{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1176,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025510204081632654,
      "grad_norm": 10.153619766235352,
      "learning_rate": 3.813559322033899e-07,
      "loss": 0.8792,
      "step": 10
    },
    {
      "epoch": 0.05102040816326531,
      "grad_norm": 4.951279640197754,
      "learning_rate": 8.050847457627118e-07,
      "loss": 0.8689,
      "step": 20
    },
    {
      "epoch": 0.07653061224489796,
      "grad_norm": 2.7297706604003906,
      "learning_rate": 1.228813559322034e-06,
      "loss": 0.8211,
      "step": 30
    },
    {
      "epoch": 0.10204081632653061,
      "grad_norm": 2.159283399581909,
      "learning_rate": 1.652542372881356e-06,
      "loss": 0.7512,
      "step": 40
    },
    {
      "epoch": 0.12755102040816327,
      "grad_norm": 2.1231689453125,
      "learning_rate": 2.076271186440678e-06,
      "loss": 0.6971,
      "step": 50
    },
    {
      "epoch": 0.15306122448979592,
      "grad_norm": 1.6915334463119507,
      "learning_rate": 2.5e-06,
      "loss": 0.6312,
      "step": 60
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.6702977418899536,
      "learning_rate": 2.9237288135593224e-06,
      "loss": 0.6424,
      "step": 70
    },
    {
      "epoch": 0.20408163265306123,
      "grad_norm": 1.6608940362930298,
      "learning_rate": 3.347457627118644e-06,
      "loss": 0.614,
      "step": 80
    },
    {
      "epoch": 0.22959183673469388,
      "grad_norm": 1.7687163352966309,
      "learning_rate": 3.7711864406779664e-06,
      "loss": 0.6052,
      "step": 90
    },
    {
      "epoch": 0.25510204081632654,
      "grad_norm": 1.6226813793182373,
      "learning_rate": 4.1949152542372886e-06,
      "loss": 0.5939,
      "step": 100
    },
    {
      "epoch": 0.28061224489795916,
      "grad_norm": 1.785929799079895,
      "learning_rate": 4.618644067796611e-06,
      "loss": 0.5806,
      "step": 110
    },
    {
      "epoch": 0.30612244897959184,
      "grad_norm": 1.9804999828338623,
      "learning_rate": 4.99998897856601e-06,
      "loss": 0.5516,
      "step": 120
    },
    {
      "epoch": 0.33163265306122447,
      "grad_norm": 1.6149433851242065,
      "learning_rate": 4.998666524068091e-06,
      "loss": 0.554,
      "step": 130
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 2.026740789413452,
      "learning_rate": 4.995141118770565e-06,
      "loss": 0.5455,
      "step": 140
    },
    {
      "epoch": 0.3826530612244898,
      "grad_norm": 1.9442322254180908,
      "learning_rate": 4.9894158708490694e-06,
      "loss": 0.5624,
      "step": 150
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 1.6927114725112915,
      "learning_rate": 4.98149582797178e-06,
      "loss": 0.5336,
      "step": 160
    },
    {
      "epoch": 0.4336734693877551,
      "grad_norm": 2.1999094486236572,
      "learning_rate": 4.9713879728491196e-06,
      "loss": 0.5282,
      "step": 170
    },
    {
      "epoch": 0.45918367346938777,
      "grad_norm": 1.8567699193954468,
      "learning_rate": 4.9591012170774605e-06,
      "loss": 0.5457,
      "step": 180
    },
    {
      "epoch": 0.4846938775510204,
      "grad_norm": 1.795466661453247,
      "learning_rate": 4.9446463932822e-06,
      "loss": 0.52,
      "step": 190
    },
    {
      "epoch": 0.5102040816326531,
      "grad_norm": 1.7791986465454102,
      "learning_rate": 4.928036245567176e-06,
      "loss": 0.5115,
      "step": 200
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 1.715412974357605,
      "learning_rate": 4.909285418278818e-06,
      "loss": 0.5345,
      "step": 210
    },
    {
      "epoch": 0.5612244897959183,
      "grad_norm": 1.9294530153274536,
      "learning_rate": 4.888410443094951e-06,
      "loss": 0.5136,
      "step": 220
    },
    {
      "epoch": 0.5867346938775511,
      "grad_norm": 1.6017546653747559,
      "learning_rate": 4.865429724449632e-06,
      "loss": 0.5226,
      "step": 230
    },
    {
      "epoch": 0.6122448979591837,
      "grad_norm": 1.7661460638046265,
