{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995656122892232,
  "eval_steps": 500,
  "global_step": 1582,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.999893525548073e-05,
      "loss": 3.5344,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 2.9992429030811932e-05,
      "loss": 3.2577,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.9980010668771548e-05,
      "loss": 3.1536,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.9961685066435444e-05,
      "loss": 3.1061,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.9937459450349648e-05,
      "loss": 3.0524,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.990734337368062e-05,
      "loss": 3.0353,
      "step": 60
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.9871348712448046e-05,
      "loss": 3.0027,
      "step": 70
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.9829489660841637e-05,
      "loss": 2.9949,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.9781782725623765e-05,
      "loss": 2.9761,
      "step": 90
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.972824671962015e-05,
      "loss": 2.9627,
      "step": 100
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.966890275430121e-05,
      "loss": 2.9518,
      "step": 110
    },
    {
      "epoch": 0.08,
      "learning_rate": 2.960377423145689e-05,
      "loss": 2.9335,
      "step": 120
    },
    {
      "epoch": 0.08,
      "learning_rate": 2.953288683396841e-05,
      "loss": 2.9325,
      "step": 130
    },
    {
      "epoch": 0.09,
      "learning_rate": 2.9456268515680404e-05,
      "loss": 2.9191,
      "step": 140
    },
    {
      "epoch": 0.09,
      "learning_rate": 2.937394949037758e-05,
      "loss": 2.9055,
      "step": 150
    },
    {
      "epoch": 0.1,
      "learning_rate": 2.928596221987015e-05,
      "loss": 2.8912,
      "step": 160
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.91923414011928e-05,
      "loss": 2.8814,
      "step": 170
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.909312395292222e-05,
      "loss": 2.869,
      "step": 180
    },
    {
      "epoch": 0.12,
      "learning_rate": 2.8988349000618563e-05,
      "loss": 2.8692,
      "step": 190
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.8878057861396606e-05,
      "loss": 2.8543,
      "step": 200
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.876229402763271e-05,
      "loss": 2.8453,
      "step": 210
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.864110314981396e-05,
      "loss": 2.8486,
      "step": 220
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 2.843,
      "step": 230
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.8382633545658654e-05,
      "loss": 2.8274,
      "step": 240
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.8245456744620734e-05,
      "loss": 2.836,
      "step": 250
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.8103056709931837e-05,
      "loss": 2.8178,
      "step": 260
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.7955489595839228e-05,
      "loss": 2.8173,
      "step": 270
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.780281359418415e-05,
      "loss": 2.8154,
      "step": 280
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.7645088911454375e-05,
      "loss": 2.8041,
      "step": 290
    },
    {
      "epoch": 0.19,
      "learning_rate": 2.748237774504229e-05,
      "loss": 2.7905,
      "step": 300
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.7314744258717864e-05,
      "loss": 2.7941,
      "step": 310
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.7142254557326233e-05,
      "loss": 2.7945,
      "step": 320
    },
    {
      "epoch": 0.21,
      "learning_rate": 2.69649766607198e-05,
      "loss": 2.7762,
      "step": 330
    },
    {
      "epoch": 0.21,
      "learning_rate": 2.678298047693518e-05,
      "loss": 2.7693,
      "step": 340
    },
    {
      "epoch": 0.22,
      "learning_rate": 2.6596337774625565e-05,
      "loss": 2.762,
      "step": 350
    },
    {
      "epoch": 0.23,
      "learning_rate": 2.6405122154759366e-05,
      "loss": 2.7738,
      "step": 360
    },
    {
      "epoch": 0.23,
      "learning_rate": 2.6209409021596286e-05,
      "loss": 2.7729,
      "step": 370
    },
    {
      "epoch": 0.24,
      "learning_rate": 2.6009275552952297e-05,
      "loss": 2.7664,
      "step": 380
    },
    {
      "epoch": 0.25,
      "learning_rate": 2.5804800669765212e-05,
      "loss": 2.7592,
      "step": 390
    },
    {
      "epoch": 0.25,
      "learning_rate": 2.5596065004972898e-05,
      "loss": 2.7376,
      "step": 400
    },
    {
      "epoch": 0.26,
      "learning_rate": 2.5383150871716342e-05,
