{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.92,
  "eval_steps": 500,
  "global_step": 2400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008,
      "grad_norm": 10.566986083984375,
      "learning_rate": 0.0,
      "loss": 14.1421,
      "step": 1
    },
    {
      "epoch": 0.008,
      "grad_norm": 12.296218872070312,
      "learning_rate": 1.730769230769231e-05,
      "loss": 13.35,
      "step": 10
    },
    {
      "epoch": 0.016,
      "grad_norm": 6.457699775695801,
      "learning_rate": 3.653846153846154e-05,
      "loss": 11.8957,
      "step": 20
    },
    {
      "epoch": 0.024,
      "grad_norm": 6.461245059967041,
      "learning_rate": 5.576923076923077e-05,
      "loss": 11.2465,
      "step": 30
    },
    {
      "epoch": 0.032,
      "grad_norm": 6.351202011108398,
      "learning_rate": 7.500000000000001e-05,
      "loss": 10.7197,
      "step": 40
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.675596714019775,
      "learning_rate": 9.423076923076924e-05,
      "loss": 10.4108,
      "step": 50
    },
    {
      "epoch": 0.048,
      "grad_norm": 6.45210599899292,
      "learning_rate": 0.00011346153846153846,
      "loss": 9.499,
      "step": 60
    },
    {
      "epoch": 0.056,
      "grad_norm": 3.97434663772583,
      "learning_rate": 0.0001326923076923077,
      "loss": 9.2464,
      "step": 70
    },
    {
      "epoch": 0.064,
      "grad_norm": 4.443643093109131,
      "learning_rate": 0.00015192307692307692,
      "loss": 9.0007,
      "step": 80
    },
    {
      "epoch": 0.072,
      "grad_norm": 4.448770046234131,
      "learning_rate": 0.00017115384615384616,
      "loss": 8.8057,
      "step": 90
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.425487041473389,
      "learning_rate": 0.00019038461538461538,
      "loss": 8.9744,
      "step": 100
    },
    {
      "epoch": 0.088,
      "grad_norm": 4.242831230163574,
      "learning_rate": 0.00019999785100910492,
      "loss": 8.9241,
      "step": 110
    },
    {
      "epoch": 0.096,
      "grad_norm": 3.6791751384735107,
      "learning_rate": 0.00019998065963611962,
      "loss": 8.8742,
      "step": 120
    },
    {
      "epoch": 0.104,
      "grad_norm": 5.0801777839660645,
      "learning_rate": 0.00019994627984564557,
      "loss": 8.8388,
      "step": 130
    },
    {
      "epoch": 0.112,
      "grad_norm": 5.117883205413818,
      "learning_rate": 0.00019989471754816785,
      "loss": 8.412,
      "step": 140
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.7691802978515625,
      "learning_rate": 0.00019982598160814377,
      "loss": 8.7482,
      "step": 150
    },
    {
      "epoch": 0.128,
      "grad_norm": 4.111888885498047,
      "learning_rate": 0.00019974008384247908,
      "loss": 8.6456,
      "step": 160
    },
    {
      "epoch": 0.136,
      "grad_norm": 3.717806816101074,
      "learning_rate": 0.0001996370390184965,
      "loss": 8.3429,
      "step": 170
    },
    {
      "epoch": 0.144,
      "grad_norm": 5.897804260253906,
      "learning_rate": 0.00019951686485139672,
      "loss": 8.5481,
      "step": 180
    },
    {
      "epoch": 0.152,
      "grad_norm": 4.452871799468994,
      "learning_rate": 0.00019937958200121303,
      "loss": 8.6357,
      "step": 190
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.103796482086182,
      "learning_rate": 0.0001992252140692594,
      "loss": 8.5245,
      "step": 200
    },
    {
      "epoch": 0.168,
      "grad_norm": 5.806966304779053,
      "learning_rate": 0.00019905378759407314,
      "loss": 8.4875,
      "step": 210
    },
    {
      "epoch": 0.176,
      "grad_norm": 4.293936729431152,
      "learning_rate": 0.00019886533204685228,
      "loss": 8.3073,
      "step": 220
    },
    {
      "epoch": 0.184,
      "grad_norm": 3.6153390407562256,
      "learning_rate": 0.00019865987982638914,
      "loss": 8.5256,
      "step": 230
    },
    {
      "epoch": 0.192,
      "grad_norm": 5.031829357147217,
      "learning_rate": 0.00019843746625350028,
      "loss": 8.3936,
      "step": 240
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.666059970855713,
      "learning_rate": 0.0001981981295649543,
      "loss": 7.9453,
      "step": 250
    },
    {
      "epoch": 0.208,
      "grad_norm": 4.338928699493408,
      "learning_rate": 0.0001979419109068982,
      "loss": 8.5403,
      "step": 260
    },
    {
      "epoch": 0.216,
      "grad_norm": 5.491336345672607,
      "learning_rate": 0.0001976688543277838,
      "loss": 8.5499,
      "step": 270
    },
    {
      "epoch": 0.224,
      "grad_norm": 4.206221580505371,
      "learning_rate": 0.00019737900677079483,
      "loss": 8.202,
      "step": 280
    },
    {
      "epoch": 0.232,
      "grad_norm": 4.248091220855713,
      "learning_rate": 0.0001970724180657768,
