,
      "step": 4610
    },
    {
      "epoch": 0.3867079601573617,
      "grad_norm": 0.3357555568218231,
      "learning_rate": 2e-05,
      "loss": 0.645,
      "step": 4620
    },
    {
      "epoch": 0.3875449903741525,
      "grad_norm": 0.37598440051078796,
      "learning_rate": 2e-05,
      "loss": 0.7059,
      "step": 4630
    },
    {
      "epoch": 0.38838202059094334,
      "grad_norm": 0.3609393239021301,
      "learning_rate": 2e-05,
      "loss": 0.669,
      "step": 4640
    },
    {
      "epoch": 0.38921905080773417,
      "grad_norm": 0.3757115304470062,
      "learning_rate": 2e-05,
      "loss": 0.6395,
      "step": 4650
    },
    {
      "epoch": 0.390056081024525,
      "grad_norm": 0.36719217896461487,
      "learning_rate": 2e-05,
      "loss": 0.6744,
      "step": 4660
    },
    {
      "epoch": 0.3908931112413158,
      "grad_norm": 0.36577510833740234,
      "learning_rate": 2e-05,
      "loss": 0.6851,
      "step": 4670
    },
    {
      "epoch": 0.39173014145810664,
      "grad_norm": 0.36640894412994385,
      "learning_rate": 2e-05,
      "loss": 0.6627,
      "step": 4680
    },
    {
      "epoch": 0.39256717167489746,
      "grad_norm": 0.3786374628543854,
      "learning_rate": 2e-05,
      "loss": 0.6888,
      "step": 4690
    },
    {
      "epoch": 0.3934042018916883,
      "grad_norm": 0.3693200349807739,
      "learning_rate": 2e-05,
      "loss": 0.6608,
      "step": 4700
    },
    {
      "epoch": 0.3942412321084791,
      "grad_norm": 0.35623887181282043,
      "learning_rate": 2e-05,
      "loss": 0.6913,
      "step": 4710
    },
    {
      "epoch": 0.39507826232526994,
      "grad_norm": 0.37266993522644043,
      "learning_rate": 2e-05,
      "loss": 0.6216,
      "step": 4720
    },
    {
      "epoch": 0.39591529254206076,
      "grad_norm": 0.48957139253616333,
      "learning_rate": 2e-05,
      "loss": 0.6577,
      "step": 4730
    },
    {
      "epoch": 0.3967523227588516,
      "grad_norm": 0.361542284488678,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 4740
    },
    {
      "epoch": 0.3975893529756424,
      "grad_norm": 0.3781006932258606,
      "learning_rate": 2e-05,
      "loss": 0.6729,
      "step": 4750
    },
    {
      "epoch": 0.39842638319243323,
      "grad_norm": 0.3931069076061249,
      "learning_rate": 2e-05,
      "loss": 0.6626,
      "step": 4760
    },
    {
      "epoch": 0.39926341340922405,
      "grad_norm": 0.3622207045555115,
      "learning_rate": 2e-05,
      "loss": 0.6719,
      "step": 4770
    },
    {
      "epoch": 0.4001004436260149,
      "grad_norm": 0.43951886892318726,
      "learning_rate": 2e-05,
      "loss": 0.6866,
      "step": 4780
    },
    {
      "epoch": 0.4009374738428057,
      "grad_norm": 0.36725106835365295,
      "learning_rate": 2e-05,
      "loss": 0.6412,
      "step": 4790
    },
    {
      "epoch": 0.4017745040595965,
      "grad_norm": 0.3468758761882782,
      "learning_rate": 2e-05,
      "loss": 0.656,
      "step": 4800
    },
    {
      "epoch": 0.40261153427638735,
      "grad_norm": 0.3753811717033386,
      "learning_rate": 2e-05,
      "loss": 0.674,
      "step": 4810
    },
    {
      "epoch": 0.40344856449317823,
      "grad_norm": 0.35826578736305237,
      "learning_rate": 2e-05,
      "loss": 0.6671,
      "step": 4820
    },
    {
      "epoch": 0.40428559470996905,
      "grad_norm": 0.5122092962265015,
      "learning_rate": 2e-05,
      "loss": 0.68,
      "step": 4830
    },
    {
      "epoch": 0.4051226249267599,
      "grad_norm": 0.34622722864151,
      "learning_rate": 2e-05,
      "loss": 0.6371,
      "step": 4840
    },
    {
      "epoch": 0.4059596551435507,
      "grad_norm": 0.354408323764801,
      "learning_rate": 2e-05,
      "loss": 0.6516,
      "step": 4850
    },
    {
      "epoch": 0.4067966853603415,
      "grad_norm": 0.3598283529281616,
      "learning_rate": 2e-05,
      "loss": 0.6357,
      "step": 4860
    },
    {
      "epoch": 0.40763371557713235,
      "grad_norm": 0.34254366159439087,
      "learning_rate": 2e-05,