      "loss": 2.7415,
      "step": 410
    },
    {
      "epoch": 0.27,
      "learning_rate": 2.516614223088018e-05,
      "loss": 2.7426,
      "step": 420
    },
    {
      "epoch": 0.27,
      "learning_rate": 2.49451246579834e-05,
      "loss": 2.7342,
      "step": 430
    },
    {
      "epoch": 0.28,
      "learning_rate": 2.472018530943342e-05,
      "loss": 2.7323,
      "step": 440
    },
    {
      "epoch": 0.28,
      "learning_rate": 2.4491412888156627e-05,
      "loss": 2.7233,
      "step": 450
    },
    {
      "epoch": 0.29,
      "learning_rate": 2.4258897608619188e-05,
      "loss": 2.714,
      "step": 460
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.4022731161251696e-05,
      "loss": 2.7176,
      "step": 470
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.3783006676291866e-05,
      "loss": 2.7241,
      "step": 480
    },
    {
      "epoch": 0.31,
      "learning_rate": 2.3539818687059375e-05,
      "loss": 2.7116,
      "step": 490
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.329326309267747e-05,
      "loss": 2.7032,
      "step": 500
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.3043437120255976e-05,
      "loss": 2.7064,
      "step": 510
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2790439286550566e-05,
      "loss": 2.7111,
      "step": 520
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2534369359113554e-05,
      "loss": 2.7055,
      "step": 530
    },
    {
      "epoch": 0.34,
      "learning_rate": 2.2275328316951384e-05,
      "loss": 2.7003,
      "step": 540
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.2013418310704422e-05,
      "loss": 2.6933,
      "step": 550
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.1748742622364718e-05,
      "loss": 2.6905,
      "step": 560
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.148140562454762e-05,
      "loss": 2.6934,
      "step": 570
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.121151273933332e-05,
      "loss": 2.6792,
      "step": 580
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.0939170396694554e-05,
      "loss": 2.6727,
      "step": 590
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.0664485992526845e-05,
      "loss": 2.6755,
      "step": 600
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.0387567846297832e-05,
      "loss": 2.6733,
      "step": 610
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.010852515833242e-05,
      "loss": 2.6637,
      "step": 620
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9827467966750533e-05,
      "loss": 2.6592,
      "step": 630
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9544507104074547e-05,
      "loss": 2.6676,
      "step": 640
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.925975415352339e-05,
      "loss": 2.6605,
      "step": 650
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.8973321405010678e-05,
      "loss": 2.6579,
      "step": 660
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.8685321810864098e-05,
      "loss": 2.6536,
      "step": 670
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8395868941283672e-05,
      "loss": 2.6464,
      "step": 680
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.8105076939556238e-05,
      "loss": 2.6525,
      "step": 690
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.781306047704409e-05,
      "loss": 2.6339,
      "step": 700
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.751993470796522e-05,
      "loss": 2.6435,
      "step": 710
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.7225815223983302e-05,
      "loss": 2.6378,
      "step": 720
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.6930818008625035e-05,
      "loss": 2.6366,
      "step": 730
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.6635059391543067e-05,
      "loss": 2.6418,
      "step": 740
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.6338656002642366e-05,
      "loss": 2.6287,
      "step": 750
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.6041724726088187e-05,
      "loss": 2.623,
      "step": 760
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.5744382654213834e-05,
      "loss": 2.6228,
      "step": 770
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.5446747041346237e-05,
      "loss": 2.6267,
      "step": 780
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.5148935257567761e-05,
      "loss": 2.6241,
      "step": 790
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.4851064742432246e-05,
      "loss": 2.6184,
      "step": 800
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.4553252958653764e-05,
      "loss": 2.6166,
      "step": 810