      "loss": 8.1605,
      "step": 290
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.153928279876709,
      "learning_rate": 0.00019674914092067015,
      "loss": 8.2001,
      "step": 300
    },
    {
      "epoch": 0.248,
      "grad_norm": 4.331130027770996,
      "learning_rate": 0.00019640923091244906,
      "loss": 8.3479,
      "step": 310
    },
    {
      "epoch": 0.256,
      "grad_norm": 4.888726711273193,
      "learning_rate": 0.0001960527464775666,
      "loss": 8.0348,
      "step": 320
    },
    {
      "epoch": 0.264,
      "grad_norm": 4.52598762512207,
      "learning_rate": 0.00019567974890190865,
      "loss": 7.9916,
      "step": 330
    },
    {
      "epoch": 0.272,
      "grad_norm": 5.109200477600098,
      "learning_rate": 0.00019529030231025776,
      "loss": 7.8621,
      "step": 340
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.465454578399658,
      "learning_rate": 0.0001948844736552688,
      "loss": 8.1529,
      "step": 350
    },
    {
      "epoch": 0.288,
      "grad_norm": 4.94981050491333,
      "learning_rate": 0.00019446233270595896,
      "loss": 7.9475,
      "step": 360
    },
    {
      "epoch": 0.296,
      "grad_norm": 4.898144245147705,
      "learning_rate": 0.00019402395203571286,
      "loss": 8.1256,
      "step": 370
    },
    {
      "epoch": 0.304,
      "grad_norm": 4.506499767303467,
      "learning_rate": 0.00019356940700980625,
      "loss": 7.9425,
      "step": 380
    },
    {
      "epoch": 0.312,
      "grad_norm": 4.715751647949219,
      "learning_rate": 0.00019309877577244924,
      "loss": 7.8867,
      "step": 390
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.232232570648193,
      "learning_rate": 0.00019261213923335194,
      "loss": 8.0137,
      "step": 400
    },
    {
      "epoch": 0.328,
      "grad_norm": 5.4095258712768555,
      "learning_rate": 0.0001921095810538148,
      "loss": 7.655,
      "step": 410
    },
    {
      "epoch": 0.336,
      "grad_norm": 8.021153450012207,
      "learning_rate": 0.00019159118763234555,
      "loss": 8.239,
      "step": 420
    },
    {
      "epoch": 0.344,
      "grad_norm": 4.821053504943848,
      "learning_rate": 0.0001910570480898061,
      "loss": 7.6991,
      "step": 430
    },
    {
      "epoch": 0.352,
      "grad_norm": 4.676478385925293,
      "learning_rate": 0.00019050725425409076,
      "loss": 7.9241,
      "step": 440
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.322430610656738,
      "learning_rate": 0.0001899419006443397,
      "loss": 8.0194,
      "step": 450
    },
    {
      "epoch": 0.368,
      "grad_norm": 3.8518083095550537,
      "learning_rate": 0.0001893610844546894,
      "loss": 7.7739,
      "step": 460
    },
    {
      "epoch": 0.376,
      "grad_norm": 4.104583263397217,
      "learning_rate": 0.00018876490553756313,
      "loss": 7.7344,
      "step": 470
    },
    {
      "epoch": 0.384,
      "grad_norm": 5.830111980438232,
      "learning_rate": 0.00018815346638650487,
      "loss": 7.4569,
      "step": 480
    },
    {
      "epoch": 0.392,
      "grad_norm": 5.279020309448242,
      "learning_rate": 0.0001875268721185585,
      "loss": 7.6329,
      "step": 490
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.486227989196777,
      "learning_rate": 0.00018688523045619674,
      "loss": 7.6998,
      "step": 500
    },
    {
      "epoch": 0.408,
      "grad_norm": 3.6189093589782715,
      "learning_rate": 0.00018622865170880151,
      "loss": 7.3692,
      "step": 510
    },
    {
      "epoch": 0.416,
      "grad_norm": 4.793766498565674,
      "learning_rate": 0.00018555724875369997,
      "loss": 7.4387,
      "step": 520
    },
    {
      "epoch": 0.424,
      "grad_norm": 3.8143932819366455,
      "learning_rate": 0.00018487113701675883,
      "loss": 7.6311,
      "step": 530
    },
    {
      "epoch": 0.432,
      "grad_norm": 5.568665027618408,
      "learning_rate": 0.00018417043445254075,
      "loss": 7.0967,
      "step": 540
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.059378623962402,
      "learning_rate": 0.00018345526152402573,
      "loss": 7.2701,
      "step": 550
    },
    {
      "epoch": 0.448,
      "grad_norm": 5.139848232269287,
      "learning_rate": 0.00018272574118190167,
      "loss": 7.2647,
      "step": 560
    },
    {
      "epoch": 0.456,
      "grad_norm": 6.435779571533203,
      "learning_rate": 0.00018198199884342673,
      "loss": 7.161,
      "step": 570
    },
    {
      "epoch": 0.464,
      "grad_norm": 5.199296951293945,
      "learning_rate": 0.0001812241623708682,
      "loss": 7.5137,
      "step": 580
    },
    {
      "epoch": 0.472,
      "grad_norm": 4.161045074462891,
      "learning_rate": 0.00018045236204952044,
      "loss": 7.356,
      "step": 590
