{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.996815286624204,
  "eval_steps": 500,
  "global_step": 3138,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01592356687898089,
      "grad_norm": 5.4758992195129395,
      "learning_rate": 1.2738853503184714e-05,
      "loss": 0.9116,
      "step": 10
    },
    {
      "epoch": 0.03184713375796178,
      "grad_norm": 4.248476028442383,
      "learning_rate": 2.5477707006369428e-05,
      "loss": 0.5347,
      "step": 20
    },
    {
      "epoch": 0.04777070063694268,
      "grad_norm": 2.722592830657959,
      "learning_rate": 3.821656050955414e-05,
      "loss": 0.3604,
      "step": 30
    },
    {
      "epoch": 0.06369426751592357,
      "grad_norm": 1.718458890914917,
      "learning_rate": 5.0955414012738855e-05,
      "loss": 0.2924,
      "step": 40
    },
    {
      "epoch": 0.07961783439490445,
      "grad_norm": 2.0852134227752686,
      "learning_rate": 6.369426751592356e-05,
      "loss": 0.2293,
      "step": 50
    },
    {
      "epoch": 0.09554140127388536,
      "grad_norm": 2.2306277751922607,
      "learning_rate": 7.643312101910829e-05,
      "loss": 0.2296,
      "step": 60
    },
    {
      "epoch": 0.11146496815286625,
      "grad_norm": 1.173863172531128,
      "learning_rate": 8.9171974522293e-05,
      "loss": 0.1754,
      "step": 70
    },
    {
      "epoch": 0.12738853503184713,
      "grad_norm": 1.4193692207336426,
      "learning_rate": 0.00010191082802547771,
      "loss": 0.1821,
      "step": 80
    },
    {
      "epoch": 0.14331210191082802,
      "grad_norm": 1.5220022201538086,
      "learning_rate": 0.00011464968152866242,
      "loss": 0.1634,
      "step": 90
    },
    {
      "epoch": 0.1592356687898089,
      "grad_norm": 1.3423898220062256,
      "learning_rate": 0.00012738853503184712,
      "loss": 0.1462,
      "step": 100
    },
    {
      "epoch": 0.1751592356687898,
      "grad_norm": 1.2148971557617188,
      "learning_rate": 0.00014012738853503185,
      "loss": 0.1386,
      "step": 110
    },
    {
      "epoch": 0.1910828025477707,
      "grad_norm": 0.9744131565093994,
      "learning_rate": 0.00015286624203821657,
      "loss": 0.1376,
      "step": 120
    },
    {
      "epoch": 0.2070063694267516,
      "grad_norm": 1.014123558998108,
      "learning_rate": 0.0001656050955414013,
      "loss": 0.1082,
      "step": 130
    },
    {
      "epoch": 0.2229299363057325,
      "grad_norm": 1.3054927587509155,
      "learning_rate": 0.000178343949044586,
      "loss": 0.1054,
      "step": 140
    },
    {
      "epoch": 0.23885350318471338,
      "grad_norm": 1.789557695388794,
      "learning_rate": 0.00019108280254777072,
      "loss": 0.1308,
      "step": 150
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 0.6913818717002869,
      "learning_rate": 0.00019999950020955923,
      "loss": 0.1197,
      "step": 160
    },
    {
      "epoch": 0.27070063694267515,
      "grad_norm": 0.9011440277099609,
      "learning_rate": 0.00019999061518514538,
      "loss": 0.0938,
      "step": 170
    },
    {
      "epoch": 0.28662420382165604,
      "grad_norm": 1.6200815439224243,
      "learning_rate": 0.00019997062484234648,
      "loss": 0.108,
      "step": 180
    },
    {
      "epoch": 0.30254777070063693,
      "grad_norm": 0.8650676608085632,
      "learning_rate": 0.0001999395314013622,
      "loss": 0.105,
      "step": 190
    },
    {
      "epoch": 0.3184713375796178,
      "grad_norm": 1.1472347974777222,
      "learning_rate": 0.00019989733831554218,
      "loss": 0.1219,
      "step": 200
    },
    {
      "epoch": 0.3343949044585987,
      "grad_norm": 1.0066096782684326,
      "learning_rate": 0.00019984405027100277,
      "loss": 0.0927,
      "step": 210
    },
    {
      "epoch": 0.3503184713375796,
      "grad_norm": 1.0204339027404785,
      "learning_rate": 0.00019977967318610639,
      "loss": 0.1119,
      "step": 220
    },
    {
      "epoch": 0.3662420382165605,
      "grad_norm": 1.074305534362793,
      "learning_rate": 0.00019970421421080427,
      "loss": 0.0903,
      "step": 230
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 0.9550490379333496,
      "learning_rate": 0.0001996176817258424,
      "loss": 0.091,
      "step": 240
    },
    {
      "epoch": 0.3980891719745223,
      "grad_norm": 1.0449860095977783,
      "learning_rate": 0.0001995200853418307,
      "loss": 0.0861,
      "step": 250
    },
    {
      "epoch": 0.4140127388535032,
      "grad_norm": 0.9337477684020996,
      "learning_rate": 0.00019941143589817558,
      "loss": 0.0907,
      "step": 260
    },
    {
      "epoch": 0.4299363057324841,
      "grad_norm": 0.7787328958511353,
      "learning_rate": 0.00019929174546187616,
      "loss": 0.0816,
      "step": 270
    },
    {
      "epoch": 0.445859872611465,
      "grad_norm": 0.5837294459342957,
      "learning_rate": 0.000199161027326184,
      "loss": 0.085,
      "step": 280
    },
    {
      "epoch": 0.46178343949044587,
      "grad_norm": 0.6995041370391846,
      "learning_rate": 0.00019901929600912678,
      "loss": 0.0767,
      "step": 290
    },
    {
      "epoch": 0.47770700636942676,
      "grad_norm": 0.9348034262657166,
      "learning_rate": 0.00019886656725189575,
      "loss": 0.08,
      "step": 300
    },
    {
      "epoch": 0.49363057324840764,
      "grad_norm": 0.458731085062027,
      "learning_rate": 0.0001987028580170976,
      "loss": 0.0819,
      "step": 310
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 0.5243947505950928,
      "learning_rate": 0.00019852818648687044,
      "loss": 0.078,
      "step": 320
    },
    {
      "epoch": 0.5254777070063694,
      "grad_norm": 0.6848501563072205,
      "learning_rate": 0.00019834257206086442,
      "loss": 0.0891,
      "step": 330
    },
    {
      "epoch": 0.5414012738853503,
      "grad_norm": 0.9554067254066467,
      "learning_rate": 0.00019814603535408724,
      "loss": 0.0759,
      "step": 340
    },
    {
      "epoch": 0.5573248407643312,
      "grad_norm": 0.49273306131362915,
      "learning_rate": 0.0001979385981946144,
      "loss": 0.0744,
      "step": 350
    },
    {
      "epoch": 0.5732484076433121,
      "grad_norm": 0.7968289852142334,
      "learning_rate": 0.00019772028362116513,
      "loss": 0.0674,
      "step": 360
    },
    {
      "epoch": 0.589171974522293,
      "grad_norm": 0.5523058176040649,
      "learning_rate": 0.00019749111588054333,
      "loss": 0.078,
      "step": 370
    },
    {
      "epoch": 0.6050955414012739,
      "grad_norm": 0.7879081964492798,
      "learning_rate": 0.00019725112042494493,
      "loss": 0.0727,
      "step": 380
    },
    {
      "epoch": 0.6210191082802548,
      "grad_norm": 0.897438645362854,
      "learning_rate": 0.00019700032390913087,
      "loss": 0.0882,
      "step": 390
    },
    {