      "learning_rate": 4.840363523306871e-06,
      "loss": 0.5273,
      "step": 240
    },
    {
      "epoch": 0.6377551020408163,
      "grad_norm": 1.8540369272232056,
      "learning_rate": 4.8132339392975384e-06,
      "loss": 0.5054,
      "step": 250
    },
    {
      "epoch": 0.6632653061224489,
      "grad_norm": 1.7291237115859985,
      "learning_rate": 4.784064891235212e-06,
      "loss": 0.5106,
      "step": 260
    },
    {
      "epoch": 0.6887755102040817,
      "grad_norm": 1.4396079778671265,
      "learning_rate": 4.752882096028146e-06,
      "loss": 0.5003,
      "step": 270
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.8574790954589844,
      "learning_rate": 4.7197130460059385e-06,
      "loss": 0.5062,
      "step": 280
    },
    {
      "epoch": 0.7397959183673469,
      "grad_norm": 1.5953553915023804,
      "learning_rate": 4.684586984680904e-06,
      "loss": 0.5085,
      "step": 290
    },
    {
      "epoch": 0.7653061224489796,
      "grad_norm": 1.7809573411941528,
      "learning_rate": 4.647534880965515e-06,
      "loss": 0.4909,
      "step": 300
    },
    {
      "epoch": 0.7908163265306123,
      "grad_norm": 1.8650221824645996,
      "learning_rate": 4.6085894018686435e-06,
      "loss": 0.4673,
      "step": 310
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 1.6842070817947388,
      "learning_rate": 4.567784883694671e-06,
      "loss": 0.4776,
      "step": 320
    },
    {
      "epoch": 0.8418367346938775,
      "grad_norm": 1.6628310680389404,
      "learning_rate": 4.525157301770865e-06,
      "loss": 0.5074,
      "step": 330
    },
    {
      "epoch": 0.8673469387755102,
      "grad_norm": 1.7108631134033203,
      "learning_rate": 4.4807442387297145e-06,
      "loss": 0.5055,
      "step": 340
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 1.7590484619140625,
      "learning_rate": 4.434584851374176e-06,
      "loss": 0.489,
      "step": 350
    },
    {
      "epoch": 0.9183673469387755,
      "grad_norm": 1.82199227809906,
      "learning_rate": 4.38671983615506e-06,
      "loss": 0.4792,
      "step": 360
    },
    {
      "epoch": 0.9438775510204082,
      "grad_norm": 1.6996486186981201,
      "learning_rate": 4.337191393290975e-06,
      "loss": 0.477,
      "step": 370
    },
    {
      "epoch": 0.9693877551020408,
      "grad_norm": 1.6575571298599243,
      "learning_rate": 4.286043189562487e-06,
      "loss": 0.5063,
      "step": 380
    },
    {
      "epoch": 0.9948979591836735,
      "grad_norm": 1.944182276725769,
      "learning_rate": 4.233320319813271e-06,
      "loss": 0.4818,
      "step": 390
    },
    {
      "epoch": 1.0204081632653061,
      "grad_norm": 1.632175326347351,
      "learning_rate": 4.1790692671922205e-06,
      "loss": 0.4184,
      "step": 400
    },
    {
      "epoch": 1.0459183673469388,
      "grad_norm": 1.8445144891738892,
      "learning_rate": 4.12333786217155e-06,
      "loss": 0.366,
      "step": 410
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 1.9345996379852295,
      "learning_rate": 4.06617524037703e-06,
      "loss": 0.4001,
      "step": 420
    },
    {
      "epoch": 1.096938775510204,
      "grad_norm": 1.8828767538070679,
      "learning_rate": 4.007631799267536e-06,
      "loss": 0.3906,
      "step": 430
    },
    {
      "epoch": 1.1224489795918366,
      "grad_norm": 1.8887412548065186,
      "learning_rate": 3.947759153702097e-06,
      "loss": 0.3783,
      "step": 440
    },
    {
      "epoch": 1.1479591836734695,
      "grad_norm": 2.035506010055542,
      "learning_rate": 3.88661009043363e-06,
      "loss": 0.3764,
      "step": 450
    },
    {
      "epoch": 1.1734693877551021,
      "grad_norm": 1.7684204578399658,
      "learning_rate": 3.824238521569462e-06,
      "loss": 0.3766,
      "step": 460
    },
    {
      "epoch": 1.1989795918367347,
      "grad_norm": 1.8956235647201538,
      "learning_rate": 3.7606994370396947e-06,
      "loss": 0.3896,
      "step": 470
    },
    {
      "epoch": 1.2244897959183674,
      "grad_norm": 1.7942086458206177,
      "learning_rate": 3.696048856115294e-06,