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.084685325622559,
      "learning_rate": 0.00017966673056530686,
      "loss": 7.3859,
      "step": 600
    },
    {
      "epoch": 0.488,
      "grad_norm": 7.625977516174316,
      "learning_rate": 0.00017886740298196863,
      "loss": 7.0858,
      "step": 610
    },
    {
      "epoch": 0.496,
      "grad_norm": 5.039551734924316,
      "learning_rate": 0.00017805451671784516,
      "loss": 7.4888,
      "step": 620
    },
    {
      "epoch": 0.504,
      "grad_norm": 5.905925273895264,
      "learning_rate": 0.00017722821152224924,
      "loss": 7.2393,
      "step": 630
    },
    {
      "epoch": 0.512,
      "grad_norm": 9.78266716003418,
      "learning_rate": 0.00017638862945144182,
      "loss": 7.4102,
      "step": 640
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.097099781036377,
      "learning_rate": 0.00017553591484421004,
      "loss": 7.297,
      "step": 650
    },
    {
      "epoch": 0.528,
      "grad_norm": 4.922943115234375,
      "learning_rate": 0.00017467021429705285,
      "loss": 7.3741,
      "step": 660
    },
    {
      "epoch": 0.536,
      "grad_norm": 5.79689884185791,
      "learning_rate": 0.00017379167663897856,
      "loss": 7.1571,
      "step": 670
    },
    {
      "epoch": 0.544,
      "grad_norm": 3.795285701751709,
      "learning_rate": 0.00017290045290591858,
      "loss": 6.8519,
      "step": 680
    },
    {
      "epoch": 0.552,
      "grad_norm": 4.697098731994629,
      "learning_rate": 0.0001719966963147616,
      "loss": 7.111,
      "step": 690
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.46597146987915,
      "learning_rate": 0.000171080562237013,
      "loss": 7.0099,
      "step": 700
    },
    {
      "epoch": 0.568,
      "grad_norm": 6.007630348205566,
      "learning_rate": 0.00017015220817208376,
      "loss": 7.326,
      "step": 710
    },
    {
      "epoch": 0.576,
      "grad_norm": 4.53595495223999,
      "learning_rate": 0.0001692117937202136,
      "loss": 7.0785,
      "step": 720
    },
    {
      "epoch": 0.584,
      "grad_norm": 6.850202560424805,
      "learning_rate": 0.00016825948055503294,
      "loss": 7.2174,
      "step": 730
    },
    {
      "epoch": 0.592,
      "grad_norm": 5.931501865386963,
      "learning_rate": 0.00016729543239576828,
      "loss": 6.995,
      "step": 740
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.900391101837158,
      "learning_rate": 0.0001663198149790961,
      "loss": 6.8995,
      "step": 750
    },
    {
      "epoch": 0.608,
      "grad_norm": 5.841864109039307,
      "learning_rate": 0.00016533279603064978,
      "loss": 6.7632,
      "step": 760
    },
    {
      "epoch": 0.616,
      "grad_norm": 5.902273178100586,
      "learning_rate": 0.00016433454523618482,
      "loss": 6.9055,
      "step": 770
    },
    {
      "epoch": 0.624,
      "grad_norm": 5.881319046020508,
      "learning_rate": 0.00016332523421240658,
      "loss": 6.686,
      "step": 780
    },
    {
      "epoch": 0.632,
      "grad_norm": 5.178507328033447,
      "learning_rate": 0.00016230503647746657,
      "loss": 6.6281,
      "step": 790
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.807728290557861,
      "learning_rate": 0.00016127412742113185,
      "loss": 6.9373,
      "step": 800
    },
    {
      "epoch": 0.648,
      "grad_norm": 5.6972975730896,
      "learning_rate": 0.000160232684274632,
      "loss": 6.7103,
      "step": 810
    },
    {
      "epoch": 0.656,
      "grad_norm": 5.271605968475342,
      "learning_rate": 0.00015918088608019043,
      "loss": 6.9209,
      "step": 820
    },
    {
      "epoch": 0.664,
      "grad_norm": 5.366481304168701,
      "learning_rate": 0.00015811891366024358,
      "loss": 6.9491,
      "step": 830
    },
    {
      "epoch": 0.672,
      "grad_norm": 4.8482866287231445,
      "learning_rate": 0.00015704694958635468,
      "loss": 6.6238,
      "step": 840
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.680044651031494,
      "learning_rate": 0.0001559651781478263,
      "loss": 6.7933,
      "step": 850
    },
    {
      "epoch": 0.688,
      "grad_norm": 4.3361496925354,
      "learning_rate": 0.00015487378532001782,
      "loss": 6.6105,
      "step": 860
    },
    {
      "epoch": 0.696,
      "grad_norm": 8.266775131225586,
      "learning_rate": 0.0001537729587323732,
      "loss": 6.5274,
      "step": 870
    },
    {
      "epoch": 0.704,
      "grad_norm": 5.751520156860352,
      "learning_rate": 0.00015266288763616403,
      "loss": 6.6585,
      "step": 880
    },
    {
      "epoch": 0.712,
      "grad_norm": 6.401413440704346,
      "learning_rate": 0.000151543762871954,
      "loss": 6.5739,
      "step": 890
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.519962310791016,