      "epoch": 0.6369426751592356,
      "grad_norm": 0.40805375576019287,
      "learning_rate": 0.00019673875418746678,
      "loss": 0.0687,
      "step": 400
    },
    {
      "epoch": 0.6528662420382165,
      "grad_norm": 0.8794530034065247,
      "learning_rate": 0.00019646644031082948,
      "loss": 0.0745,
      "step": 410
    },
    {
      "epoch": 0.6687898089171974,
      "grad_norm": 0.671391487121582,
      "learning_rate": 0.00019618341252338026,
      "loss": 0.0773,
      "step": 420
    },
    {
      "epoch": 0.6847133757961783,
      "grad_norm": 0.7118439078330994,
      "learning_rate": 0.00019588970225920612,
      "loss": 0.0646,
      "step": 430
    },
    {
      "epoch": 0.7006369426751592,
      "grad_norm": 0.8737181425094604,
      "learning_rate": 0.00019558534213882838,
      "loss": 0.0667,
      "step": 440
    },
    {
      "epoch": 0.7165605095541401,
      "grad_norm": 0.9923394322395325,
      "learning_rate": 0.00019527036596557983,
      "loss": 0.0699,
      "step": 450
    },
    {
      "epoch": 0.732484076433121,
      "grad_norm": 0.8922154307365417,
      "learning_rate": 0.0001949448087218504,
      "loss": 0.062,
      "step": 460
    },
    {
      "epoch": 0.7484076433121019,
      "grad_norm": 0.6825265884399414,
      "learning_rate": 0.0001946087065652018,
      "loss": 0.0672,
      "step": 470
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 0.6685843467712402,
      "learning_rate": 0.0001942620968243519,
      "loss": 0.057,
      "step": 480
    },
    {
      "epoch": 0.7802547770700637,
      "grad_norm": 0.5671578049659729,
      "learning_rate": 0.00019390501799502872,
      "loss": 0.0543,
      "step": 490
    },
    {
      "epoch": 0.7961783439490446,
      "grad_norm": 0.5407360196113586,
      "learning_rate": 0.00019353750973569498,
      "loss": 0.0526,
      "step": 500
    },
    {
      "epoch": 0.8121019108280255,
      "grad_norm": 0.45758092403411865,
      "learning_rate": 0.0001931596128631436,
      "loss": 0.0523,
      "step": 510
    },
    {
      "epoch": 0.8280254777070064,
      "grad_norm": 0.5699948668479919,
      "learning_rate": 0.0001927713693479643,
      "loss": 0.0508,
      "step": 520
    },
    {
      "epoch": 0.8439490445859873,
      "grad_norm": 0.6552209258079529,
      "learning_rate": 0.00019237282230988222,
      "loss": 0.0736,
      "step": 530
    },
    {
      "epoch": 0.8598726114649682,
      "grad_norm": 0.6645591855049133,
      "learning_rate": 0.00019196401601296898,
      "loss": 0.067,
      "step": 540
    },
    {
      "epoch": 0.8757961783439491,
      "grad_norm": 0.5558132529258728,
      "learning_rate": 0.00019154499586072646,
      "loss": 0.0562,
      "step": 550
    },
    {
      "epoch": 0.89171974522293,
      "grad_norm": 0.42415371537208557,
      "learning_rate": 0.00019111580839104418,
      "loss": 0.064,
      "step": 560
    },
    {
      "epoch": 0.9076433121019108,
      "grad_norm": 0.6379917860031128,
      "learning_rate": 0.00019067650127103055,
      "loss": 0.0628,
      "step": 570
    },
    {
      "epoch": 0.9235668789808917,
      "grad_norm": 0.603799045085907,
      "learning_rate": 0.0001902271232917189,
      "loss": 0.0668,
      "step": 580
    },
    {
      "epoch": 0.9394904458598726,
      "grad_norm": 0.569402813911438,
      "learning_rate": 0.0001897677243626485,
      "loss": 0.0613,
      "step": 590
    },
    {
      "epoch": 0.9554140127388535,
      "grad_norm": 0.7091385722160339,
      "learning_rate": 0.00018929835550632145,
      "loss": 0.0665,
      "step": 600
    },
    {
      "epoch": 0.9713375796178344,
      "grad_norm": 0.5904159545898438,
      "learning_rate": 0.00018881906885253593,
      "loss": 0.0766,
      "step": 610
    },
    {
      "epoch": 0.9872611464968153,
      "grad_norm": 0.6689817905426025,
      "learning_rate": 0.00018832991763259647,
      "loss": 0.065,
      "step": 620
    },
    {
      "epoch": 1.0031847133757963,
      "grad_norm": 0.7255253791809082,
      "learning_rate": 0.00018783095617340193,
      "loss": 0.053,
      "step": 630
    },
    {
      "epoch": 1.019108280254777,
      "grad_norm": 0.41102227568626404,
      "learning_rate": 0.0001873222398914117,
      "loss": 0.0584,
      "step": 640
    },
    {
      "epoch": 1.035031847133758,
      "grad_norm": 0.5301253199577332,
      "learning_rate": 0.00018680382528649093,
      "loss": 0.0619,
      "step": 650
    },
    {
      "epoch": 1.0509554140127388,
      "grad_norm": 0.5525597929954529,
      "learning_rate": 0.00018627576993563563,
      "loss": 0.0556,
      "step": 660
    },
    {
      "epoch": 1.0668789808917198,
      "grad_norm": 0.5564507246017456,
      "learning_rate": 0.0001857381324865777,
      "loss": 0.0428,
      "step": 670
    },
    {
      "epoch": 1.0828025477707006,
      "grad_norm": 0.5977344512939453,
      "learning_rate": 0.0001851909726512714,
      "loss": 0.0576,
      "step": 680
    },
    {
      "epoch": 1.0987261146496816,
      "grad_norm": 0.49446621537208557,
      "learning_rate": 0.00018463435119926176,
      "loss": 0.0527,
      "step": 690
    },
    {
      "epoch": 1.1146496815286624,
      "grad_norm": 0.4952321946620941,
      "learning_rate": 0.00018406832995093486,
      "loss": 0.0472,
      "step": 700
    },
    {
      "epoch": 1.1305732484076434,
      "grad_norm": 0.40190061926841736,
      "learning_rate": 0.0001834929717706522,
      "loss": 0.0557,
      "step": 710
    },
    {
      "epoch": 1.1464968152866242,
      "grad_norm": 0.44065698981285095,
      "learning_rate": 0.00018290834055976855,
      "loss": 0.0473,
      "step": 720
    },
    {
      "epoch": 1.1624203821656052,
      "grad_norm": 0.5594244003295898,
      "learning_rate": 0.00018231450124953495,
      "loss": 0.0588,
      "step": 730
    },
    {
      "epoch": 1.178343949044586,
      "grad_norm": 0.49133527278900146,
      "learning_rate": 0.00018171151979388714,
      "loss": 0.0497,
      "step": 740
    },
    {
      "epoch": 1.194267515923567,
      "grad_norm": 0.5907294750213623,
      "learning_rate": 0.00018109946316212052,
      "loss": 0.0525,
      "step": 750
    },
    {
      "epoch": 1.2101910828025477,
      "grad_norm": 0.5373413562774658,
      "learning_rate": 0.00018047839933145232,
      "loss": 0.0553,
      "step": 760
    },
    {
      "epoch": 1.2261146496815287,
      "grad_norm": 0.5720402598381042,
      "learning_rate": 0.00017984839727947164,
      "loss": 0.0505,
      "step": 770
    },
    {
      "epoch": 1.2420382165605095,
      "grad_norm": 0.3708595037460327,
      "learning_rate": 0.00017920952697647886,
      "loss": 0.0544,
      "step": 780
    },
    {
      "epoch": 1.2579617834394905,
      "grad_norm": 0.48515430092811584,
      "learning_rate": 0.00017856185937771416,