      "loss": 0.6617,
      "step": 4870
    },
    {
      "epoch": 0.4084707457939232,
      "grad_norm": 0.4542747437953949,
      "learning_rate": 2e-05,
      "loss": 0.6532,
      "step": 4880
    },
    {
      "epoch": 0.409307776010714,
      "grad_norm": 0.35308054089546204,
      "learning_rate": 2e-05,
      "loss": 0.6558,
      "step": 4890
    },
    {
      "epoch": 0.4101448062275048,
      "grad_norm": 0.37054240703582764,
      "learning_rate": 2e-05,
      "loss": 0.6681,
      "step": 4900
    },
    {
      "epoch": 0.41098183644429565,
      "grad_norm": 0.3926886022090912,
      "learning_rate": 2e-05,
      "loss": 0.6739,
      "step": 4910
    },
    {
      "epoch": 0.41181886666108647,
      "grad_norm": 0.4271712601184845,
      "learning_rate": 2e-05,
      "loss": 0.655,
      "step": 4920
    },
    {
      "epoch": 0.4126558968778773,
      "grad_norm": 0.4078362286090851,
      "learning_rate": 2e-05,
      "loss": 0.5931,
      "step": 4930
    },
    {
      "epoch": 0.4134929270946681,
      "grad_norm": 0.3612181544303894,
      "learning_rate": 2e-05,
      "loss": 0.6747,
      "step": 4940
    },
    {
      "epoch": 0.41432995731145894,
      "grad_norm": 0.36205559968948364,
      "learning_rate": 2e-05,
      "loss": 0.6643,
      "step": 4950
    },
    {
      "epoch": 0.41516698752824976,
      "grad_norm": 0.33852431178092957,
      "learning_rate": 2e-05,
      "loss": 0.6516,
      "step": 4960
    },
    {
      "epoch": 0.4160040177450406,
      "grad_norm": 0.360599160194397,
      "learning_rate": 2e-05,
      "loss": 0.6424,
      "step": 4970
    },
    {
      "epoch": 0.4168410479618314,
      "grad_norm": 0.3367842733860016,
      "learning_rate": 2e-05,
      "loss": 0.6583,
      "step": 4980
    },
    {
      "epoch": 0.41767807817862224,
      "grad_norm": 0.37129393219947815,
      "learning_rate": 2e-05,
      "loss": 0.6406,
      "step": 4990
    },
    {
      "epoch": 0.41851510839541306,
      "grad_norm": 0.40336930751800537,
      "learning_rate": 2e-05,
      "loss": 0.6615,
      "step": 5000
    },
    {
      "epoch": 0.4193521386122039,
      "grad_norm": 0.3625902831554413,
      "learning_rate": 2e-05,
      "loss": 0.6726,
      "step": 5010
    },
    {
      "epoch": 0.4201891688289947,
      "grad_norm": 0.3696928322315216,
      "learning_rate": 2e-05,
      "loss": 0.6929,
      "step": 5020
    },
    {
      "epoch": 0.42102619904578553,
      "grad_norm": 0.35005995631217957,
      "learning_rate": 2e-05,
      "loss": 0.6734,
      "step": 5030
    },
    {
      "epoch": 0.42186322926257636,
      "grad_norm": 0.3266897201538086,
      "learning_rate": 2e-05,
      "loss": 0.6882,
      "step": 5040
    },
    {
      "epoch": 0.4227002594793672,
      "grad_norm": 0.4611261188983917,
      "learning_rate": 2e-05,
      "loss": 0.7007,
      "step": 5050
    },
    {
      "epoch": 0.423537289696158,
      "grad_norm": 0.39426982402801514,
      "learning_rate": 2e-05,
      "loss": 0.6638,
      "step": 5060
    },
    {
      "epoch": 0.4243743199129489,
      "grad_norm": 0.3529796302318573,
      "learning_rate": 2e-05,
      "loss": 0.6559,
      "step": 5070
    },
    {
      "epoch": 0.4252113501297397,
      "grad_norm": 0.39787614345550537,
      "learning_rate": 2e-05,
      "loss": 0.6387,
      "step": 5080
    },
    {
      "epoch": 0.42604838034653053,
      "grad_norm": 0.3493591547012329,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.5847,
      "step": 5090
    },
    {
      "epoch": 0.42688541056332135,
      "grad_norm": 0.3723395764827728,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.6512,
      "step": 5100
    },
    {
      "epoch": 0.4277224407801122,
      "grad_norm": 0.40137779712677,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.6725,
      "step": 5110
    },
    {
      "epoch": 0.428559470996903,
      "grad_norm": 0.3983379900455475,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.6721,
      "step": 5120
    },
    {
      "epoch": 0.4293965012136938,