      "learning_rate": 0.0001504157768367901,
      "loss": 6.6899,
      "step": 900
    },
    {
      "epoch": 0.728,
      "grad_norm": 5.094054222106934,
      "learning_rate": 0.00014927912345112616,
      "loss": 6.2008,
      "step": 910
    },
    {
      "epoch": 0.736,
      "grad_norm": 6.107059955596924,
      "learning_rate": 0.0001481339981254846,
      "loss": 6.2149,
      "step": 920
    },
    {
      "epoch": 0.744,
      "grad_norm": 6.355636119842529,
      "learning_rate": 0.00014698059772686202,
      "loss": 6.7521,
      "step": 930
    },
    {
      "epoch": 0.752,
      "grad_norm": 7.230486869812012,
      "learning_rate": 0.00014581912054488413,
      "loss": 6.3823,
      "step": 940
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.960805416107178,
      "learning_rate": 0.00014464976625771654,
      "loss": 6.6149,
      "step": 950
    },
    {
      "epoch": 0.768,
      "grad_norm": 5.269943714141846,
      "learning_rate": 0.00014347273589773637,
      "loss": 6.1001,
      "step": 960
    },
    {
      "epoch": 0.776,
      "grad_norm": 6.840855598449707,
      "learning_rate": 0.0001422882318169716,
      "loss": 6.3677,
      "step": 970
    },
    {
      "epoch": 0.784,
      "grad_norm": 5.1594038009643555,
      "learning_rate": 0.00014109645765231278,
      "loss": 6.3929,
      "step": 980
    },
    {
      "epoch": 0.792,
      "grad_norm": 5.096086025238037,
      "learning_rate": 0.00013989761829050475,
      "loss": 6.1354,
      "step": 990
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.235525608062744,
      "learning_rate": 0.00013869191983292283,
      "loss": 6.4954,
      "step": 1000
    },
    {
      "epoch": 0.808,
      "grad_norm": 5.518918991088867,
      "learning_rate": 0.00013747956956014037,
      "loss": 6.449,
      "step": 1010
    },
    {
      "epoch": 0.816,
      "grad_norm": 4.848990440368652,
      "learning_rate": 0.00013626077589629367,
      "loss": 6.392,
      "step": 1020
    },
    {
      "epoch": 0.824,
      "grad_norm": 7.234468460083008,
      "learning_rate": 0.00013503574837325015,
      "loss": 6.5465,
      "step": 1030
    },
    {
      "epoch": 0.832,
      "grad_norm": 6.593731880187988,
      "learning_rate": 0.00013380469759458643,
      "loss": 6.574,
      "step": 1040
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.687368392944336,
      "learning_rate": 0.00013256783519938154,
      "loss": 6.1995,
      "step": 1050
    },
    {
      "epoch": 0.848,
      "grad_norm": 4.857635498046875,
      "learning_rate": 0.00013132537382583274,
      "loss": 5.8422,
      "step": 1060
    },
    {
      "epoch": 0.856,
      "grad_norm": 7.068734645843506,
      "learning_rate": 0.00013007752707469924,
      "loss": 6.0601,
      "step": 1070
    },
    {
      "epoch": 0.864,
      "grad_norm": 4.396754741668701,
      "learning_rate": 0.00012882450947258045,
      "loss": 5.8387,
      "step": 1080
    },
    {
      "epoch": 0.872,
      "grad_norm": 9.501909255981445,
      "learning_rate": 0.0001275665364350352,
      "loss": 5.9831,
      "step": 1090
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.957056522369385,
      "learning_rate": 0.00012630382422954795,
      "loss": 6.1359,
      "step": 1100
    },
    {
      "epoch": 0.888,
      "grad_norm": 5.782343864440918,
      "learning_rate": 0.00012503658993834885,
      "loss": 6.0754,
      "step": 1110
    },
    {
      "epoch": 0.896,
      "grad_norm": 5.452831268310547,
      "learning_rate": 0.0001237650514210932,
      "loss": 5.6186,
      "step": 1120
    },
    {
      "epoch": 0.904,
      "grad_norm": 6.382038593292236,
      "learning_rate": 0.00012248942727740783,
      "loss": 5.7174,
      "step": 1130
    },
    {
      "epoch": 0.912,
      "grad_norm": 6.288851261138916,
      "learning_rate": 0.00012120993680931003,
      "loss": 5.6529,
      "step": 1140
    },
    {
      "epoch": 0.92,
      "grad_norm": 6.7387166023254395,
      "learning_rate": 0.0001199267999835055,
      "loss": 5.603,
      "step": 1150
    },
    {
      "epoch": 0.928,
      "grad_norm": 5.694065093994141,
      "learning_rate": 0.00011864023739357235,
      "loss": 5.2627,
      "step": 1160
    },
    {
      "epoch": 0.936,
      "grad_norm": 6.711731910705566,
      "learning_rate": 0.00011735047022203741,
      "loss": 5.4706,
      "step": 1170
    },
    {
      "epoch": 0.944,
      "grad_norm": 5.517411708831787,
      "learning_rate": 0.00011605772020235072,
      "loss": 5.6277,
      "step": 1180
    },
    {
      "epoch": 0.952,
      "grad_norm": 6.785055160522461,
      "learning_rate": 0.00011476220958076607,
      "loss": 5.9611,
      "step": 1190
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.702793121337891,
