{
  "best_metric": 0.019110437482595444,
  "best_model_checkpoint": "runs/deepseek_lora_20240422-151242/checkpoint-2500",
  "epoch": 1.5665387326701654,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.4764970541000366,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 1.5176,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.0680618286132812,
      "learning_rate": 8.000000000000001e-07,
      "loss": 1.5115,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.892202138900757,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.4821,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8446696996688843,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.4669,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.351251244544983,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.5485,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4435632228851318,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.3646,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.500685691833496,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 1.5509,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.027316093444824,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.4173,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5012431144714355,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 1.4232,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.025063991546631,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.3952,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2457600831985474,
      "learning_rate": 4.4e-06,
      "loss": 1.1888,
      "step": 110
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2635124921798706,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.2259,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.6183345317840576,
      "learning_rate": 5.2e-06,
      "loss": 1.0052,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.38128924369812,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.9928,
      "step": 140
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.641148328781128,
      "learning_rate": 6e-06,
      "loss": 0.9287,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0428839921951294,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.7218,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0025560855865479,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.6115,
      "step": 170
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.962586522102356,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.5337,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.46322694420814514,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.3379,
      "step": 190
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.5135998725891113,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3204,
      "step": 200
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4679415225982666,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.2534,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6532612442970276,
      "learning_rate": 8.8e-06,
      "loss": 0.1868,
      "step": 220
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.841340184211731,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.2609,
      "step": 230
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7561472058296204,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.3831,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.38061589002609253,
      "learning_rate": 1e-05,
      "loss": 0.3377,
      "step": 250
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.24261775612831116,
      "learning_rate": 1.04e-05,
      "loss": 0.1452,
      "step": 260
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.345815896987915,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.0759,
      "step": 270
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.30060628056526184,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.233,
      "step": 280
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.45316824316978455,
      "learning_rate": 1.16e-05,
      "loss": 0.1512,
      "step": 290
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.31342798471450806,
      "learning_rate": 1.2e-05,
      "loss": 0.1612,
      "step": 300
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.0591862201690674,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.1128,
      "step": 310
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.04360930994153023,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.1209,
      "step": 320
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2543979585170746,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.139,
      "step": 330
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.55256462097168,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.1386,
      "step": 340
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.032540880143642426,
      "learning_rate": 1.4e-05,
      "loss": 0.0659,
      "step": 350
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.020825587213039398,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.0491,
      "step": 360
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.304826736450195,
      "learning_rate": 1.48e-05,
      "loss": 0.1335,
      "step": 370
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.03812132403254509,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0341,
      "step": 380
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.0986132025718689,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.1627,
      "step": 390
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4882945716381073,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0757,
      "step": 400
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.015025519765913486,
      "learning_rate": 1.64e-05,
      "loss": 0.0658,
      "step": 410
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16647090017795563,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0375,
      "step": 420
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.016350600868463516,
      "learning_rate": 1.72e-05,
      "loss": 0.1766,
      "step": 430
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.011344443075358868,
      "learning_rate": 1.76e-05,
      "loss": 0.0802,
      "step": 440
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.05872802808880806,
      "learning_rate": 1.8e-05,
      "loss": 0.0866,
      "step": 450
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.6746442317962646,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.0869,