      "loss": 0.3838,
      "step": 480
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.8298710584640503,
      "learning_rate": 3.6303437780186745e-06,
      "loss": 0.3854,
      "step": 490
    },
    {
      "epoch": 1.2755102040816326,
      "grad_norm": 1.7128721475601196,
      "learning_rate": 3.5636421316702975e-06,
      "loss": 0.3617,
      "step": 500
    },
    {
      "epoch": 1.3010204081632653,
      "grad_norm": 1.7735494375228882,
      "learning_rate": 3.4960027246156043e-06,
      "loss": 0.3852,
      "step": 510
    },
    {
      "epoch": 1.3265306122448979,
      "grad_norm": 1.6894264221191406,
      "learning_rate": 3.427485191177311e-06,
      "loss": 0.3831,
      "step": 520
    },
    {
      "epoch": 1.3520408163265305,
      "grad_norm": 1.9257900714874268,
      "learning_rate": 3.358149939878772e-06,
      "loss": 0.3756,
      "step": 530
    },
    {
      "epoch": 1.3775510204081631,
      "grad_norm": 2.079662322998047,
      "learning_rate": 3.288058100184772e-06,
      "loss": 0.3799,
      "step": 540
    },
    {
      "epoch": 1.403061224489796,
      "grad_norm": 1.7804886102676392,
      "learning_rate": 3.217271468606694e-06,
      "loss": 0.3767,
      "step": 550
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.8842053413391113,
      "learning_rate": 3.1458524542195932e-06,
      "loss": 0.3591,
      "step": 560
    },
    {
      "epoch": 1.4540816326530612,
      "grad_norm": 1.6243324279785156,
      "learning_rate": 3.073864023639197e-06,
      "loss": 0.3904,
      "step": 570
    },
    {
      "epoch": 1.4795918367346939,
      "grad_norm": 1.8911751508712769,
      "learning_rate": 3.0013696455073504e-06,
      "loss": 0.3729,
      "step": 580
    },
    {
      "epoch": 1.5051020408163265,
      "grad_norm": 1.9872528314590454,
      "learning_rate": 2.9284332345348554e-06,
      "loss": 0.3748,
      "step": 590
    },
    {
      "epoch": 1.5306122448979593,
      "grad_norm": 1.7591272592544556,
      "learning_rate": 2.8551190951510225e-06,
      "loss": 0.3891,
      "step": 600
    },
    {
      "epoch": 1.556122448979592,
      "grad_norm": 2.0258169174194336,
      "learning_rate": 2.7814918648096408e-06,
      "loss": 0.3609,
      "step": 610
    },
    {
      "epoch": 1.5816326530612246,
      "grad_norm": 3.180363178253174,
      "learning_rate": 2.707616457001321e-06,
      "loss": 0.3775,
      "step": 620
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 2.0159478187561035,
      "learning_rate": 2.633558004022484e-06,
      "loss": 0.3734,
      "step": 630
    },
    {
      "epoch": 1.6326530612244898,
      "grad_norm": 1.8687129020690918,
      "learning_rate": 2.559381799551433e-06,
      "loss": 0.3649,
      "step": 640
    },
    {
      "epoch": 1.6581632653061225,
      "grad_norm": 1.742733120918274,
      "learning_rate": 2.4851532410821415e-06,
      "loss": 0.363,
      "step": 650
    },
    {
      "epoch": 1.683673469387755,
      "grad_norm": 1.7738467454910278,
      "learning_rate": 2.4109377722665175e-06,
      "loss": 0.3683,
      "step": 660
    },
    {
      "epoch": 1.7091836734693877,
      "grad_norm": 1.6583940982818604,
      "learning_rate": 2.3368008252159673e-06,
      "loss": 0.3733,
      "step": 670
    },
    {
      "epoch": 1.7346938775510203,
      "grad_norm": 1.785640835762024,
      "learning_rate": 2.262807762813137e-06,
      "loss": 0.3604,
      "step": 680
    },
    {
      "epoch": 1.760204081632653,
      "grad_norm": 2.008305072784424,
      "learning_rate": 2.189023821084689e-06,
      "loss": 0.3624,
      "step": 690
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 1.9097682237625122,
      "learning_rate": 2.1155140516859228e-06,
      "loss": 0.3632,
      "step": 700
    },
    {
      "epoch": 1.8112244897959182,
      "grad_norm": 1.626064419746399,
      "learning_rate": 2.042343264547941e-06,
      "loss": 0.3718,
      "step": 710
    },
    {
      "epoch": 1.836734693877551,
      "grad_norm": 1.9117207527160645,
      "learning_rate": 1.9695759707379406e-06,