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.4255617345786167e-05,
      "loss": 2.612,
      "step": 820
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.3958275273911812e-05,
      "loss": 2.6187,
      "step": 830
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.366134399735764e-05,
      "loss": 2.6108,
      "step": 840
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.3364940608456937e-05,
      "loss": 2.6053,
      "step": 850
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.3069181991374968e-05,
      "loss": 2.6057,
      "step": 860
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.2774184776016704e-05,
      "loss": 2.6014,
      "step": 870
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.2480065292034782e-05,
      "loss": 2.6042,
      "step": 880
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.2186939522955918e-05,
      "loss": 2.6045,
      "step": 890
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.1894923060443763e-05,
      "loss": 2.6042,
      "step": 900
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.160413105871633e-05,
      "loss": 2.5976,
      "step": 910
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.13146781891359e-05,
      "loss": 2.5961,
      "step": 920
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.102667859498933e-05,
      "loss": 2.5942,
      "step": 930
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.0740245846476613e-05,
      "loss": 2.5886,
      "step": 940
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.0455492895925456e-05,
      "loss": 2.5824,
      "step": 950
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.0172532033249467e-05,
      "loss": 2.5877,
      "step": 960
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.891474841667585e-06,
      "loss": 2.5852,
      "step": 970
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.61243215370217e-06,
      "loss": 2.5889,
      "step": 980
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.335514007473155e-06,
      "loss": 2.5794,
      "step": 990
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.060829603305446e-06,
      "loss": 2.5715,
      "step": 1000
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.788487260666683e-06,
      "loss": 2.5798,
      "step": 1010
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.518594375452384e-06,
      "loss": 2.5705,
      "step": 1020
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.251257377635285e-06,
      "loss": 2.5763,
      "step": 1030
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.986581689295577e-06,
      "loss": 2.5734,
      "step": 1040
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.724671683048615e-06,
      "loss": 2.5701,
      "step": 1050
    },
    {
      "epoch": 0.67,
      "learning_rate": 7.465630640886447e-06,
      "loss": 2.5609,
      "step": 1060
    },
    {
      "epoch": 0.68,
      "learning_rate": 7.209560713449439e-06,
      "loss": 2.5639,
      "step": 1070
    },
    {
      "epoch": 0.68,
      "learning_rate": 6.956562879744027e-06,
      "loss": 2.5749,
      "step": 1080
    },
    {
      "epoch": 0.69,
      "learning_rate": 6.706736907322525e-06,
      "loss": 2.5661,
      "step": 1090
    },
    {
      "epoch": 0.7,
      "learning_rate": 6.460181312940628e-06,
      "loss": 2.5743,
      "step": 1100
    },
    {
      "epoch": 0.7,
      "learning_rate": 6.2169933237081386e-06,
      "loss": 2.5721,
      "step": 1110
    },
    {
      "epoch": 0.71,
      "learning_rate": 5.977268838748303e-06,
      "loss": 2.5746,
      "step": 1120
    },
    {
      "epoch": 0.71,
      "learning_rate": 5.7411023913808115e-06,
      "loss": 2.5565,
      "step": 1130
    },
    {
      "epoch": 0.72,
      "learning_rate": 5.508587111843373e-06,
      "loss": 2.5607,
      "step": 1140
    },
    {
      "epoch": 0.73,
      "learning_rate": 5.279814690566585e-06,
      "loss": 2.5569,
      "step": 1150
    },
    {
      "epoch": 0.73,
      "learning_rate": 5.054875342016599e-06,
      "loss": 2.5594,
      "step": 1160
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.833857769119822e-06,
      "loss": 2.5568,
      "step": 1170
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.616849128283658e-06,
      "loss": 2.5606,
      "step": 1180
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.403934995027105e-06,
      "loss": 2.555,
      "step": 1190
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.195199330234789e-06,
      "loss": 2.5511,
      "step": 1200
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.990724447047702e-06,
      "loss": 2.5439,
      "step": 1210