      "step": 460
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.01068535540252924,
      "learning_rate": 1.88e-05,
      "loss": 0.2154,
      "step": 470
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.020563703030347824,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.101,
      "step": 480
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.17447501420974731,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.1116,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.7621819972991943,
      "learning_rate": 2e-05,
      "loss": 0.1465,
      "step": 500
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.05017045885324478,
      "eval_runtime": 61.9228,
      "eval_samples_per_second": 16.149,
      "eval_steps_per_second": 16.149,
      "step": 500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.015244358219206333,
      "learning_rate": 1.9955555555555557e-05,
      "loss": 0.0218,
      "step": 510
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.20013077557086945,
      "learning_rate": 1.9911111111111112e-05,
      "loss": 0.0845,
      "step": 520
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.007578707300126553,
      "learning_rate": 1.9866666666666667e-05,
      "loss": 0.1364,
      "step": 530
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.6152952909469604,
      "learning_rate": 1.9822222222222226e-05,
      "loss": 0.1135,
      "step": 540
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.0188419818878174,
      "learning_rate": 1.977777777777778e-05,
      "loss": 0.1452,
      "step": 550
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.007765072397887707,
      "learning_rate": 1.9733333333333336e-05,
      "loss": 0.1082,
      "step": 560
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11964543163776398,
      "learning_rate": 1.968888888888889e-05,
      "loss": 0.0912,
      "step": 570
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.242813587188721,
      "learning_rate": 1.9644444444444447e-05,
      "loss": 0.1243,
      "step": 580
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.00892120786011219,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.1229,
      "step": 590
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.008041107095777988,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 0.1782,
      "step": 600
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.013454565778374672,
      "learning_rate": 1.9511111111111113e-05,
      "loss": 0.0647,
      "step": 610
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15643712878227234,
      "learning_rate": 1.9466666666666668e-05,
      "loss": 0.0449,
      "step": 620
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.01665414497256279,
      "learning_rate": 1.9422222222222223e-05,
      "loss": 0.0636,
      "step": 630
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.01023570355027914,
      "learning_rate": 1.9377777777777778e-05,
      "loss": 0.0566,
      "step": 640
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.00858321599662304,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 0.2039,
      "step": 650
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.273760795593262,
      "learning_rate": 1.928888888888889e-05,
      "loss": 0.0662,
      "step": 660
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.022378291934728622,
      "learning_rate": 1.9244444444444444e-05,
      "loss": 0.1163,
      "step": 670
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.08007720112800598,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0781,
      "step": 680
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.018208255991339684,
      "learning_rate": 1.9155555555555558e-05,
      "loss": 0.099,
      "step": 690
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.07240291684865952,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 0.0522,
      "step": 700
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5902518033981323,
      "learning_rate": 1.9066666666666668e-05,
      "loss": 0.0421,
      "step": 710
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.01304863765835762,
      "learning_rate": 1.9022222222222223e-05,
      "loss": 0.0968,
      "step": 720
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.013874703086912632,
      "learning_rate": 1.897777777777778e-05,
      "loss": 0.016,
      "step": 730
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.01731191948056221,
      "learning_rate": 1.8933333333333334e-05,
      "loss": 0.0936,
      "step": 740
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.513253688812256,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.1283,
      "step": 750
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9819907546043396,
      "learning_rate": 1.8844444444444444e-05,
      "loss": 0.1182,
      "step": 760
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.303761005401611,
      "learning_rate": 1.88e-05,
      "loss": 0.0491,
      "step": 770
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.017575740814209,
      "learning_rate": 1.8755555555555558e-05,
      "loss": 0.115,
      "step": 780
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.07593460381031036,
      "learning_rate": 1.8711111111111113e-05,
      "loss": 0.0237,
      "step": 790
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.1718034744262695,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.1243,
      "step": 800
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.009034764021635056,
      "learning_rate": 1.8622222222222224e-05,
      "loss": 0.0435,
      "step": 810
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.008039540611207485,
      "learning_rate": 1.857777777777778e-05,
      "loss": 0.0797,
      "step": 820
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.030577246099710464,
      "learning_rate": 1.8533333333333334e-05,
      "loss": 0.0902,
      "step": 830
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.01614721678197384,
      "learning_rate": 1.848888888888889e-05,
      "loss": 0.0443,
      "step": 840
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.00745397200807929,
      "learning_rate": 1.8444444444444448e-05,
      "loss": 0.013,
      "step": 850
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.0033397756051272154,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.0177,
      "step": 860
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.005004946142435074,
      "learning_rate": 1.835555555555556e-05,
      "loss": 0.0414,
      "step": 870