      "loss": 0.041,
      "step": 790
    },
    {
      "epoch": 1.2738853503184713,
      "grad_norm": 0.6001298427581787,
      "learning_rate": 0.00017790546641547722,
      "loss": 0.0478,
      "step": 800
    },
    {
      "epoch": 1.2898089171974523,
      "grad_norm": 0.4815957844257355,
      "learning_rate": 0.00017724042099113792,
      "loss": 0.0439,
      "step": 810
    },
    {
      "epoch": 1.305732484076433,
      "grad_norm": 0.6788195371627808,
      "learning_rate": 0.00017656679696703996,
      "loss": 0.0573,
      "step": 820
    },
    {
      "epoch": 1.321656050955414,
      "grad_norm": 0.3975256383419037,
      "learning_rate": 0.0001758846691582972,
      "loss": 0.0472,
      "step": 830
    },
    {
      "epoch": 1.3375796178343948,
      "grad_norm": 0.4403384327888489,
      "learning_rate": 0.00017519411332448444,
      "loss": 0.0445,
      "step": 840
    },
    {
      "epoch": 1.3535031847133758,
      "grad_norm": 0.39884060621261597,
      "learning_rate": 0.00017449520616122344,
      "loss": 0.0473,
      "step": 850
    },
    {
      "epoch": 1.3694267515923566,
      "grad_norm": 0.41912099719047546,
      "learning_rate": 0.00017378802529166462,
      "loss": 0.0529,
      "step": 860
    },
    {
      "epoch": 1.3853503184713376,
      "grad_norm": 0.43497881293296814,
      "learning_rate": 0.00017307264925786622,
      "loss": 0.0462,
      "step": 870
    },
    {
      "epoch": 1.4012738853503186,
      "grad_norm": 0.6406415700912476,
      "learning_rate": 0.0001723491575120708,
      "loss": 0.0515,
      "step": 880
    },
    {
      "epoch": 1.4171974522292994,
      "grad_norm": 0.5098315477371216,
      "learning_rate": 0.0001716176304078813,
      "loss": 0.0537,
      "step": 890
    },
    {
      "epoch": 1.4331210191082802,
      "grad_norm": 0.39417779445648193,
      "learning_rate": 0.0001708781491913365,
      "loss": 0.0462,
      "step": 900
    },
    {
      "epoch": 1.4490445859872612,
      "grad_norm": 0.31453198194503784,
      "learning_rate": 0.00017013079599188758,
      "loss": 0.0496,
      "step": 910
    },
    {
      "epoch": 1.4649681528662422,
      "grad_norm": 0.467495322227478,
      "learning_rate": 0.00016937565381327661,
      "loss": 0.0518,
      "step": 920
    },
    {
      "epoch": 1.480891719745223,
      "grad_norm": 0.5229964256286621,
      "learning_rate": 0.00016861280652431758,
      "loss": 0.0588,
      "step": 930
    },
    {
      "epoch": 1.4968152866242037,
      "grad_norm": 0.5483565926551819,
      "learning_rate": 0.00016784233884958212,
      "loss": 0.0454,
      "step": 940
    },
    {
      "epoch": 1.5127388535031847,
      "grad_norm": 0.6096242070198059,
      "learning_rate": 0.00016706433635998912,
      "loss": 0.0458,
      "step": 950
    },
    {
      "epoch": 1.5286624203821657,
      "grad_norm": 0.34970104694366455,
      "learning_rate": 0.00016627888546330138,
      "loss": 0.0498,
      "step": 960
    },
    {
      "epoch": 1.5445859872611465,
      "grad_norm": 0.691208004951477,
      "learning_rate": 0.00016548607339452853,
      "loss": 0.0446,
      "step": 970
    },
    {
      "epoch": 1.5605095541401273,
      "grad_norm": 0.32399702072143555,
      "learning_rate": 0.00016468598820623855,
      "loss": 0.0452,
      "step": 980
    },
    {
      "epoch": 1.5764331210191083,
      "grad_norm": 0.36962831020355225,
      "learning_rate": 0.0001638787187587784,
      "loss": 0.0415,
      "step": 990
    },
    {
      "epoch": 1.5923566878980893,
      "grad_norm": 0.45716020464897156,
      "learning_rate": 0.0001630643547104047,
      "loss": 0.0439,
      "step": 1000
    },
    {
      "epoch": 1.60828025477707,
      "grad_norm": 0.3454309403896332,
      "learning_rate": 0.00016224298650732597,
      "loss": 0.0532,
      "step": 1010
    },
    {
      "epoch": 1.6242038216560508,
      "grad_norm": 0.6210212707519531,
      "learning_rate": 0.00016141470537365754,
      "loss": 0.0414,
      "step": 1020
    },
    {
      "epoch": 1.6401273885350318,
      "grad_norm": 0.3841806650161743,
      "learning_rate": 0.00016057960330128973,
      "loss": 0.0484,
      "step": 1030
    },
    {
      "epoch": 1.6560509554140128,
      "grad_norm": 0.3426171839237213,
      "learning_rate": 0.0001597377730396708,
      "loss": 0.0423,
      "step": 1040
    },
    {
      "epoch": 1.6719745222929936,
      "grad_norm": 0.33869680762290955,
      "learning_rate": 0.0001588893080855061,
      "loss": 0.0436,
      "step": 1050
    },
    {
      "epoch": 1.6878980891719744,
      "grad_norm": 0.40435710549354553,
      "learning_rate": 0.00015803430267237383,
      "loss": 0.0517,
      "step": 1060
    },
    {
      "epoch": 1.7038216560509554,
      "grad_norm": 0.25334039330482483,
      "learning_rate": 0.00015717285176025913,
      "loss": 0.0339,
      "step": 1070
    },
    {
      "epoch": 1.7197452229299364,
      "grad_norm": 0.40064361691474915,
      "learning_rate": 0.00015630505102500754,
      "loss": 0.0472,
      "step": 1080
    },
    {
      "epoch": 1.7356687898089171,
      "grad_norm": 0.5399286150932312,
      "learning_rate": 0.0001554309968476988,
      "loss": 0.0448,
      "step": 1090
    },
    {
      "epoch": 1.7515923566878981,
      "grad_norm": 0.36550888419151306,
      "learning_rate": 0.0001545507863039426,
      "loss": 0.0348,
      "step": 1100
    },
    {
      "epoch": 1.767515923566879,
      "grad_norm": 0.5114671587944031,
      "learning_rate": 0.00015366451715309674,
      "loss": 0.0409,
      "step": 1110
    },
    {
      "epoch": 1.78343949044586,
      "grad_norm": 0.4588598608970642,
      "learning_rate": 0.00015277228782740988,
      "loss": 0.0394,
      "step": 1120
    },
    {
      "epoch": 1.799363057324841,
      "grad_norm": 0.35981065034866333,
      "learning_rate": 0.0001518741974210892,
      "loss": 0.0412,
      "step": 1130
    },
    {
      "epoch": 1.8152866242038217,
      "grad_norm": 0.39596328139305115,
      "learning_rate": 0.00015097034567929457,
      "loss": 0.0369,
      "step": 1140
    },
    {
      "epoch": 1.8312101910828025,
      "grad_norm": 0.45480477809906006,
      "learning_rate": 0.0001500608329870606,
      "loss": 0.0439,
      "step": 1150
    },
    {
      "epoch": 1.8471337579617835,
      "grad_norm": 0.28407368063926697,
      "learning_rate": 0.00014914576035814748,
      "loss": 0.0483,
      "step": 1160
    },
    {
      "epoch": 1.8630573248407645,
      "grad_norm": 0.4121032953262329,
      "learning_rate": 0.000148225229423822,
      "loss": 0.0488,
      "step": 1170
    },
    {
      "epoch": 1.8789808917197452,
      "grad_norm": 0.44136524200439453,
      "learning_rate": 0.00014729934242157004,
      "loss": 0.0434,
      "step": 1180
    },
    {