      "learning_rate": 0.00011346416107813267,
      "loss": 5.9226,
      "step": 1200
    },
    {
      "epoch": 0.968,
      "grad_norm": 7.974459648132324,
      "learning_rate": 0.00011216379785160578,
      "loss": 5.8474,
      "step": 1210
    },
    {
      "epoch": 0.976,
      "grad_norm": 5.838559150695801,
      "learning_rate": 0.000110861343456282,
      "loss": 6.0194,
      "step": 1220
    },
    {
      "epoch": 0.984,
      "grad_norm": 7.7551703453063965,
      "learning_rate": 0.00010955702180676632,
      "loss": 5.7078,
      "step": 1230
    },
    {
      "epoch": 0.992,
      "grad_norm": 7.658422470092773,
      "learning_rate": 0.00010825105713867724,
      "loss": 5.6615,
      "step": 1240
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.197235584259033,
      "learning_rate": 0.0001069436739700968,
      "loss": 5.7792,
      "step": 1250
    },
    {
      "epoch": 1.008,
      "grad_norm": 6.991013526916504,
      "learning_rate": 0.00010563509706297188,
      "loss": 5.3963,
      "step": 1260
    },
    {
      "epoch": 1.016,
      "grad_norm": 7.606940269470215,
      "learning_rate": 0.00010432555138447404,
      "loss": 5.9168,
      "step": 1270
    },
    {
      "epoch": 1.024,
      "grad_norm": 6.3427042961120605,
      "learning_rate": 0.0001030152620683233,
      "loss": 5.6694,
      "step": 1280
    },
    {
      "epoch": 1.032,
      "grad_norm": 5.197830677032471,
      "learning_rate": 0.00010170445437608403,
      "loss": 5.4628,
      "step": 1290
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.1471381187438965,
      "learning_rate": 0.00010039335365843851,
      "loss": 5.4371,
      "step": 1300
    },
    {
      "epoch": 1.048,
      "grad_norm": 6.546257495880127,
      "learning_rate": 9.908218531644521e-05,
      "loss": 5.7931,
      "step": 1310
    },
    {
      "epoch": 1.056,
      "grad_norm": 8.735387802124023,
      "learning_rate": 9.77711747627883e-05,
      "loss": 5.354,
      "step": 1320
    },
    {
      "epoch": 1.064,
      "grad_norm": 6.352960109710693,
      "learning_rate": 9.646054738302551e-05,
      "loss": 5.0677,
      "step": 1330
    },
    {
      "epoch": 1.072,
      "grad_norm": 7.88266658782959,
      "learning_rate": 9.515052849684019e-05,
      "loss": 5.4004,
      "step": 1340
    },
    {
      "epoch": 1.08,
      "grad_norm": 5.571359634399414,
      "learning_rate": 9.384134331930513e-05,
      "loss": 5.3187,
      "step": 1350
    },
    {
      "epoch": 1.088,
      "grad_norm": 8.573116302490234,
      "learning_rate": 9.253321692216402e-05,
      "loss": 5.2719,
      "step": 1360
    },
    {
      "epoch": 1.096,
      "grad_norm": 9.162277221679688,
      "learning_rate": 9.122637419513778e-05,
      "loss": 5.6781,
      "step": 1370
    },
    {
      "epoch": 1.104,
      "grad_norm": 11.947822570800781,
      "learning_rate": 8.992103980726207e-05,
      "loss": 5.593,
      "step": 1380
    },
    {
      "epoch": 1.112,
      "grad_norm": 8.124156951904297,
      "learning_rate": 8.861743816826274e-05,
      "loss": 5.3142,
      "step": 1390
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.134088039398193,
      "learning_rate": 8.731579338997594e-05,
      "loss": 5.3372,
      "step": 1400
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 8.928452491760254,
      "learning_rate": 8.601632924781935e-05,
      "loss": 5.4594,
      "step": 1410
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 5.617236614227295,
      "learning_rate": 8.471926914232137e-05,
      "loss": 5.0226,
      "step": 1420
    },
    {
      "epoch": 1.144,
      "grad_norm": 6.080244064331055,
      "learning_rate": 8.34248360607145e-05,
      "loss": 5.3677,
      "step": 1430
    },
    {
      "epoch": 1.152,
      "grad_norm": 9.846885681152344,
      "learning_rate": 8.213325253860013e-05,
      "loss": 5.2629,
      "step": 1440
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.537776947021484,
      "learning_rate": 8.084474062169071e-05,
      "loss": 5.5022,
      "step": 1450
    },
    {
      "epoch": 1.168,
      "grad_norm": 5.898927688598633,
      "learning_rate": 7.955952182763624e-05,
      "loss": 5.2349,
      "step": 1460
    },
    {
      "epoch": 1.176,
      "grad_norm": 6.621062755584717,
      "learning_rate": 7.827781710794146e-05,
      "loss": 5.4929,
      "step": 1470
    },
    {
      "epoch": 1.184,
      "grad_norm": 6.793276309967041,
      "learning_rate": 7.699984680998063e-05,
      "loss": 5.136,
      "step": 1480
    },
    {
      "epoch": 1.192,
      "grad_norm": 8.44278335571289,
      "learning_rate": 7.57258306391157e-05,
      "loss": 4.9147,
      "step": 1490
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.787193298339844,
      "learning_rate": 7.445598762092537e-05,