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.005703146569430828,
      "learning_rate": 1.8311111111111114e-05,
      "loss": 0.1875,
      "step": 880
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.011084661819040775,
      "learning_rate": 1.826666666666667e-05,
      "loss": 0.0093,
      "step": 890
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.5405566692352295,
      "learning_rate": 1.8222222222222224e-05,
      "loss": 0.074,
      "step": 900
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.003973611164838076,
      "learning_rate": 1.817777777777778e-05,
      "loss": 0.0506,
      "step": 910
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.04966599866747856,
      "learning_rate": 1.8133333333333335e-05,
      "loss": 0.0544,
      "step": 920
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.01420037355273962,
      "learning_rate": 1.808888888888889e-05,
      "loss": 0.0153,
      "step": 930
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.305298328399658,
      "learning_rate": 1.8044444444444445e-05,
      "loss": 0.0359,
      "step": 940
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.4599621295928955,
      "learning_rate": 1.8e-05,
      "loss": 0.0808,
      "step": 950
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.250840187072754,
      "learning_rate": 1.7955555555555556e-05,
      "loss": 0.1756,
      "step": 960
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.541337251663208,
      "learning_rate": 1.791111111111111e-05,
      "loss": 0.0827,
      "step": 970
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.04627770185470581,
      "learning_rate": 1.7866666666666666e-05,
      "loss": 0.0777,
      "step": 980
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.06008163094520569,
      "learning_rate": 1.782222222222222e-05,
      "loss": 0.0921,
      "step": 990
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.03849005699157715,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 0.0555,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.028874782845377922,
      "eval_runtime": 61.928,
      "eval_samples_per_second": 16.148,
      "eval_steps_per_second": 16.148,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.006071585696190596,
      "learning_rate": 1.7733333333333335e-05,
      "loss": 0.0399,
      "step": 1010
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.012464660219848156,
      "learning_rate": 1.768888888888889e-05,
      "loss": 0.075,
      "step": 1020
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.0500749833881855,
      "learning_rate": 1.7644444444444446e-05,
      "loss": 0.1375,
      "step": 1030
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.1470282077789307,
      "learning_rate": 1.76e-05,
      "loss": 0.0725,
      "step": 1040
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.040950775146484,
      "learning_rate": 1.7555555555555556e-05,
      "loss": 0.0757,
      "step": 1050
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.010639629326760769,
      "learning_rate": 1.751111111111111e-05,
      "loss": 0.0495,
      "step": 1060
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.009533255361020565,
      "learning_rate": 1.7466666666666667e-05,
      "loss": 0.0125,
      "step": 1070
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.4316189289093018,
      "learning_rate": 1.7422222222222222e-05,
      "loss": 0.089,
      "step": 1080
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.18850259482860565,
      "learning_rate": 1.737777777777778e-05,
      "loss": 0.0359,
      "step": 1090
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.012972812168300152,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.0618,
      "step": 1100
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8784104585647583,
      "learning_rate": 1.728888888888889e-05,
      "loss": 0.106,
      "step": 1110
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.02918626181781292,
      "learning_rate": 1.7244444444444446e-05,
      "loss": 0.0367,
      "step": 1120
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.1463236808776855,
      "learning_rate": 1.72e-05,
      "loss": 0.0784,
      "step": 1130
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.754035234451294,
      "learning_rate": 1.7155555555555557e-05,
      "loss": 0.08,
      "step": 1140
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.014391366392374039,
      "learning_rate": 1.7111111111111112e-05,
      "loss": 0.1047,
      "step": 1150
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.016152633354067802,
      "learning_rate": 1.706666666666667e-05,
      "loss": 0.0453,
      "step": 1160
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.020005827769637108,
      "learning_rate": 1.7022222222222226e-05,
      "loss": 0.0388,
      "step": 1170
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.004559040535241365,
      "learning_rate": 1.697777777777778e-05,
      "loss": 0.0306,
      "step": 1180
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5455616116523743,
      "learning_rate": 1.6933333333333336e-05,
      "loss": 0.0595,
      "step": 1190
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.00697706313803792,
      "learning_rate": 1.688888888888889e-05,
      "loss": 0.1236,
      "step": 1200
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.926461935043335,
      "learning_rate": 1.6844444444444447e-05,
      "loss": 0.0155,
      "step": 1210
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.4559099674224854,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0083,
      "step": 1220
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.033357810229063034,
      "learning_rate": 1.6755555555555557e-05,
      "loss": 0.0337,
      "step": 1230
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.008271641097962856,
      "learning_rate": 1.6711111111111112e-05,
      "loss": 0.0618,
      "step": 1240
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.003830987960100174,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0097,
      "step": 1250
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.021310480311512947,
      "learning_rate": 1.6622222222222223e-05,
      "loss": 0.0717,
      "step": 1260
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.166944980621338,
      "learning_rate": 1.6577777777777778e-05,
      "loss": 0.0135,
      "step": 1270
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.0050102900713682175,
      "learning_rate": 1.6533333333333333e-05,
      "loss": 0.02,