      "epoch": 1.894904458598726,
      "grad_norm": 0.43993085622787476,
      "learning_rate": 0.00014636820218374177,
      "loss": 0.0457,
      "step": 1190
    },
    {
      "epoch": 1.910828025477707,
      "grad_norm": 0.3353908360004425,
      "learning_rate": 0.00014543191212613047,
      "loss": 0.0486,
      "step": 1200
    },
    {
      "epoch": 1.926751592356688,
      "grad_norm": 0.4279186427593231,
      "learning_rate": 0.0001444905762364872,
      "loss": 0.0526,
      "step": 1210
    },
    {
      "epoch": 1.9426751592356688,
      "grad_norm": 0.5773082375526428,
      "learning_rate": 0.00014354429906297125,
      "loss": 0.0462,
      "step": 1220
    },
    {
      "epoch": 1.9585987261146496,
      "grad_norm": 0.39543482661247253,
      "learning_rate": 0.00014259318570253863,
      "loss": 0.0417,
      "step": 1230
    },
    {
      "epoch": 1.9745222929936306,
      "grad_norm": 0.39346298575401306,
      "learning_rate": 0.0001416373417892698,
      "loss": 0.0361,
      "step": 1240
    },
    {
      "epoch": 1.9904458598726116,
      "grad_norm": 0.32164451479911804,
      "learning_rate": 0.0001406768734826375,
      "loss": 0.0394,
      "step": 1250
    },
    {
      "epoch": 2.0063694267515926,
      "grad_norm": 0.551193356513977,
      "learning_rate": 0.00013971188745571616,
      "loss": 0.042,
      "step": 1260
    },
    {
      "epoch": 2.022292993630573,
      "grad_norm": 0.3745879530906677,
      "learning_rate": 0.00013874249088333468,
      "loss": 0.0432,
      "step": 1270
    },
    {
      "epoch": 2.038216560509554,
      "grad_norm": 0.36253151297569275,
      "learning_rate": 0.00013776879143017294,
      "loss": 0.0395,
      "step": 1280
    },
    {
      "epoch": 2.054140127388535,
      "grad_norm": 0.5354855060577393,
      "learning_rate": 0.00013679089723880427,
      "loss": 0.044,
      "step": 1290
    },
    {
      "epoch": 2.070063694267516,
      "grad_norm": 0.5604785680770874,
      "learning_rate": 0.00013580891691768483,
      "loss": 0.0377,
      "step": 1300
    },
    {
      "epoch": 2.0859872611464967,
      "grad_norm": 0.42235568165779114,
      "learning_rate": 0.00013482295952909102,
      "loss": 0.0381,
      "step": 1310
    },
    {
      "epoch": 2.1019108280254777,
      "grad_norm": 0.3224591612815857,
      "learning_rate": 0.0001338331345770068,
      "loss": 0.0396,
      "step": 1320
    },
    {
      "epoch": 2.1178343949044587,
      "grad_norm": 0.36872801184654236,
      "learning_rate": 0.00013283955199496156,
      "loss": 0.0469,
      "step": 1330
    },
    {
      "epoch": 2.1337579617834397,
      "grad_norm": 0.38453301787376404,
      "learning_rate": 0.0001318423221338207,
      "loss": 0.0433,
      "step": 1340
    },
    {
      "epoch": 2.1496815286624202,
      "grad_norm": 0.3108746409416199,
      "learning_rate": 0.0001308415557495295,
      "loss": 0.0373,
      "step": 1350
    },
    {
      "epoch": 2.1656050955414012,
      "grad_norm": 0.41351979970932007,
      "learning_rate": 0.00012983736399081242,
      "loss": 0.0362,
      "step": 1360
    },
    {
      "epoch": 2.1815286624203822,
      "grad_norm": 0.39941781759262085,
      "learning_rate": 0.0001288298583868282,
      "loss": 0.0444,
      "step": 1370
    },
    {
      "epoch": 2.1974522292993632,
      "grad_norm": 0.5019637942314148,
      "learning_rate": 0.00012781915083478332,
      "loss": 0.0362,
      "step": 1380
    },
    {
      "epoch": 2.213375796178344,
      "grad_norm": 0.44153285026550293,
      "learning_rate": 0.00012680535358750402,
      "loss": 0.048,
      "step": 1390
    },
    {
      "epoch": 2.229299363057325,
      "grad_norm": 0.3250869810581207,
      "learning_rate": 0.00012578857924096934,
      "loss": 0.038,
      "step": 1400
    },
    {
      "epoch": 2.245222929936306,
      "grad_norm": 0.38918137550354004,
      "learning_rate": 0.00012476894072180575,
      "loss": 0.0507,
      "step": 1410
    },
    {
      "epoch": 2.261146496815287,
      "grad_norm": 0.42312440276145935,
      "learning_rate": 0.00012374655127474487,
      "loss": 0.0401,
      "step": 1420
    },
    {
      "epoch": 2.2770700636942673,
      "grad_norm": 0.6262978911399841,
      "learning_rate": 0.00012272152445004647,
      "loss": 0.0476,
      "step": 1430
    },
    {
      "epoch": 2.2929936305732483,
      "grad_norm": 0.4742312431335449,
      "learning_rate": 0.00012169397409088695,
      "loss": 0.0375,
      "step": 1440
    },
    {
      "epoch": 2.3089171974522293,
      "grad_norm": 0.5429434180259705,
      "learning_rate": 0.00012066401432071551,
      "loss": 0.0486,
      "step": 1450
    },
    {
      "epoch": 2.3248407643312103,
      "grad_norm": 0.34995678067207336,
      "learning_rate": 0.00011963175953057924,
      "loss": 0.0398,
      "step": 1460
    },
    {
      "epoch": 2.340764331210191,
      "grad_norm": 0.26490700244903564,
      "learning_rate": 0.00011859732436641851,
      "loss": 0.0285,
      "step": 1470
    },
    {
      "epoch": 2.356687898089172,
      "grad_norm": 0.30831092596054077,
      "learning_rate": 0.00011756082371633373,
      "loss": 0.0344,
      "step": 1480
    },
    {
      "epoch": 2.372611464968153,
      "grad_norm": 0.35261446237564087,
      "learning_rate": 0.00011652237269782573,
      "loss": 0.0327,
      "step": 1490
    },
    {
      "epoch": 2.388535031847134,
      "grad_norm": 0.31007587909698486,
      "learning_rate": 0.00011548208664501034,
      "loss": 0.049,
      "step": 1500
    },
    {
      "epoch": 2.404458598726115,
      "grad_norm": 0.49864161014556885,
      "learning_rate": 0.00011444008109580884,
      "loss": 0.0318,
      "step": 1510
    },
    {
      "epoch": 2.4203821656050954,
      "grad_norm": 0.2871188521385193,
      "learning_rate": 0.00011339647177911599,
      "loss": 0.0335,
      "step": 1520
    },
    {
      "epoch": 2.4363057324840764,
      "grad_norm": 0.38693806529045105,
      "learning_rate": 0.00011235137460194676,
      "loss": 0.0334,
      "step": 1530
    },
    {
      "epoch": 2.4522292993630574,
      "grad_norm": 0.34477829933166504,
      "learning_rate": 0.00011130490563656326,
      "loss": 0.0278,
      "step": 1540
    },
    {
      "epoch": 2.468152866242038,
      "grad_norm": 0.21558038890361786,
      "learning_rate": 0.00011025718110758338,
      "loss": 0.0425,
      "step": 1550
    },
    {
      "epoch": 2.484076433121019,
      "grad_norm": 0.3483816683292389,
      "learning_rate": 0.0001092083173790724,
      "loss": 0.0308,
      "step": 1560
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.365614652633667,
      "learning_rate": 0.00010815843094161927,
      "loss": 0.0309,
      "step": 1570
    },
    {
      "epoch": 2.515923566878981,
      "grad_norm": 0.4408479928970337,