      "loss": 5.4473,
      "step": 1500
    },
    {
      "epoch": 1.208,
      "grad_norm": 7.027050495147705,
      "learning_rate": 7.319053606355061e-05,
      "loss": 5.1231,
      "step": 1510
    },
    {
      "epoch": 1.216,
      "grad_norm": 6.161197662353516,
      "learning_rate": 7.192969352016383e-05,
      "loss": 5.1066,
      "step": 1520
    },
    {
      "epoch": 1.224,
      "grad_norm": 5.8758745193481445,
      "learning_rate": 7.067367675156758e-05,
      "loss": 5.1743,
      "step": 1530
    },
    {
      "epoch": 1.232,
      "grad_norm": 8.138237953186035,
      "learning_rate": 6.942270168892959e-05,
      "loss": 5.2096,
      "step": 1540
    },
    {
      "epoch": 1.24,
      "grad_norm": 7.498131275177002,
      "learning_rate": 6.817698339666066e-05,
      "loss": 5.1718,
      "step": 1550
    },
    {
      "epoch": 1.248,
      "grad_norm": 5.885447025299072,
      "learning_rate": 6.693673603544097e-05,
      "loss": 5.4034,
      "step": 1560
    },
    {
      "epoch": 1.256,
      "grad_norm": 5.564108848571777,
      "learning_rate": 6.570217282540247e-05,
      "loss": 5.0762,
      "step": 1570
    },
    {
      "epoch": 1.264,
      "grad_norm": 5.904695510864258,
      "learning_rate": 6.447350600947236e-05,
      "loss": 4.9994,
      "step": 1580
    },
    {
      "epoch": 1.272,
      "grad_norm": 6.91570520401001,
      "learning_rate": 6.325094681688503e-05,
      "loss": 5.1286,
      "step": 1590
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.768303394317627,
      "learning_rate": 6.203470542686798e-05,
      "loss": 4.975,
      "step": 1600
    },
    {
      "epoch": 1.288,
      "grad_norm": 9.351409912109375,
      "learning_rate": 6.082499093250831e-05,
      "loss": 4.9613,
      "step": 1610
    },
    {
      "epoch": 1.296,
      "grad_norm": 9.572948455810547,
      "learning_rate": 5.962201130480618e-05,
      "loss": 4.7479,
      "step": 1620
    },
    {
      "epoch": 1.304,
      "grad_norm": 7.169508457183838,
      "learning_rate": 5.842597335692067e-05,
      "loss": 5.2066,
      "step": 1630
    },
    {
      "epoch": 1.312,
      "grad_norm": 7.029344081878662,
      "learning_rate": 5.7237082708615186e-05,
      "loss": 4.7997,
      "step": 1640
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.824025630950928,
      "learning_rate": 5.605554375090784e-05,
      "loss": 4.6027,
      "step": 1650
    },
    {
      "epoch": 1.328,
      "grad_norm": 12.187256813049316,
      "learning_rate": 5.488155961093298e-05,
      "loss": 5.03,
      "step": 1660
    },
    {
      "epoch": 1.336,
      "grad_norm": 5.913212776184082,
      "learning_rate": 5.3715332117020154e-05,
      "loss": 4.7387,
      "step": 1670
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 6.245068550109863,
      "learning_rate": 5.2557061763996354e-05,
      "loss": 5.0907,
      "step": 1680
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 9.76169490814209,
      "learning_rate": 5.140694767871733e-05,
      "loss": 5.1469,
      "step": 1690
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 10.653762817382812,
      "learning_rate": 5.026518758583434e-05,
      "loss": 4.9712,
      "step": 1700
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 7.157083511352539,
      "learning_rate": 4.9131977773801765e-05,
      "loss": 4.6749,
      "step": 1710
    },
    {
      "epoch": 1.376,
      "grad_norm": 8.09749698638916,
      "learning_rate": 4.8007513061131736e-05,
      "loss": 4.5003,
      "step": 1720
    },
    {
      "epoch": 1.384,
      "grad_norm": 6.536433219909668,
      "learning_rate": 4.6891986762901406e-05,
      "loss": 4.2644,
      "step": 1730
    },
    {
      "epoch": 1.392,
      "grad_norm": 7.866100311279297,
      "learning_rate": 4.578559065751873e-05,
      "loss": 4.721,
      "step": 1740
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.806945323944092,
      "learning_rate": 4.4688514953752515e-05,
      "loss": 4.6537,
      "step": 1750
    },
    {
      "epoch": 1.408,
      "grad_norm": 8.229887962341309,
      "learning_rate": 4.360094825803203e-05,
      "loss": 4.7068,
      "step": 1760
    },
    {
      "epoch": 1.416,
      "grad_norm": 10.401880264282227,
      "learning_rate": 4.25230775420224e-05,
      "loss": 4.5295,
      "step": 1770
    },
    {
      "epoch": 1.424,
      "grad_norm": 7.166602611541748,
      "learning_rate": 4.145508811048089e-05,
      "loss": 4.5715,
      "step": 1780
    },
    {
      "epoch": 1.432,
      "grad_norm": 8.82656478881836,
      "learning_rate": 4.039716356939981e-05,
      "loss": 4.8924,
      "step": 1790
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.747090339660645,
      "learning_rate": 3.9349485794441395e-05,
      "loss": 4.5401,
      "step": 1800