      "step": 1280
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.044907063245773315,
      "learning_rate": 1.648888888888889e-05,
      "loss": 0.0688,
      "step": 1290
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.06957779079675674,
      "learning_rate": 1.6444444444444444e-05,
      "loss": 0.0887,
      "step": 1300
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.025342583656311,
      "learning_rate": 1.64e-05,
      "loss": 0.0536,
      "step": 1310
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.011287701316177845,
      "learning_rate": 1.6355555555555557e-05,
      "loss": 0.0663,
      "step": 1320
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.220019578933716,
      "learning_rate": 1.6311111111111113e-05,
      "loss": 0.0495,
      "step": 1330
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6111379861831665,
      "learning_rate": 1.6266666666666668e-05,
      "loss": 0.0658,
      "step": 1340
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.004258489701896906,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 0.0022,
      "step": 1350
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0197640657424927,
      "learning_rate": 1.617777777777778e-05,
      "loss": 0.0969,
      "step": 1360
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.003954906482249498,
      "learning_rate": 1.6133333333333334e-05,
      "loss": 0.0474,
      "step": 1370
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.326837539672852,
      "learning_rate": 1.608888888888889e-05,
      "loss": 0.0683,
      "step": 1380
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.008893176913261414,
      "learning_rate": 1.6044444444444444e-05,
      "loss": 0.0577,
      "step": 1390
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.007870793342590332,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0658,
      "step": 1400
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.003645359305664897,
      "learning_rate": 1.5955555555555558e-05,
      "loss": 0.004,
      "step": 1410
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6187639236450195,
      "learning_rate": 1.5911111111111113e-05,
      "loss": 0.0814,
      "step": 1420
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7969435453414917,
      "learning_rate": 1.586666666666667e-05,
      "loss": 0.0154,
      "step": 1430
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.8159172534942627,
      "learning_rate": 1.5822222222222224e-05,
      "loss": 0.1169,
      "step": 1440
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7929564714431763,
      "learning_rate": 1.577777777777778e-05,
      "loss": 0.1261,
      "step": 1450
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2537187337875366,
      "learning_rate": 1.5733333333333334e-05,
      "loss": 0.0193,
      "step": 1460
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.004017004277557135,
      "learning_rate": 1.5688888888888893e-05,
      "loss": 0.0246,
      "step": 1470
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.6317222118377686,
      "learning_rate": 1.5644444444444448e-05,
      "loss": 0.0987,
      "step": 1480
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.007867836393415928,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.0535,
      "step": 1490
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.21237313747406,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.0514,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.021845491603016853,
      "eval_runtime": 61.8947,
      "eval_samples_per_second": 16.156,
      "eval_steps_per_second": 16.156,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.022142503410577774,
      "learning_rate": 1.5511111111111114e-05,
      "loss": 0.0561,
      "step": 1510
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5647516250610352,
      "learning_rate": 1.546666666666667e-05,
      "loss": 0.0338,
      "step": 1520
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.00992211140692234,
      "learning_rate": 1.5422222222222224e-05,
      "loss": 0.0595,
      "step": 1530
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.0023656291887164116,
      "learning_rate": 1.537777777777778e-05,
      "loss": 0.0406,
      "step": 1540
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.043642736971378326,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 0.1006,
      "step": 1550
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14472796022891998,
      "learning_rate": 1.528888888888889e-05,
      "loss": 0.0087,
      "step": 1560
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.002529141725972295,
      "learning_rate": 1.5244444444444447e-05,
      "loss": 0.0283,
      "step": 1570
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.0024450107011944056,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0695,
      "step": 1580
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.016819190233945847,
      "learning_rate": 1.5155555555555557e-05,
      "loss": 0.0606,
      "step": 1590
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9701637029647827,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 0.0807,
      "step": 1600
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.029984062537550926,
      "learning_rate": 1.5066666666666668e-05,
      "loss": 0.0693,
      "step": 1610
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3928338587284088,
      "learning_rate": 1.5022222222222223e-05,
      "loss": 0.0741,
      "step": 1620
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.005559510085731745,
      "learning_rate": 1.497777777777778e-05,
      "loss": 0.0725,
      "step": 1630
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.0025930607225745916,
      "learning_rate": 1.4933333333333335e-05,
      "loss": 0.1192,
      "step": 1640
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.3535702228546143,
      "learning_rate": 1.488888888888889e-05,
      "loss": 0.0606,
      "step": 1650
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.1475963592529297,
      "learning_rate": 1.4844444444444445e-05,
      "loss": 0.0956,
      "step": 1660
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.003930664621293545,
      "learning_rate": 1.48e-05,
      "loss": 0.0246,
      "step": 1670
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.315476655960083,
      "learning_rate": 1.4755555555555556e-05,
      "loss": 0.0953,
      "step": 1680
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8841826319694519,
      "learning_rate": 1.4711111111111111e-05,