      "grad_norm": 0.4244665801525116,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.6925,
      "step": 5130
    },
    {
      "epoch": 0.43023353143048465,
      "grad_norm": 0.39193373918533325,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.6566,
      "step": 5140
    },
    {
      "epoch": 0.4310705616472755,
      "grad_norm": 0.3795602321624756,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.653,
      "step": 5150
    },
    {
      "epoch": 0.4319075918640663,
      "grad_norm": 0.35238948464393616,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.6435,
      "step": 5160
    },
    {
      "epoch": 0.4327446220808571,
      "grad_norm": 0.3624531030654907,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.6457,
      "step": 5170
    },
    {
      "epoch": 0.43358165229764795,
      "grad_norm": 0.34817200899124146,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.6873,
      "step": 5180
    },
    {
      "epoch": 0.43441868251443877,
      "grad_norm": 0.4171640872955322,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.6788,
      "step": 5190
    },
    {
      "epoch": 0.4352557127312296,
      "grad_norm": 0.37277984619140625,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.6728,
      "step": 5200
    },
    {
      "epoch": 0.4360927429480204,
      "grad_norm": 0.38008707761764526,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.6409,
      "step": 5210
    },
    {
      "epoch": 0.43692977316481124,
      "grad_norm": 0.34851887822151184,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.6645,
      "step": 5220
    },
    {
      "epoch": 0.43776680338160207,
      "grad_norm": 0.3664775788784027,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.6616,
      "step": 5230
    },
    {
      "epoch": 0.4386038335983929,
      "grad_norm": 0.3701815605163574,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.6364,
      "step": 5240
    },
    {
      "epoch": 0.4394408638151837,
      "grad_norm": 0.362636923789978,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.6473,
      "step": 5250
    },
    {
      "epoch": 0.44027789403197454,
      "grad_norm": 0.342526376247406,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.6731,
      "step": 5260
    },
    {
      "epoch": 0.44111492424876536,
      "grad_norm": 0.32191959023475647,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.6668,
      "step": 5270
    },
    {
      "epoch": 0.4419519544655562,
      "grad_norm": 0.3483056128025055,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.6675,
      "step": 5280
    },
    {
      "epoch": 0.442788984682347,
      "grad_norm": 0.3659847378730774,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.6675,
      "step": 5290
    },
    {
      "epoch": 0.44362601489913783,
      "grad_norm": 0.36162468791007996,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.648,
      "step": 5300
    },
    {
      "epoch": 0.4444630451159287,
      "grad_norm": 0.36163952946662903,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.6463,
      "step": 5310
    },
    {
      "epoch": 0.44530007533271954,
      "grad_norm": 0.5795994997024536,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.6575,
      "step": 5320
    },
    {
      "epoch": 0.44613710554951036,
      "grad_norm": 0.35306987166404724,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.6597,
      "step": 5330
    },
    {
      "epoch": 0.4469741357663012,
      "grad_norm": 0.33831799030303955,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.6277,
      "step": 5340
    },
    {
      "epoch": 0.447811165983092,
      "grad_norm": 0.3335249423980713,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.6637,
      "step": 5350
    },
    {
      "epoch": 0.44864819619988283,
      "grad_norm": 0.3490908443927765,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.6529,
      "step": 5360
    },
    {
      "epoch": 0.44948522641667366,
      "grad_norm": 0.36650386452674866,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.6497,
      "step": 5370