      "learning_rate": 0.00010710763839939857,
      "loss": 0.034,
      "step": 1580
    },
    {
      "epoch": 2.531847133757962,
      "grad_norm": 0.2580474615097046,
      "learning_rate": 0.00010605605645722024,
      "loss": 0.0349,
      "step": 1590
    },
    {
      "epoch": 2.5477707006369426,
      "grad_norm": 0.3337418735027313,
      "learning_rate": 0.0001050038019075678,
      "loss": 0.0342,
      "step": 1600
    },
    {
      "epoch": 2.5636942675159236,
      "grad_norm": 0.41936802864074707,
      "learning_rate": 0.00010395099161762698,
      "loss": 0.0404,
      "step": 1610
    },
    {
      "epoch": 2.5796178343949046,
      "grad_norm": 0.3816032111644745,
      "learning_rate": 0.00010289774251630602,
      "loss": 0.0442,
      "step": 1620
    },
    {
      "epoch": 2.595541401273885,
      "grad_norm": 0.30350789427757263,
      "learning_rate": 0.00010184417158124915,
      "loss": 0.036,
      "step": 1630
    },
    {
      "epoch": 2.611464968152866,
      "grad_norm": 0.20693911612033844,
      "learning_rate": 0.00010079039582584457,
      "loss": 0.0293,
      "step": 1640
    },
    {
      "epoch": 2.627388535031847,
      "grad_norm": 0.31562432646751404,
      "learning_rate": 9.973653228622867e-05,
      "loss": 0.0268,
      "step": 1650
    },
    {
      "epoch": 2.643312101910828,
      "grad_norm": 0.3229486048221588,
      "learning_rate": 9.86826980082874e-05,
      "loss": 0.0385,
      "step": 1660
    },
    {
      "epoch": 2.659235668789809,
      "grad_norm": 0.20771260559558868,
      "learning_rate": 9.762901003465679e-05,
      "loss": 0.0293,
      "step": 1670
    },
    {
      "epoch": 2.6751592356687897,
      "grad_norm": 0.2522100508213043,
      "learning_rate": 9.657558539172376e-05,
      "loss": 0.0273,
      "step": 1680
    },
    {
      "epoch": 2.6910828025477707,
      "grad_norm": 0.30410313606262207,
      "learning_rate": 9.552254107662888e-05,
      "loss": 0.039,
      "step": 1690
    },
    {
      "epoch": 2.7070063694267517,
      "grad_norm": 0.5962004661560059,
      "learning_rate": 9.446999404427209e-05,
      "loss": 0.0355,
      "step": 1700
    },
    {
      "epoch": 2.722929936305732,
      "grad_norm": 0.3611944913864136,
      "learning_rate": 9.341806119432324e-05,
      "loss": 0.0299,
      "step": 1710
    },
    {
      "epoch": 2.738853503184713,
      "grad_norm": 0.23662623763084412,
      "learning_rate": 9.2366859358239e-05,
      "loss": 0.0305,
      "step": 1720
    },
    {
      "epoch": 2.754777070063694,
      "grad_norm": 0.27245596051216125,
      "learning_rate": 9.131650528628687e-05,
      "loss": 0.0265,
      "step": 1730
    },
    {
      "epoch": 2.770700636942675,
      "grad_norm": 0.23096629977226257,
      "learning_rate": 9.026711563457868e-05,
      "loss": 0.0373,
      "step": 1740
    },
    {
      "epoch": 2.786624203821656,
      "grad_norm": 0.23357895016670227,
      "learning_rate": 8.921880695211431e-05,
      "loss": 0.0301,
      "step": 1750
    },
    {
      "epoch": 2.802547770700637,
      "grad_norm": 0.36908674240112305,
      "learning_rate": 8.817169566783723e-05,
      "loss": 0.0276,
      "step": 1760
    },
    {
      "epoch": 2.8184713375796178,
      "grad_norm": 0.20657651126384735,
      "learning_rate": 8.712589807770349e-05,
      "loss": 0.0299,
      "step": 1770
    },
    {
      "epoch": 2.8343949044585988,
      "grad_norm": 0.27385637164115906,
      "learning_rate": 8.60815303317657e-05,
      "loss": 0.0312,
      "step": 1780
    },
    {
      "epoch": 2.8503184713375798,
      "grad_norm": 0.4029058516025543,
      "learning_rate": 8.503870842127267e-05,
      "loss": 0.0391,
      "step": 1790
    },
    {
      "epoch": 2.8662420382165603,
      "grad_norm": 0.2755729556083679,
      "learning_rate": 8.399754816578718e-05,
      "loss": 0.0367,
      "step": 1800
    },
    {
      "epoch": 2.8821656050955413,
      "grad_norm": 0.2416078746318817,
      "learning_rate": 8.295816520032271e-05,
      "loss": 0.0274,
      "step": 1810
    },
    {
      "epoch": 2.8980891719745223,
      "grad_norm": 0.22337061166763306,
      "learning_rate": 8.192067496250038e-05,
      "loss": 0.0219,
      "step": 1820
    },
    {
      "epoch": 2.9140127388535033,
      "grad_norm": 0.21762487292289734,
      "learning_rate": 8.088519267972813e-05,
      "loss": 0.023,
      "step": 1830
    },
    {
      "epoch": 2.9299363057324843,
      "grad_norm": 0.32507646083831787,
      "learning_rate": 7.985183335640331e-05,
      "loss": 0.0284,
      "step": 1840
    },
    {
      "epoch": 2.945859872611465,
      "grad_norm": 0.3356756269931793,
      "learning_rate": 7.882071176113967e-05,
      "loss": 0.0304,
      "step": 1850
    },
    {
      "epoch": 2.961783439490446,
      "grad_norm": 0.31162557005882263,
      "learning_rate": 7.779194241402071e-05,
      "loss": 0.0282,
      "step": 1860
    },
    {
      "epoch": 2.977707006369427,
      "grad_norm": 0.24151593446731567,
      "learning_rate": 7.676563957388097e-05,
      "loss": 0.0248,
      "step": 1870
    },
    {
      "epoch": 2.9936305732484074,
      "grad_norm": 0.41317158937454224,
      "learning_rate": 7.574191722561562e-05,
      "loss": 0.0303,
      "step": 1880
    },
    {
      "epoch": 3.0095541401273884,
      "grad_norm": 0.33723485469818115,
      "learning_rate": 7.472088906752124e-05,
      "loss": 0.0326,
      "step": 1890
    },
    {
      "epoch": 3.0254777070063694,
      "grad_norm": 0.402401328086853,
      "learning_rate": 7.37026684986678e-05,
      "loss": 0.0327,
      "step": 1900
    },
    {
      "epoch": 3.0414012738853504,
      "grad_norm": 0.2733253538608551,
      "learning_rate": 7.26873686063045e-05,
      "loss": 0.0286,
      "step": 1910
    },
    {
      "epoch": 3.0573248407643314,
      "grad_norm": 0.2582903802394867,
      "learning_rate": 7.16751021532994e-05,
      "loss": 0.0232,
      "step": 1920
    },
    {
      "epoch": 3.073248407643312,
      "grad_norm": 0.3948001265525818,
      "learning_rate": 7.066598156561606e-05,
      "loss": 0.0331,
      "step": 1930
    },
    {
      "epoch": 3.089171974522293,
      "grad_norm": 0.23859384655952454,
      "learning_rate": 6.966011891982678e-05,
      "loss": 0.0253,
      "step": 1940
    },
    {
      "epoch": 3.105095541401274,
      "grad_norm": 0.2874588072299957,
      "learning_rate": 6.865762593066513e-05,
      "loss": 0.0316,
      "step": 1950
    },
    {
      "epoch": 3.121019108280255,
      "grad_norm": 0.40092089772224426,
      "learning_rate": 6.76586139386184e-05,
      "loss": 0.0294,
      "step": 1960
    },
    {
      "epoch": 3.1369426751592355,
      "grad_norm": 0.2927519679069519,
      "learning_rate": 6.666319389756189e-05,
      "loss": 0.0323,
      "step": 1970
    },
    {