      "loss": 0.035,
      "step": 1690
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.01530434563755989,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 0.0383,
      "step": 1700
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.836932182312012,
      "learning_rate": 1.4622222222222225e-05,
      "loss": 0.0981,
      "step": 1710
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.0168469101190567,
      "learning_rate": 1.457777777777778e-05,
      "loss": 0.1095,
      "step": 1720
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.15857402980327606,
      "learning_rate": 1.4533333333333335e-05,
      "loss": 0.0574,
      "step": 1730
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5915563106536865,
      "learning_rate": 1.448888888888889e-05,
      "loss": 0.0305,
      "step": 1740
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.039337050169706345,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 0.0177,
      "step": 1750
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.0033416992519050837,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.0009,
      "step": 1760
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.00678109098225832,
      "learning_rate": 1.4355555555555556e-05,
      "loss": 0.07,
      "step": 1770
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.4513912200927734,
      "learning_rate": 1.4311111111111111e-05,
      "loss": 0.0865,
      "step": 1780
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.966366708278656,
      "learning_rate": 1.4266666666666668e-05,
      "loss": 0.0032,
      "step": 1790
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.9548590183258057,
      "learning_rate": 1.4222222222222224e-05,
      "loss": 0.0721,
      "step": 1800
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.815798044204712,
      "learning_rate": 1.4177777777777779e-05,
      "loss": 0.032,
      "step": 1810
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.7791800498962402,
      "learning_rate": 1.4133333333333334e-05,
      "loss": 0.0763,
      "step": 1820
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.004075185861438513,
      "learning_rate": 1.408888888888889e-05,
      "loss": 0.0085,
      "step": 1830
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.0038460579235106707,
      "learning_rate": 1.4044444444444445e-05,
      "loss": 0.0696,
      "step": 1840
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.005857239011675119,
      "learning_rate": 1.4e-05,
      "loss": 0.0559,
      "step": 1850
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1006237044930458,
      "learning_rate": 1.3955555555555558e-05,
      "loss": 0.059,
      "step": 1860
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.004246099852025509,
      "learning_rate": 1.3911111111111114e-05,
      "loss": 0.0866,
      "step": 1870
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.00666034035384655,
      "learning_rate": 1.3866666666666669e-05,
      "loss": 0.0452,
      "step": 1880
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.0027115230914205313,
      "learning_rate": 1.3822222222222224e-05,
      "loss": 0.033,
      "step": 1890
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.016477108001709,
      "learning_rate": 1.377777777777778e-05,
      "loss": 0.0437,
      "step": 1900
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4079822301864624,
      "learning_rate": 1.3733333333333335e-05,
      "loss": 0.0474,
      "step": 1910
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.0025102542713284492,
      "learning_rate": 1.368888888888889e-05,
      "loss": 0.0631,
      "step": 1920
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.394373893737793,
      "learning_rate": 1.3644444444444445e-05,
      "loss": 0.0446,
      "step": 1930
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.0866854190826416,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.0073,
      "step": 1940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.007402149029076099,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 0.0311,
      "step": 1950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.004811545833945274,
      "learning_rate": 1.3511111111111112e-05,
      "loss": 0.0861,
      "step": 1960
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.004000263754278421,
      "learning_rate": 1.3466666666666668e-05,
      "loss": 0.0342,
      "step": 1970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.003166991053149104,
      "learning_rate": 1.3422222222222223e-05,
      "loss": 0.0897,
      "step": 1980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.006424722261726856,
      "learning_rate": 1.3377777777777778e-05,
      "loss": 0.0161,
      "step": 1990
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.0020944385323673487,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0358,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.017703723162412643,
      "eval_runtime": 61.8922,
      "eval_samples_per_second": 16.157,
      "eval_steps_per_second": 16.157,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.003021718468517065,
      "learning_rate": 1.3288888888888889e-05,
      "loss": 0.0445,
      "step": 2010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.010362805798649788,
      "learning_rate": 1.3244444444444447e-05,
      "loss": 0.0308,
      "step": 2020
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.32867729663848877,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.0514,
      "step": 2030
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.0034784479066729546,
      "learning_rate": 1.3155555555555558e-05,
      "loss": 0.0002,
      "step": 2040
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.08931613713502884,
      "learning_rate": 1.3111111111111113e-05,
      "loss": 0.0351,
      "step": 2050
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.0024316832423210144,
      "learning_rate": 1.3066666666666668e-05,
      "loss": 0.0822,
      "step": 2060
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6956457495689392,
      "learning_rate": 1.3022222222222223e-05,
      "loss": 0.0656,
      "step": 2070
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.006866452284157276,
      "learning_rate": 1.2977777777777779e-05,
      "loss": 0.0709,
      "step": 2080
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6907499432563782,
      "learning_rate": 1.2933333333333334e-05,
      "loss": 0.0474,
      "step": 2090
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.3049083948135376,