      "loss": 0.3683,
      "step": 720
    },
    {
      "epoch": 1.8622448979591837,
      "grad_norm": 2.007272958755493,
      "learning_rate": 1.8972763255829876e-06,
      "loss": 0.381,
      "step": 730
    },
    {
      "epoch": 1.8877551020408163,
      "grad_norm": 1.879835605621338,
      "learning_rate": 1.8255080721074391e-06,
      "loss": 0.3782,
      "step": 740
    },
    {
      "epoch": 1.913265306122449,
      "grad_norm": 1.857400894165039,
      "learning_rate": 1.7543344848338705e-06,
      "loss": 0.3574,
      "step": 750
    },
    {
      "epoch": 1.9387755102040818,
      "grad_norm": 2.0696310997009277,
      "learning_rate": 1.6838183139970494e-06,
      "loss": 0.3731,
      "step": 760
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 1.7536237239837646,
      "learning_rate": 1.614021730220161e-06,
      "loss": 0.3604,
      "step": 770
    },
    {
      "epoch": 1.989795918367347,
      "grad_norm": 1.9241575002670288,
      "learning_rate": 1.5450062697020362e-06,
      "loss": 0.3762,
      "step": 780
    },
    {
      "epoch": 2.0153061224489797,
      "grad_norm": 1.6344062089920044,
      "learning_rate": 1.4768327799637314e-06,
      "loss": 0.3237,
      "step": 790
    },
    {
      "epoch": 2.0408163265306123,
      "grad_norm": 1.8893767595291138,
      "learning_rate": 1.4095613662022722e-06,
      "loss": 0.2731,
      "step": 800
    },
    {
      "epoch": 2.066326530612245,
      "grad_norm": 2.045369863510132,
      "learning_rate": 1.3432513382988726e-06,
      "loss": 0.2794,
      "step": 810
    },
    {
      "epoch": 2.0918367346938775,
      "grad_norm": 1.9920740127563477,
      "learning_rate": 1.2779611585283442e-06,
      "loss": 0.2642,
      "step": 820
    },
    {
      "epoch": 2.11734693877551,
      "grad_norm": 1.981793761253357,
      "learning_rate": 1.2137483900157964e-06,
      "loss": 0.2638,
      "step": 830
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 2.1838555335998535,
      "learning_rate": 1.1506696459860744e-06,
      "loss": 0.266,
      "step": 840
    },
    {
      "epoch": 2.1683673469387754,
      "grad_norm": 1.7283031940460205,
      "learning_rate": 1.0887805398506757e-06,
      "loss": 0.2623,
      "step": 850
    },
    {
      "epoch": 2.193877551020408,
      "grad_norm": 1.952677607536316,
      "learning_rate": 1.0281356361761539e-06,
      "loss": 0.2768,
      "step": 860
    },
    {
      "epoch": 2.2193877551020407,
      "grad_norm": 1.7542058229446411,
      "learning_rate": 9.687884025772364e-07,
      "loss": 0.2592,
      "step": 870
    },
    {
      "epoch": 2.2448979591836733,
      "grad_norm": 2.021303415298462,
      "learning_rate": 9.107911625770718e-07,
      "loss": 0.2654,
      "step": 880
    },
    {
      "epoch": 2.270408163265306,
      "grad_norm": 1.9793764352798462,
      "learning_rate": 8.541950494761633e-07,
      "loss": 0.2655,
      "step": 890
    },
    {
      "epoch": 2.295918367346939,
      "grad_norm": 1.9583169221878052,
      "learning_rate": 7.99049961270666e-07,
      "loss": 0.2658,
      "step": 900
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 1.8489506244659424,
      "learning_rate": 7.454045166597884e-07,
      "loss": 0.2751,
      "step": 910
    },
    {
      "epoch": 2.3469387755102042,
      "grad_norm": 1.9560836553573608,
      "learning_rate": 6.933060121810886e-07,
      "loss": 0.258,
      "step": 920
    },
    {
      "epoch": 2.372448979591837,
      "grad_norm": 2.220918655395508,
      "learning_rate": 6.428003805114544e-07,
      "loss": 0.2557,
      "step": 930
    },
    {
      "epoch": 2.3979591836734695,
      "grad_norm": 2.1624178886413574,
      "learning_rate": 5.939321499705311e-07,
      "loss": 0.2713,
      "step": 940
    },
    {
      "epoch": 2.423469387755102,
      "grad_norm": 1.792274832725525,
      "learning_rate": 5.467444052623034e-07,
      "loss": 0.2513,
      "step": 950
    },
    {
      "epoch": 2.4489795918367347,
      "grad_norm": 2.3812448978424072,