      "epoch": 3.1528662420382165,
      "grad_norm": 0.2560182809829712,
      "learning_rate": 6.567147636243576e-05,
      "loss": 0.0292,
      "step": 1980
    },
    {
      "epoch": 3.1687898089171975,
      "grad_norm": 0.18076066672801971,
      "learning_rate": 6.468357147696653e-05,
      "loss": 0.0223,
      "step": 1990
    },
    {
      "epoch": 3.1847133757961785,
      "grad_norm": 0.32147490978240967,
      "learning_rate": 6.36995889614342e-05,
      "loss": 0.028,
      "step": 2000
    },
    {
      "epoch": 3.200636942675159,
      "grad_norm": 0.24778632819652557,
      "learning_rate": 6.271963810048607e-05,
      "loss": 0.0244,
      "step": 2010
    },
    {
      "epoch": 3.21656050955414,
      "grad_norm": 0.2591859698295593,
      "learning_rate": 6.174382773099938e-05,
      "loss": 0.0252,
      "step": 2020
    },
    {
      "epoch": 3.232484076433121,
      "grad_norm": 0.37974950671195984,
      "learning_rate": 6.077226622999355e-05,
      "loss": 0.0293,
      "step": 2030
    },
    {
      "epoch": 3.248407643312102,
      "grad_norm": 0.36128807067871094,
      "learning_rate": 5.980506150259323e-05,
      "loss": 0.0253,
      "step": 2040
    },
    {
      "epoch": 3.2643312101910826,
      "grad_norm": 0.2899702489376068,
      "learning_rate": 5.88423209700441e-05,
      "loss": 0.0199,
      "step": 2050
    },
    {
      "epoch": 3.2802547770700636,
      "grad_norm": 0.24200858175754547,
      "learning_rate": 5.7884151557782305e-05,
      "loss": 0.0334,
      "step": 2060
    },
    {
      "epoch": 3.2961783439490446,
      "grad_norm": 0.24898533523082733,
      "learning_rate": 5.693065968355888e-05,
      "loss": 0.0269,
      "step": 2070
    },
    {
      "epoch": 3.3121019108280256,
      "grad_norm": 0.33028173446655273,
      "learning_rate": 5.598195124562049e-05,
      "loss": 0.0309,
      "step": 2080
    },
    {
      "epoch": 3.328025477707006,
      "grad_norm": 0.32554179430007935,
      "learning_rate": 5.5038131610948326e-05,
      "loss": 0.0274,
      "step": 2090
    },
    {
      "epoch": 3.343949044585987,
      "grad_norm": 0.25384682416915894,
      "learning_rate": 5.409930560355527e-05,
      "loss": 0.0235,
      "step": 2100
    },
    {
      "epoch": 3.359872611464968,
      "grad_norm": 0.3368367850780487,
      "learning_rate": 5.316557749284401e-05,
      "loss": 0.0259,
      "step": 2110
    },
    {
      "epoch": 3.375796178343949,
      "grad_norm": 0.3751080334186554,
      "learning_rate": 5.223705098202647e-05,
      "loss": 0.0405,
      "step": 2120
    },
    {
      "epoch": 3.3917197452229297,
      "grad_norm": 0.3862003684043884,
      "learning_rate": 5.131382919660609e-05,
      "loss": 0.0311,
      "step": 2130
    },
    {
      "epoch": 3.4076433121019107,
      "grad_norm": 0.32009872794151306,
      "learning_rate": 5.03960146729244e-05,
      "loss": 0.0252,
      "step": 2140
    },
    {
      "epoch": 3.4235668789808917,
      "grad_norm": 0.2071923166513443,
      "learning_rate": 4.9483709346772896e-05,
      "loss": 0.0248,
      "step": 2150
    },
    {
      "epoch": 3.4394904458598727,
      "grad_norm": 0.3461016118526459,
      "learning_rate": 4.85770145420718e-05,
      "loss": 0.0336,
      "step": 2160
    },
    {
      "epoch": 3.4554140127388537,
      "grad_norm": 0.3042682111263275,
      "learning_rate": 4.7676030959616526e-05,
      "loss": 0.0211,
      "step": 2170
    },
    {
      "epoch": 3.4713375796178343,
      "grad_norm": 0.16671477258205414,
      "learning_rate": 4.678085866589348e-05,
      "loss": 0.018,
      "step": 2180
    },
    {
      "epoch": 3.4872611464968153,
      "grad_norm": 0.20116400718688965,
      "learning_rate": 4.5891597081966583e-05,
      "loss": 0.0224,
      "step": 2190
    },
    {
      "epoch": 3.5031847133757963,
      "grad_norm": 0.23196861147880554,
      "learning_rate": 4.500834497243478e-05,
      "loss": 0.0221,
      "step": 2200
    },
    {
      "epoch": 3.519108280254777,
      "grad_norm": 0.2727397084236145,
      "learning_rate": 4.413120043446322e-05,
      "loss": 0.0272,
      "step": 2210
    },
    {
      "epoch": 3.535031847133758,
      "grad_norm": 0.283627450466156,
      "learning_rate": 4.3260260886888206e-05,
      "loss": 0.0259,
      "step": 2220
    },
    {
      "epoch": 3.550955414012739,
      "grad_norm": 0.40015652775764465,
      "learning_rate": 4.239562305939725e-05,
      "loss": 0.0298,
      "step": 2230
    },
    {
      "epoch": 3.56687898089172,
      "grad_norm": 0.30534714460372925,
      "learning_rate": 4.153738298178623e-05,
      "loss": 0.0245,
      "step": 2240
    },
    {
      "epoch": 3.582802547770701,
      "grad_norm": 0.22230559587478638,
      "learning_rate": 4.068563597329379e-05,
      "loss": 0.0231,
      "step": 2250
    },
    {
      "epoch": 3.5987261146496814,
      "grad_norm": 0.3529864549636841,
      "learning_rate": 3.984047663201492e-05,
      "loss": 0.04,
      "step": 2260
    },
    {
      "epoch": 3.6146496815286624,
      "grad_norm": 0.2945915460586548,
      "learning_rate": 3.9001998824394526e-05,
      "loss": 0.0251,
      "step": 2270
    },
    {
      "epoch": 3.6305732484076434,
      "grad_norm": 0.24476458132266998,
      "learning_rate": 3.817029567480228e-05,
      "loss": 0.0245,
      "step": 2280
    },
    {
      "epoch": 3.646496815286624,
      "grad_norm": 0.1646440178155899,
      "learning_rate": 3.7345459555189885e-05,
      "loss": 0.0246,
      "step": 2290
    },
    {
      "epoch": 3.662420382165605,
      "grad_norm": 0.21083424985408783,
      "learning_rate": 3.652758207483197e-05,
      "loss": 0.0272,
      "step": 2300
    },
    {
      "epoch": 3.678343949044586,
      "grad_norm": 0.17064619064331055,
      "learning_rate": 3.5716754070151524e-05,
      "loss": 0.0169,
      "step": 2310
    },
    {
      "epoch": 3.694267515923567,
      "grad_norm": 0.18861030042171478,
      "learning_rate": 3.4913065594631434e-05,
      "loss": 0.0257,
      "step": 2320
    },
    {
      "epoch": 3.710191082802548,
      "grad_norm": 0.31544387340545654,
      "learning_rate": 3.4116605908812595e-05,
      "loss": 0.0237,
      "step": 2330
    },
    {
      "epoch": 3.7261146496815285,
      "grad_norm": 0.5185452103614807,
      "learning_rate": 3.3327463470380524e-05,
      "loss": 0.0229,
      "step": 2340
    },
    {
      "epoch": 3.7420382165605095,
      "grad_norm": 0.2828991115093231,
      "learning_rate": 3.2545725924340876e-05,
      "loss": 0.0294,
      "step": 2350
    },
    {
      "epoch": 3.7579617834394905,
      "grad_norm": 0.2944129407405853,
      "learning_rate": 3.1771480093285186e-05,
      "loss": 0.0218,
      "step": 2360
    },
    {
      "epoch": 3.7738853503184715,
      "grad_norm": 0.24150818586349487,