      "learning_rate": 1.288888888888889e-05,
      "loss": 0.096,
      "step": 2100
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.015370524488389492,
      "learning_rate": 1.2844444444444446e-05,
      "loss": 0.0433,
      "step": 2110
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.02007671445608139,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0422,
      "step": 2120
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.00284277624450624,
      "learning_rate": 1.2755555555555556e-05,
      "loss": 0.0432,
      "step": 2130
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.0024309258442372084,
      "learning_rate": 1.2711111111111112e-05,
      "loss": 0.0462,
      "step": 2140
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.016368340700864792,
      "learning_rate": 1.2666666666666667e-05,
      "loss": 0.0804,
      "step": 2150
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.005066912155598402,
      "learning_rate": 1.2622222222222222e-05,
      "loss": 0.0807,
      "step": 2160
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.002282659988850355,
      "learning_rate": 1.257777777777778e-05,
      "loss": 0.0395,
      "step": 2170
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8781790137290955,
      "learning_rate": 1.2533333333333336e-05,
      "loss": 0.0168,
      "step": 2180
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.0021755376365035772,
      "learning_rate": 1.2488888888888891e-05,
      "loss": 0.1348,
      "step": 2190
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.0018356326036155224,
      "learning_rate": 1.2444444444444446e-05,
      "loss": 0.0419,
      "step": 2200
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.0016609402373433113,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.0738,
      "step": 2210
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.0487405061721802,
      "learning_rate": 1.2355555555555557e-05,
      "loss": 0.0209,
      "step": 2220
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.0034472872503101826,
      "learning_rate": 1.2311111111111112e-05,
      "loss": 0.0322,
      "step": 2230
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.6054458618164062,
      "learning_rate": 1.2266666666666667e-05,
      "loss": 0.1416,
      "step": 2240
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.022294044494629,
      "learning_rate": 1.2222222222222224e-05,
      "loss": 0.0846,
      "step": 2250
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.010752925649285316,
      "learning_rate": 1.217777777777778e-05,
      "loss": 0.0093,
      "step": 2260
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9160808324813843,
      "learning_rate": 1.2133333333333335e-05,
      "loss": 0.0751,
      "step": 2270
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.002021084539592266,
      "learning_rate": 1.208888888888889e-05,
      "loss": 0.0271,
      "step": 2280
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0034389556385576725,
      "learning_rate": 1.2044444444444445e-05,
      "loss": 0.0233,
      "step": 2290
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0031041186302900314,
      "learning_rate": 1.2e-05,
      "loss": 0.068,
      "step": 2300
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0017185291508212686,
      "learning_rate": 1.1955555555555556e-05,
      "loss": 0.0369,
      "step": 2310
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.008447349071502686,
      "learning_rate": 1.191111111111111e-05,
      "loss": 0.0549,
      "step": 2320
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.0062310476787388325,
      "learning_rate": 1.186666666666667e-05,
      "loss": 0.0575,
      "step": 2330
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7583394646644592,
      "learning_rate": 1.1822222222222225e-05,
      "loss": 0.0256,
      "step": 2340
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4174383878707886,
      "learning_rate": 1.177777777777778e-05,
      "loss": 0.0805,
      "step": 2350
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.003498498583212495,
      "learning_rate": 1.1733333333333335e-05,
      "loss": 0.0472,
      "step": 2360
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.035282135009766,
      "learning_rate": 1.168888888888889e-05,
      "loss": 0.0662,
      "step": 2370
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.0031760677229613066,
      "learning_rate": 1.1644444444444446e-05,
      "loss": 0.0367,
      "step": 2380
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.006041550077497959,
      "learning_rate": 1.16e-05,
      "loss": 0.0408,
      "step": 2390
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.004563288297504187,
      "learning_rate": 1.1555555555555556e-05,
      "loss": 0.0133,
      "step": 2400
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.7603859901428223,
      "learning_rate": 1.1511111111111113e-05,
      "loss": 0.0268,
      "step": 2410
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.0016988619463518262,
      "learning_rate": 1.1466666666666668e-05,
      "loss": 0.0542,
      "step": 2420
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.005680690053850412,
      "learning_rate": 1.1422222222222223e-05,
      "loss": 0.0311,
      "step": 2430
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.9060475826263428,
      "learning_rate": 1.1377777777777779e-05,
      "loss": 0.1159,
      "step": 2440
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.005553722381592,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 0.0648,
      "step": 2450
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.020055294036865,
      "learning_rate": 1.1288888888888889e-05,
      "loss": 0.0312,
      "step": 2460
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.021578742191195488,
      "learning_rate": 1.1244444444444444e-05,
      "loss": 0.0285,
      "step": 2470
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.00151452433783561,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0666,
      "step": 2480
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.127124309539795,
      "learning_rate": 1.1155555555555556e-05,
      "loss": 0.0362,
      "step": 2490
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0015013846568763256,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.0575,
      "step": 2500
    },
    {
      "epoch": 0.78,
      "eval_loss": 0.019110437482595444,
      "eval_runtime": 61.8844,