    },
    {
      "epoch": 1.448,
      "grad_norm": 12.537016868591309,
      "learning_rate": 3.831223489967025e-05,
      "loss": 4.7768,
      "step": 1810
    },
    {
      "epoch": 1.456,
      "grad_norm": 6.526289939880371,
      "learning_rate": 3.728558920658868e-05,
      "loss": 4.4916,
      "step": 1820
    },
    {
      "epoch": 1.464,
      "grad_norm": 7.882378101348877,
      "learning_rate": 3.6269725213479846e-05,
      "loss": 4.5072,
      "step": 1830
    },
    {
      "epoch": 1.472,
      "grad_norm": 10.795116424560547,
      "learning_rate": 3.526481756506498e-05,
      "loss": 4.7454,
      "step": 1840
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.107118606567383,
      "learning_rate": 3.4271039022478694e-05,
      "loss": 4.3667,
      "step": 1850
    },
    {
      "epoch": 1.488,
      "grad_norm": 8.725346565246582,
      "learning_rate": 3.328856043356837e-05,
      "loss": 4.8818,
      "step": 1860
    },
    {
      "epoch": 1.496,
      "grad_norm": 8.258055686950684,
      "learning_rate": 3.231755070352249e-05,
      "loss": 4.4017,
      "step": 1870
    },
    {
      "epoch": 1.504,
      "grad_norm": 9.932522773742676,
      "learning_rate": 3.1358176765832915e-05,
      "loss": 4.3575,
      "step": 1880
    },
    {
      "epoch": 1.512,
      "grad_norm": 7.825517654418945,
      "learning_rate": 3.041060355359594e-05,
      "loss": 4.2122,
      "step": 1890
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.71921443939209,
      "learning_rate": 2.9474993971157605e-05,
      "loss": 4.4648,
      "step": 1900
    },
    {
      "epoch": 1.528,
      "grad_norm": 9.51145076751709,
      "learning_rate": 2.8551508866107514e-05,
      "loss": 4.4388,
      "step": 1910
    },
    {
      "epoch": 1.536,
      "grad_norm": 6.032364845275879,
      "learning_rate": 2.764030700162633e-05,
      "loss": 4.6544,
      "step": 1920
    },
    {
      "epoch": 1.544,
      "grad_norm": 6.442480564117432,
      "learning_rate": 2.6741545029191674e-05,
      "loss": 4.3281,
      "step": 1930
    },
    {
      "epoch": 1.552,
      "grad_norm": 9.142407417297363,
      "learning_rate": 2.5855377461646902e-05,
      "loss": 4.7258,
      "step": 1940
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.672632217407227,
      "learning_rate": 2.4981956646637815e-05,
      "loss": 4.076,
      "step": 1950
    },
    {
      "epoch": 1.568,
      "grad_norm": 10.532658576965332,
      "learning_rate": 2.412143274042129e-05,
      "loss": 4.3765,
      "step": 1960
    },
    {
      "epoch": 1.576,
      "grad_norm": 6.73195743560791,
      "learning_rate": 2.327395368205084e-05,
      "loss": 4.4419,
      "step": 1970
    },
    {
      "epoch": 1.584,
      "grad_norm": 7.2371745109558105,
      "learning_rate": 2.243966516794338e-05,
      "loss": 4.5395,
      "step": 1980
    },
    {
      "epoch": 1.592,
      "grad_norm": 9.002269744873047,
      "learning_rate": 2.161871062683145e-05,
      "loss": 4.5577,
      "step": 1990
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.184229850769043,
      "learning_rate": 2.0811231195105186e-05,
      "loss": 4.3631,
      "step": 2000
    },
    {
      "epoch": 1.608,
      "grad_norm": 7.962112903594971,
      "learning_rate": 2.0017365692548717e-05,
      "loss": 4.6822,
      "step": 2010
    },
    {
      "epoch": 1.616,
      "grad_norm": 7.3697190284729,
      "learning_rate": 1.9237250598474564e-05,
      "loss": 4.1389,
      "step": 2020
    },
    {
      "epoch": 1.624,
      "grad_norm": 8.01472282409668,
      "learning_rate": 1.8471020028260368e-05,
      "loss": 3.9598,
      "step": 2030
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 5.376262187957764,
      "learning_rate": 1.7718805710292208e-05,
      "loss": 4.3192,
      "step": 2040
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 10.927371978759766,
      "learning_rate": 1.6980736963318177e-05,
      "loss": 4.192,
      "step": 2050
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 6.14822244644165,
      "learning_rate": 1.625694067421626e-05,
      "loss": 4.4155,
      "step": 2060
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 8.27953815460205,
      "learning_rate": 1.554754127618019e-05,
      "loss": 4.5526,
      "step": 2070
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 6.160037517547607,
      "learning_rate": 1.4852660727327361e-05,
      "loss": 4.453,
      "step": 2080
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 7.098100662231445,
      "learning_rate": 1.4172418489731942e-05,
      "loss": 4.4741,
      "step": 2090
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 6.903653144836426,
      "learning_rate": 1.3506931508887333e-05,
      "loss": 4.0234,
      "step": 2100
    },
    {