      "learning_rate": 5.0127874948944e-07,
      "loss": 0.2835,
      "step": 960
    },
    {
      "epoch": 2.4744897959183674,
      "grad_norm": 1.8628736734390259,
      "learning_rate": 4.5757526747389506e-07,
      "loss": 0.2544,
      "step": 970
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.102025270462036,
      "learning_rate": 4.156724904161008e-07,
      "loss": 0.271,
      "step": 980
    },
    {
      "epoch": 2.5255102040816326,
      "grad_norm": 2.079012870788574,
      "learning_rate": 3.7560736192391174e-07,
      "loss": 0.2578,
      "step": 990
    },
    {
      "epoch": 2.5510204081632653,
      "grad_norm": 1.9921406507492065,
      "learning_rate": 3.3741520544125243e-07,
      "loss": 0.2596,
      "step": 1000
    },
    {
      "epoch": 2.576530612244898,
      "grad_norm": 2.3191800117492676,
      "learning_rate": 3.011296931051813e-07,
      "loss": 0.2641,
      "step": 1010
    },
    {
      "epoch": 2.6020408163265305,
      "grad_norm": 1.9748038053512573,
      "learning_rate": 2.6678281605883206e-07,
      "loss": 0.2624,
      "step": 1020
    },
    {
      "epoch": 2.627551020408163,
      "grad_norm": 2.1532864570617676,
      "learning_rate": 2.3440485624640325e-07,
      "loss": 0.2717,
      "step": 1030
    },
    {
      "epoch": 2.6530612244897958,
      "grad_norm": 2.0552635192871094,
      "learning_rate": 2.040243597150654e-07,
      "loss": 0.2767,
      "step": 1040
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 1.7890598773956299,
      "learning_rate": 1.7566811144732076e-07,
      "loss": 0.2622,
      "step": 1050
    },
    {
      "epoch": 2.704081632653061,
      "grad_norm": 1.9225808382034302,
      "learning_rate": 1.4936111174600808e-07,
      "loss": 0.2707,
      "step": 1060
    },
    {
      "epoch": 2.729591836734694,
      "grad_norm": 2.1159281730651855,
      "learning_rate": 1.2512655419277152e-07,
      "loss": 0.264,
      "step": 1070
    },
    {
      "epoch": 2.7551020408163263,
      "grad_norm": 2.0600357055664062,
      "learning_rate": 1.0298580519942536e-07,
      "loss": 0.2663,
      "step": 1080
    },
    {
      "epoch": 2.7806122448979593,
      "grad_norm": 2.0410473346710205,
      "learning_rate": 8.29583851702448e-08,
      "loss": 0.255,
      "step": 1090
    },
    {
      "epoch": 2.806122448979592,
      "grad_norm": 1.9862523078918457,
      "learning_rate": 6.506195129179021e-08,
      "loss": 0.2662,
      "step": 1100
    },
    {
      "epoch": 2.8316326530612246,
      "grad_norm": 2.0106465816497803,
      "learning_rate": 4.931228196543847e-08,
      "loss": 0.2616,
      "step": 1110
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 1.9957739114761353,
      "learning_rate": 3.572326289634415e-08,
      "loss": 0.2637,
      "step": 1120
    },
    {
      "epoch": 2.88265306122449,
      "grad_norm": 1.9279589653015137,
      "learning_rate": 2.4306874851103834e-08,
      "loss": 0.2702,
      "step": 1130
    },
    {
      "epoch": 2.9081632653061225,
      "grad_norm": 1.8473937511444092,
      "learning_rate": 1.5073183094901632e-08,
      "loss": 0.2748,
      "step": 1140
    },
    {
      "epoch": 2.933673469387755,
      "grad_norm": 1.8530915975570679,
      "learning_rate": 8.03032851746488e-09,
      "loss": 0.279,
      "step": 1150
    },
    {
      "epoch": 2.9591836734693877,
      "grad_norm": 2.213524580001831,
      "learning_rate": 3.184520455642992e-09,
      "loss": 0.2869,
      "step": 1160
    },
    {
      "epoch": 2.9846938775510203,
      "grad_norm": 2.0284934043884277,
      "learning_rate": 5.400312189426937e-10,
      "loss": 0.2537,
      "step": 1170
    },
    {
      "epoch": 3.0,
      "step": 1176,
      "total_flos": 1.400695163851899e+18,
      "train_loss": 0.4022732187493318,
      "train_runtime": 68284.6725,
      "train_samples_per_second": 0.138,
      "train_steps_per_second": 0.017
    }
  ],
  "logging_steps": 10,
  "max_steps": 1176,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.400695163851899e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}