      "eval_samples_per_second": 16.159,
      "eval_steps_per_second": 16.159,
      "step": 2500
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.0033025951124727726,
      "learning_rate": 1.1066666666666669e-05,
      "loss": 0.081,
      "step": 2510
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.0032579610124230385,
      "learning_rate": 1.1022222222222224e-05,
      "loss": 0.0207,
      "step": 2520
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.838714838027954,
      "learning_rate": 1.0977777777777779e-05,
      "loss": 0.1599,
      "step": 2530
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0018555274000391364,
      "learning_rate": 1.0933333333333334e-05,
      "loss": 0.002,
      "step": 2540
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0020195182878524065,
      "learning_rate": 1.088888888888889e-05,
      "loss": 0.008,
      "step": 2550
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0014295022701844573,
      "learning_rate": 1.0844444444444446e-05,
      "loss": 0.0276,
      "step": 2560
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.145263433456421,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.0248,
      "step": 2570
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.27980375289917,
      "learning_rate": 1.0755555555555557e-05,
      "loss": 0.0358,
      "step": 2580
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.667983293533325,
      "learning_rate": 1.0711111111111112e-05,
      "loss": 0.0651,
      "step": 2590
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.0018497890559956431,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.0387,
      "step": 2600
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.0013170776655897498,
      "learning_rate": 1.0622222222222223e-05,
      "loss": 0.0173,
      "step": 2610
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.004894789308309555,
      "learning_rate": 1.0577777777777778e-05,
      "loss": 0.043,
      "step": 2620
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.04001269116997719,
      "learning_rate": 1.0533333333333333e-05,
      "loss": 0.0926,
      "step": 2630
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.0019531013676896691,
      "learning_rate": 1.048888888888889e-05,
      "loss": 0.0511,
      "step": 2640
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.29565167427063,
      "learning_rate": 1.0444444444444445e-05,
      "loss": 0.0803,
      "step": 2650
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.7071011066436768,
      "learning_rate": 1.04e-05,
      "loss": 0.0604,
      "step": 2660
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.0376460887491703,
      "learning_rate": 1.0355555555555557e-05,
      "loss": 0.0155,
      "step": 2670
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.001858079107478261,
      "learning_rate": 1.0311111111111113e-05,
      "loss": 0.0665,
      "step": 2680
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.0015587485395371914,
      "learning_rate": 1.0266666666666668e-05,
      "loss": 0.0613,
      "step": 2690
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.047985997051000595,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 0.0032,
      "step": 2700
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.2135114669799805,
      "learning_rate": 1.0177777777777778e-05,
      "loss": 0.0508,
      "step": 2710
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.1044480800628662,
      "learning_rate": 1.0133333333333335e-05,
      "loss": 0.0333,
      "step": 2720
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.047782231122255325,
      "learning_rate": 1.008888888888889e-05,
      "loss": 0.0256,
      "step": 2730
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.0029533940833061934,
      "learning_rate": 1.0044444444444446e-05,
      "loss": 0.0107,
      "step": 2740
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.003864617319777608,
      "learning_rate": 1e-05,
      "loss": 0.0508,
      "step": 2750
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.006308398675173521,
      "learning_rate": 9.955555555555556e-06,
      "loss": 0.0599,
      "step": 2760
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0014458250952884555,
      "learning_rate": 9.911111111111113e-06,
      "loss": 0.0457,
      "step": 2770
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0018096743151545525,
      "learning_rate": 9.866666666666668e-06,
      "loss": 0.0773,
      "step": 2780
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.002803443931043148,
      "learning_rate": 9.822222222222223e-06,
      "loss": 0.0229,
      "step": 2790
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0039856089279055595,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.0738,
      "step": 2800
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0016119845677167177,
      "learning_rate": 9.733333333333334e-06,
      "loss": 0.1069,
      "step": 2810
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.044685885310173035,
      "learning_rate": 9.688888888888889e-06,
      "loss": 0.0718,
      "step": 2820
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5644834041595459,
      "learning_rate": 9.644444444444444e-06,
      "loss": 0.0539,
      "step": 2830
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.6698065400123596,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0671,
      "step": 2840
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.009338573552668095,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.0238,
      "step": 2850
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.002303204732015729,
      "learning_rate": 9.511111111111112e-06,
      "loss": 0.0531,
      "step": 2860
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.004544748924672604,
      "learning_rate": 9.466666666666667e-06,
      "loss": 0.133,
      "step": 2870
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.004034551791846752,
      "learning_rate": 9.422222222222222e-06,
      "loss": 0.0174,
      "step": 2880
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.0032248455099761486,
      "learning_rate": 9.377777777777779e-06,
      "loss": 0.0571,
      "step": 2890
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2451166957616806,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.0304,
      "step": 2900
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.001918753026984632,
      "learning_rate": 9.28888888888889e-06,
      "loss": 0.0826,
      "step": 2910