      "epoch": 1.688,
      "grad_norm": 6.967901706695557,
      "learning_rate": 1.2856314193601216e-05,
      "loss": 4.2137,
      "step": 2110
    },
    {
      "epoch": 1.696,
      "grad_norm": 7.635384559631348,
      "learning_rate": 1.2220678396326678e-05,
      "loss": 4.38,
      "step": 2120
    },
    {
      "epoch": 1.704,
      "grad_norm": 9.351762771606445,
      "learning_rate": 1.160013339393281e-05,
      "loss": 4.4418,
      "step": 2130
    },
    {
      "epoch": 1.712,
      "grad_norm": 7.01410436630249,
      "learning_rate": 1.0994785868918101e-05,
      "loss": 4.4396,
      "step": 2140
    },
    {
      "epoch": 1.72,
      "grad_norm": 6.974796772003174,
      "learning_rate": 1.040473989106988e-05,
      "loss": 4.5075,
      "step": 2150
    },
    {
      "epoch": 1.728,
      "grad_norm": 7.545105934143066,
      "learning_rate": 9.830096899572927e-06,
      "loss": 4.7132,
      "step": 2160
    },
    {
      "epoch": 1.736,
      "grad_norm": 7.074202537536621,
      "learning_rate": 9.270955685570226e-06,
      "loss": 4.2393,
      "step": 2170
    },
    {
      "epoch": 1.744,
      "grad_norm": 6.088277339935303,
      "learning_rate": 8.727412375179156e-06,
      "loss": 4.8092,
      "step": 2180
    },
    {
      "epoch": 1.752,
      "grad_norm": 5.859469413757324,
      "learning_rate": 8.199560412965634e-06,
      "loss": 4.6403,
      "step": 2190
    },
    {
      "epoch": 1.76,
      "grad_norm": 6.770120143890381,
      "learning_rate": 7.687490545879461e-06,
      "loss": 4.2544,
      "step": 2200
    },
    {
      "epoch": 1.768,
      "grad_norm": 8.092095375061035,
      "learning_rate": 7.191290807653251e-06,
      "loss": 4.454,
      "step": 2210
    },
    {
      "epoch": 1.776,
      "grad_norm": 6.498937606811523,
      "learning_rate": 6.711046503667983e-06,
      "loss": 4.0243,
      "step": 2220
    },
    {
      "epoch": 1.784,
      "grad_norm": 5.894200801849365,
      "learning_rate": 6.24684019628744e-06,
      "loss": 4.0666,
      "step": 2230
    },
    {
      "epoch": 1.792,
      "grad_norm": 6.309505462646484,
      "learning_rate": 5.79875169066435e-06,
      "loss": 4.2441,
      "step": 2240
    },
    {
      "epoch": 1.8,
      "grad_norm": 6.20737886428833,
      "learning_rate": 5.366858021020471e-06,
      "loss": 4.1951,
      "step": 2250
    },
    {
      "epoch": 1.808,
      "grad_norm": 8.314438819885254,
      "learning_rate": 4.951233437403102e-06,
      "loss": 4.4043,
      "step": 2260
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 6.5354108810424805,
      "learning_rate": 4.551949392920118e-06,
      "loss": 4.1528,
      "step": 2270
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 8.503190040588379,
      "learning_rate": 4.169074531456063e-06,
      "loss": 3.8358,
      "step": 2280
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 7.091247081756592,
      "learning_rate": 3.802674675870932e-06,
      "loss": 4.3702,
      "step": 2290
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 6.265818119049072,
      "learning_rate": 3.4528128166842033e-06,
      "loss": 4.3354,
      "step": 2300
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 7.026565074920654,
      "learning_rate": 3.119549101245567e-06,
      "loss": 3.9832,
      "step": 2310
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 6.407781600952148,
      "learning_rate": 2.8029408233946177e-06,
      "loss": 4.1522,
      "step": 2320
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 13.10364055633545,
      "learning_rate": 2.503042413611001e-06,
      "loss": 4.354,
      "step": 2330
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 12.458951950073242,
      "learning_rate": 2.219905429656899e-06,
      "loss": 4.5043,
      "step": 2340
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.545763969421387,
      "learning_rate": 1.9535785477133195e-06,
      "loss": 4.2234,
      "step": 2350
    },
    {
      "epoch": 1.888,
      "grad_norm": 9.243309020996094,
      "learning_rate": 1.7041075540118578e-06,
      "loss": 4.2382,
      "step": 2360
    },
    {
      "epoch": 1.896,
      "grad_norm": 6.87458610534668,
      "learning_rate": 1.4715353369631924e-06,
      "loss": 3.9852,
      "step": 2370
    },
    {
      "epoch": 1.904,
      "grad_norm": 6.957976341247559,
      "learning_rate": 1.2559018797838384e-06,
      "loss": 3.9335,
      "step": 2380
    },
    {
      "epoch": 1.912,
      "grad_norm": 5.689143657684326,
      "learning_rate": 1.0572442536223692e-06,
      "loss": 4.0,
      "step": 2390
    },
    {
      "epoch": 1.92,
      "grad_norm": 5.896793842315674,
      "learning_rate": 8.755966111861913e-07,
      "loss": 4.2807,
      "step": 2400
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}