      "learning_rate": 3.1004811967748134e-05,
      "loss": 0.024,
      "step": 2370
    },
    {
      "epoch": 3.789808917197452,
      "grad_norm": 0.2700318396091461,
      "learning_rate": 3.024580669665711e-05,
      "loss": 0.0244,
      "step": 2380
    },
    {
      "epoch": 3.805732484076433,
      "grad_norm": 0.4214876592159271,
      "learning_rate": 2.9494548577875192e-05,
      "loss": 0.0264,
      "step": 2390
    },
    {
      "epoch": 3.821656050955414,
      "grad_norm": 0.2855050265789032,
      "learning_rate": 2.8751121048838836e-05,
      "loss": 0.02,
      "step": 2400
    },
    {
      "epoch": 3.837579617834395,
      "grad_norm": 0.19058647751808167,
      "learning_rate": 2.801560667729093e-05,
      "loss": 0.0243,
      "step": 2410
    },
    {
      "epoch": 3.853503184713376,
      "grad_norm": 0.32941293716430664,
      "learning_rate": 2.7288087152110565e-05,
      "loss": 0.0268,
      "step": 2420
    },
    {
      "epoch": 3.8694267515923566,
      "grad_norm": 0.4154011905193329,
      "learning_rate": 2.656864327424027e-05,
      "loss": 0.0206,
      "step": 2430
    },
    {
      "epoch": 3.8853503184713376,
      "grad_norm": 0.3098050057888031,
      "learning_rate": 2.585735494771222e-05,
      "loss": 0.023,
      "step": 2440
    },
    {
      "epoch": 3.9012738853503186,
      "grad_norm": 0.27963143587112427,
      "learning_rate": 2.515430117077354e-05,
      "loss": 0.0273,
      "step": 2450
    },
    {
      "epoch": 3.917197452229299,
      "grad_norm": 0.29374557733535767,
      "learning_rate": 2.4459560027112504e-05,
      "loss": 0.0244,
      "step": 2460
    },
    {
      "epoch": 3.93312101910828,
      "grad_norm": 0.1925758719444275,
      "learning_rate": 2.377320867718654e-05,
      "loss": 0.0214,
      "step": 2470
    },
    {
      "epoch": 3.949044585987261,
      "grad_norm": 0.3169631361961365,
      "learning_rate": 2.3095323349652153e-05,
      "loss": 0.0258,
      "step": 2480
    },
    {
      "epoch": 3.964968152866242,
      "grad_norm": 0.2621577978134155,
      "learning_rate": 2.242597933289894e-05,
      "loss": 0.0238,
      "step": 2490
    },
    {
      "epoch": 3.980891719745223,
      "grad_norm": 0.2187499850988388,
      "learning_rate": 2.176525096668769e-05,
      "loss": 0.0266,
      "step": 2500
    },
    {
      "epoch": 3.9968152866242037,
      "grad_norm": 0.2935671806335449,
      "learning_rate": 2.1113211633894003e-05,
      "loss": 0.0214,
      "step": 2510
    },
    {
      "epoch": 4.012738853503185,
      "grad_norm": 0.1593521386384964,
      "learning_rate": 2.046993375235804e-05,
      "loss": 0.0166,
      "step": 2520
    },
    {
      "epoch": 4.028662420382165,
      "grad_norm": 0.32574114203453064,
      "learning_rate": 1.9835488766841615e-05,
      "loss": 0.0232,
      "step": 2530
    },
    {
      "epoch": 4.044585987261146,
      "grad_norm": 0.2108948677778244,
      "learning_rate": 1.920994714109323e-05,
      "loss": 0.0255,
      "step": 2540
    },
    {
      "epoch": 4.060509554140127,
      "grad_norm": 0.2054058164358139,
      "learning_rate": 1.8593378350022183e-05,
      "loss": 0.0156,
      "step": 2550
    },
    {
      "epoch": 4.076433121019108,
      "grad_norm": 0.28134655952453613,
      "learning_rate": 1.798585087198228e-05,
      "loss": 0.0244,
      "step": 2560
    },
    {
      "epoch": 4.092356687898089,
      "grad_norm": 0.26843348145484924,
      "learning_rate": 1.7387432181166642e-05,
      "loss": 0.0294,
      "step": 2570
    },
    {
      "epoch": 4.10828025477707,
      "grad_norm": 0.32004109025001526,
      "learning_rate": 1.6798188740113484e-05,
      "loss": 0.0263,
      "step": 2580
    },
    {
      "epoch": 4.124203821656051,
      "grad_norm": 0.6330613493919373,
      "learning_rate": 1.621818599232475e-05,
      "loss": 0.0178,
      "step": 2590
    },
    {
      "epoch": 4.140127388535032,
      "grad_norm": 0.21530304849147797,
      "learning_rate": 1.564748835499773e-05,
      "loss": 0.0277,
      "step": 2600
    },
    {
      "epoch": 4.156050955414012,
      "grad_norm": 0.20998333394527435,
      "learning_rate": 1.5086159211870442e-05,
      "loss": 0.0214,
      "step": 2610
    },
    {
      "epoch": 4.171974522292993,
      "grad_norm": 0.3173789381980896,
      "learning_rate": 1.4534260906182228e-05,
      "loss": 0.0198,
      "step": 2620
    },
    {
      "epoch": 4.187898089171974,
      "grad_norm": 0.18961144983768463,
      "learning_rate": 1.399185473374961e-05,
      "loss": 0.021,
      "step": 2630
    },
    {
      "epoch": 4.203821656050955,
      "grad_norm": 0.36942291259765625,
      "learning_rate": 1.3459000936158528e-05,
      "loss": 0.0239,
      "step": 2640
    },
    {
      "epoch": 4.219745222929936,
      "grad_norm": 0.28133276104927063,
      "learning_rate": 1.293575869407373e-05,
      "loss": 0.0244,
      "step": 2650
    },
    {
      "epoch": 4.235668789808917,
      "grad_norm": 0.2719907760620117,
      "learning_rate": 1.2422186120665935e-05,
      "loss": 0.0203,
      "step": 2660
    },
    {
      "epoch": 4.251592356687898,
      "grad_norm": 0.15558955073356628,
      "learning_rate": 1.1918340255157612e-05,
      "loss": 0.0214,
      "step": 2670
    },
    {
      "epoch": 4.267515923566879,
      "grad_norm": 0.23012204468250275,
      "learning_rate": 1.1424277056487864e-05,
      "loss": 0.0232,
      "step": 2680
    },
    {
      "epoch": 4.2834394904458595,
      "grad_norm": 0.29646140336990356,
      "learning_rate": 1.0940051397097673e-05,
      "loss": 0.0179,
      "step": 2690
    },
    {
      "epoch": 4.2993630573248405,
      "grad_norm": 0.282185435295105,
      "learning_rate": 1.0465717056835355e-05,
      "loss": 0.0171,
      "step": 2700
    },
    {
      "epoch": 4.3152866242038215,
      "grad_norm": 0.3244917392730713,
      "learning_rate": 1.0001326716983584e-05,
      "loss": 0.0161,
      "step": 2710
    },
    {
      "epoch": 4.3312101910828025,
      "grad_norm": 0.2768123149871826,
      "learning_rate": 9.546931954408622e-06,
      "loss": 0.0248,
      "step": 2720
    },
    {
      "epoch": 4.3471337579617835,
      "grad_norm": 0.14225247502326965,
      "learning_rate": 9.10258323583173e-06,
      "loss": 0.0171,
      "step": 2730
    },
    {
      "epoch": 4.3630573248407645,
      "grad_norm": 0.16188742220401764,
      "learning_rate": 8.668329912224337e-06,
      "loss": 0.0177,
      "step": 2740
    },
    {
      "epoch": 4.3789808917197455,
      "grad_norm": 0.30127179622650146,
      "learning_rate": 8.244220213326958e-06,
      "loss": 0.0213,
      "step": 2750
    },
    {
      "epoch": 4.3949044585987265,
      "grad_norm": 0.1804376244544983,
      "learning_rate": 7.83030124229246e-06,
      "loss": 0.0232,
      "step": 2760