    },
    {
      "epoch": 0.4503222566334645,
      "grad_norm": 0.346248596906662,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.6752,
      "step": 5380
    },
    {
      "epoch": 0.4511592868502553,
      "grad_norm": 0.34524908661842346,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.6394,
      "step": 5390
    },
    {
      "epoch": 0.4519963170670461,
      "grad_norm": 0.3573249578475952,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.6787,
      "step": 5400
    },
    {
      "epoch": 0.45283334728383695,
      "grad_norm": 0.35597580671310425,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.6971,
      "step": 5410
    },
    {
      "epoch": 0.4536703775006278,
      "grad_norm": 0.36199232935905457,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.6682,
      "step": 5420
    },
    {
      "epoch": 0.4545074077174186,
      "grad_norm": 0.393062025308609,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.6573,
      "step": 5430
    },
    {
      "epoch": 0.4553444379342094,
      "grad_norm": 0.3516090512275696,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.6321,
      "step": 5440
    },
    {
      "epoch": 0.45618146815100025,
      "grad_norm": 0.387460857629776,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.6756,
      "step": 5450
    },
    {
      "epoch": 0.45701849836779107,
      "grad_norm": 0.3646886348724365,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.6385,
      "step": 5460
    },
    {
      "epoch": 0.4578555285845819,
      "grad_norm": 0.33883002400398254,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.6826,
      "step": 5470
    },
    {
      "epoch": 0.4586925588013727,
      "grad_norm": 0.3687443137168884,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.6485,
      "step": 5480
    },
    {
      "epoch": 0.45952958901816354,
      "grad_norm": 0.36745738983154297,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.6697,
      "step": 5490
    },
    {
      "epoch": 0.46036661923495437,
      "grad_norm": 0.3611207604408264,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.6132,
      "step": 5500
    },
    {
      "epoch": 0.4612036494517452,
      "grad_norm": 0.34954649209976196,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.6436,
      "step": 5510
    },
    {
      "epoch": 0.462040679668536,
      "grad_norm": 0.3694668710231781,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.6035,
      "step": 5520
    },
    {
      "epoch": 0.46287770988532684,
      "grad_norm": 0.33643293380737305,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.616,
      "step": 5530
    },
    {
      "epoch": 0.46371474010211766,
      "grad_norm": 0.35816338658332825,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.622,
      "step": 5540
    },
    {
      "epoch": 0.4645517703189085,
      "grad_norm": 0.3566610515117645,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.6435,
      "step": 5550
    },
    {
      "epoch": 0.46538880053569937,
      "grad_norm": 0.32019272446632385,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.6681,
      "step": 5560
    },
    {
      "epoch": 0.4662258307524902,
      "grad_norm": 0.33671486377716064,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.6293,
      "step": 5570
    },
    {
      "epoch": 0.467062860969281,
      "grad_norm": 0.3729787766933441,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.6354,
      "step": 5580
    },
    {
      "epoch": 0.46789989118607184,
      "grad_norm": 0.3181573152542114,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.6365,
      "step": 5590
    },
    {
      "epoch": 0.4684858123378254,
      "step": 5597,
      "total_flos": 2.2737574718397743e+19,
      "train_loss": 0.02192150756293415,
      "train_runtime": 1760.7014,
      "train_samples_per_second": 38.146,
      "train_steps_per_second": 3.179
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 159,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2737574718397743e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}