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.790590978403716e-06,
      "loss": 2.5611,
      "step": 1220
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.594877845240638e-06,
      "loss": 2.5441,
      "step": 1230
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.403662225374437e-06,
      "loss": 2.5446,
      "step": 1240
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.217019523064825e-06,
      "loss": 2.5496,
      "step": 1250
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.035023339280201e-06,
      "loss": 2.5526,
      "step": 1260
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.8577454426737635e-06,
      "loss": 2.5444,
      "step": 1270
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.685255741282137e-06,
      "loss": 2.5514,
      "step": 1280
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.5176222549577134e-06,
      "loss": 2.5537,
      "step": 1290
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.354911088545625e-06,
      "loss": 2.5433,
      "step": 1300
    },
    {
      "epoch": 0.83,
      "learning_rate": 2.1971864058158507e-06,
      "loss": 2.548,
      "step": 1310
    },
    {
      "epoch": 0.83,
      "learning_rate": 2.044510404160774e-06,
      "loss": 2.5431,
      "step": 1320
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.8969432900681672e-06,
      "loss": 2.5386,
      "step": 1330
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.7545432553792683e-06,
      "loss": 2.5459,
      "step": 1340
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6173664543413435e-06,
      "loss": 2.5442,
      "step": 1350
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 2.538,
      "step": 1360
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.3588968501860449e-06,
      "loss": 2.5495,
      "step": 1370
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.237705972367294e-06,
      "loss": 2.5392,
      "step": 1380
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.1219421386033958e-06,
      "loss": 2.5355,
      "step": 1390
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.0116509993814388e-06,
      "loss": 2.5495,
      "step": 1400
    },
    {
      "epoch": 0.89,
      "learning_rate": 9.068760470777821e-07,
      "loss": 2.5433,
      "step": 1410
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.076585988072033e-07,
      "loss": 2.5463,
      "step": 1420
    },
    {
      "epoch": 0.9,
      "learning_rate": 7.140377801298542e-07,
      "loss": 2.535,
      "step": 1430
    },
    {
      "epoch": 0.91,
      "learning_rate": 6.260505096224222e-07,
      "loss": 2.5427,
      "step": 1440
    },
    {
      "epoch": 0.92,
      "learning_rate": 5.437314843195962e-07,
      "loss": 2.5426,
      "step": 1450
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.6711316603159084e-07,
      "loss": 2.5384,
      "step": 1460
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.962257685431103e-07,
      "loss": 2.5463,
      "step": 1470
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.310972456987937e-07,
      "loss": 2.5353,
      "step": 1480
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.7175328037985115e-07,
      "loss": 2.5289,
      "step": 1490
    },
    {
      "epoch": 0.95,
      "learning_rate": 2.1821727437623672e-07,
      "loss": 2.5392,
      "step": 1500
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.7051033915836233e-07,
      "loss": 2.5441,
      "step": 1510
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.2865128755195522e-07,
      "loss": 2.531,
      "step": 1520
    },
    {
      "epoch": 0.97,
      "learning_rate": 9.265662631938399e-08,
      "loss": 2.5296,
      "step": 1530
    },
    {
      "epoch": 0.97,
      "learning_rate": 6.254054965035394e-08,
      "loss": 2.5368,
      "step": 1540
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.8314933564556066e-08,
      "loss": 2.5445,
      "step": 1550
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.9989331228453257e-08,
      "loss": 2.5476,
      "step": 1560
    },
    {
      "epoch": 0.99,
      "learning_rate": 7.570969188070432e-09,
      "loss": 2.5443,
      "step": 1570
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.064744519270744e-09,
      "loss": 2.5448,
      "step": 1580
    },
    {
      "epoch": 1.0,
      "step": 1582,
      "total_flos": 1.413137776868967e+20,
      "train_loss": 2.6770026192502336,
      "train_runtime": 195293.7522,
      "train_samples_per_second": 8.299,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 10,
  "max_steps": 1582,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 1.413137776868967e+20,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}