    },
    {
      "epoch": 4.4108280254777075,
      "grad_norm": 0.24196097254753113,
      "learning_rate": 7.426618970454757e-06,
      "loss": 0.0163,
      "step": 2770
    },
    {
      "epoch": 4.426751592356688,
      "grad_norm": 0.4308320879936218,
      "learning_rate": 7.033218232223027e-06,
      "loss": 0.0165,
      "step": 2780
    },
    {
      "epoch": 4.442675159235669,
      "grad_norm": 0.3199467062950134,
      "learning_rate": 6.650142720102248e-06,
      "loss": 0.024,
      "step": 2790
    },
    {
      "epoch": 4.45859872611465,
      "grad_norm": 0.2603253722190857,
      "learning_rate": 6.2774349798405465e-06,
      "loss": 0.0232,
      "step": 2800
    },
    {
      "epoch": 4.474522292993631,
      "grad_norm": 0.20752651989459991,
      "learning_rate": 5.915136405703858e-06,
      "loss": 0.0223,
      "step": 2810
    },
    {
      "epoch": 4.490445859872612,
      "grad_norm": 0.17525236308574677,
      "learning_rate": 5.563287235878678e-06,
      "loss": 0.0168,
      "step": 2820
    },
    {
      "epoch": 4.506369426751593,
      "grad_norm": 0.20458337664604187,
      "learning_rate": 5.221926548002876e-06,
      "loss": 0.0208,
      "step": 2830
    },
    {
      "epoch": 4.522292993630574,
      "grad_norm": 0.353657066822052,
      "learning_rate": 4.891092254825725e-06,
      "loss": 0.0224,
      "step": 2840
    },
    {
      "epoch": 4.538216560509554,
      "grad_norm": 0.23591625690460205,
      "learning_rate": 4.570821099997169e-06,
      "loss": 0.0186,
      "step": 2850
    },
    {
      "epoch": 4.554140127388535,
      "grad_norm": 0.2248276323080063,
      "learning_rate": 4.261148653986824e-06,
      "loss": 0.0201,
      "step": 2860
    },
    {
      "epoch": 4.570063694267516,
      "grad_norm": 0.19413244724273682,
      "learning_rate": 3.962109310133544e-06,
      "loss": 0.0209,
      "step": 2870
    },
    {
      "epoch": 4.585987261146497,
      "grad_norm": 0.1779373288154602,
      "learning_rate": 3.6737362808254793e-06,
      "loss": 0.0164,
      "step": 2880
    },
    {
      "epoch": 4.601910828025478,
      "grad_norm": 0.13120533525943756,
      "learning_rate": 3.3960615938114172e-06,
      "loss": 0.02,
      "step": 2890
    },
    {
      "epoch": 4.617834394904459,
      "grad_norm": 0.24661269783973694,
      "learning_rate": 3.1291160886436575e-06,
      "loss": 0.0199,
      "step": 2900
    },
    {
      "epoch": 4.63375796178344,
      "grad_norm": 0.20324034988880157,
      "learning_rate": 2.872929413252867e-06,
      "loss": 0.0168,
      "step": 2910
    },
    {
      "epoch": 4.649681528662421,
      "grad_norm": 0.1891164630651474,
      "learning_rate": 2.6275300206552667e-06,
      "loss": 0.0239,
      "step": 2920
    },
    {
      "epoch": 4.665605095541402,
      "grad_norm": 0.446866512298584,
      "learning_rate": 2.3929451657925482e-06,
      "loss": 0.0215,
      "step": 2930
    },
    {
      "epoch": 4.681528662420382,
      "grad_norm": 0.1874268352985382,
      "learning_rate": 2.1692009025048422e-06,
      "loss": 0.0194,
      "step": 2940
    },
    {
      "epoch": 4.697452229299363,
      "grad_norm": 0.3293200135231018,
      "learning_rate": 1.956322080637052e-06,
      "loss": 0.0245,
      "step": 2950
    },
    {
      "epoch": 4.713375796178344,
      "grad_norm": 0.21030674874782562,
      "learning_rate": 1.754332343278986e-06,
      "loss": 0.0218,
      "step": 2960
    },
    {
      "epoch": 4.729299363057325,
      "grad_norm": 0.1421198844909668,
      "learning_rate": 1.5632541241394461e-06,
      "loss": 0.017,
      "step": 2970
    },
    {
      "epoch": 4.745222929936306,
      "grad_norm": 0.30587485432624817,
      "learning_rate": 1.3831086450546648e-06,
      "loss": 0.0165,
      "step": 2980
    },
    {
      "epoch": 4.761146496815287,
      "grad_norm": 0.28812679648399353,
      "learning_rate": 1.213915913631314e-06,
      "loss": 0.0221,
      "step": 2990
    },
    {
      "epoch": 4.777070063694268,
      "grad_norm": 0.20168578624725342,
      "learning_rate": 1.055694721024425e-06,
      "loss": 0.0197,
      "step": 3000
    },
    {
      "epoch": 4.792993630573249,
      "grad_norm": 0.31528469920158386,
      "learning_rate": 9.084626398503604e-07,
      "loss": 0.0278,
      "step": 3010
    },
    {
      "epoch": 4.80891719745223,
      "grad_norm": 0.2268596887588501,
      "learning_rate": 7.722360222350955e-07,
      "loss": 0.0196,
      "step": 3020
    },
    {
      "epoch": 4.82484076433121,
      "grad_norm": 0.21379084885120392,
      "learning_rate": 6.470299979981609e-07,
      "loss": 0.0242,
      "step": 3030
    },
    {
      "epoch": 4.840764331210191,
      "grad_norm": 0.4038665294647217,
      "learning_rate": 5.328584729722197e-07,
      "loss": 0.0208,
      "step": 3040
    },
    {
      "epoch": 4.856687898089172,
      "grad_norm": 0.1791532039642334,
      "learning_rate": 4.297341274586475e-07,
      "loss": 0.0248,
      "step": 3050
    },
    {
      "epoch": 4.872611464968153,
      "grad_norm": 0.2981935739517212,
      "learning_rate": 3.376684148192255e-07,
      "loss": 0.0166,
      "step": 3060
    },
    {
      "epoch": 4.888535031847134,
      "grad_norm": 0.37263256311416626,
      "learning_rate": 2.5667156020408034e-07,
      "loss": 0.0161,
      "step": 3070
    },
    {
      "epoch": 4.904458598726115,
      "grad_norm": 0.19920985400676727,
      "learning_rate": 1.8675255941604797e-07,
      "loss": 0.0296,
      "step": 3080
    },
    {
      "epoch": 4.920382165605096,
      "grad_norm": 0.17917926609516144,
      "learning_rate": 1.2791917791152853e-07,
      "loss": 0.024,
      "step": 3090
    },
    {
      "epoch": 4.936305732484076,
      "grad_norm": 0.2005498707294464,
      "learning_rate": 8.017794993808725e-08,
      "loss": 0.0223,
      "step": 3100
    },
    {
      "epoch": 4.952229299363057,
      "grad_norm": 0.23752698302268982,
      "learning_rate": 4.353417780872393e-08,
      "loss": 0.0199,
      "step": 3110
    },
    {
      "epoch": 4.968152866242038,
      "grad_norm": 0.20469646155834198,
      "learning_rate": 1.7991931312921762e-08,
      "loss": 0.021,
      "step": 3120
    },
    {
      "epoch": 4.984076433121019,
      "grad_norm": 0.20618374645709991,
      "learning_rate": 3.5540472647088707e-09,
      "loss": 0.0183,
      "step": 3130
    },
    {
      "epoch": 4.996815286624204,
      "step": 3138,
      "total_flos": 1.1375719484380992e+17,
      "train_loss": 0.04994222255384489,
      "train_runtime": 1406.117,
      "train_samples_per_second": 35.707,
      "train_steps_per_second": 2.232
    }
  ],
  "logging_steps": 10,
  "max_steps": 3138,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1375719484380992e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}