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.0014626473421230912,
      "learning_rate": 9.244444444444445e-06,
      "loss": 0.0288,
      "step": 2920
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.003409222001209855,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.041,
      "step": 2930
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.002315065125003457,
      "learning_rate": 9.155555555555557e-06,
      "loss": 0.0322,
      "step": 2940
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.002485796343535185,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.0132,
      "step": 2950
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0027836193330585957,
      "learning_rate": 9.066666666666667e-06,
      "loss": 0.0424,
      "step": 2960
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.2203580141067505,
      "learning_rate": 9.022222222222223e-06,
      "loss": 0.0607,
      "step": 2970
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0014286149526014924,
      "learning_rate": 8.977777777777778e-06,
      "loss": 0.0082,
      "step": 2980
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0011963992146775126,
      "learning_rate": 8.933333333333333e-06,
      "loss": 0.0447,
      "step": 2990
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0012861526338383555,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.0394,
      "step": 3000
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.021464653313159943,
      "eval_runtime": 62.0041,
      "eval_samples_per_second": 16.128,
      "eval_steps_per_second": 16.128,
      "step": 3000
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.6348252296447754,
      "learning_rate": 8.844444444444445e-06,
      "loss": 0.1294,
      "step": 3010
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.0012073700781911612,
      "learning_rate": 8.8e-06,
      "loss": 0.0304,
      "step": 3020
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.003195586148649454,
      "learning_rate": 8.755555555555556e-06,
      "loss": 0.0161,
      "step": 3030
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.9521057605743408,
      "learning_rate": 8.711111111111111e-06,
      "loss": 0.1185,
      "step": 3040
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.035520199686288834,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.0357,
      "step": 3050
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0031184712424874306,
      "learning_rate": 8.622222222222223e-06,
      "loss": 0.0505,
      "step": 3060
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8742607831954956,
      "learning_rate": 8.577777777777778e-06,
      "loss": 0.0607,
      "step": 3070
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.001605386845767498,
      "learning_rate": 8.533333333333335e-06,
      "loss": 0.048,
      "step": 3080
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.8198726177215576,
      "learning_rate": 8.48888888888889e-06,
      "loss": 0.029,
      "step": 3090
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.0011747152311727405,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.0745,
      "step": 3100
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.0037345418240875006,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.1246,
      "step": 3110
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.03968101739883423,
      "learning_rate": 8.355555555555556e-06,
      "loss": 0.028,
      "step": 3120
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.0017666455823928118,
      "learning_rate": 8.311111111111111e-06,
      "loss": 0.0398,
      "step": 3130
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.0016725193709135056,
      "learning_rate": 8.266666666666667e-06,
      "loss": 0.0874,
      "step": 3140
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.0152335166931152,
      "learning_rate": 8.222222222222222e-06,
      "loss": 0.0233,
      "step": 3150
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.003967667929828167,
      "learning_rate": 8.177777777777779e-06,
      "loss": 0.0322,
      "step": 3160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.0012324820272624493,
      "learning_rate": 8.133333333333334e-06,
      "loss": 0.01,
      "step": 3170
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.001890141749754548,
      "learning_rate": 8.08888888888889e-06,
      "loss": 0.0617,
      "step": 3180
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0017083596903830767,
      "learning_rate": 8.044444444444444e-06,
      "loss": 0.031,
      "step": 3190
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.1974055767059326,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0136,
      "step": 3200
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.8210293054580688,
      "learning_rate": 7.955555555555557e-06,
      "loss": 0.0174,
      "step": 3210
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.1489944458007812,
      "learning_rate": 7.911111111111112e-06,
      "loss": 0.0776,
      "step": 3220
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.0014464000705629587,
      "learning_rate": 7.866666666666667e-06,
      "loss": 0.0164,
      "step": 3230
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.5747243762016296,
      "learning_rate": 7.822222222222224e-06,
      "loss": 0.0093,
      "step": 3240
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.009251836687326431,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.0171,
      "step": 3250
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2069794237613678,
      "learning_rate": 7.733333333333334e-06,
      "loss": 0.0301,
      "step": 3260
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.251051664352417,
      "learning_rate": 7.68888888888889e-06,
      "loss": 0.0707,
      "step": 3270
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.0010599648812785745,
      "learning_rate": 7.644444444444445e-06,
      "loss": 0.0053,
      "step": 3280
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.0016899543115869164,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.0006,
      "step": 3290
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.003940447699278593,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.014,
      "step": 3300
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.7399486899375916,
      "learning_rate": 7.511111111111111e-06,
      "loss": 0.0309,
      "step": 3310
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.0006520531023852527,
      "learning_rate": 7.4666666666666675e-06,
      "loss": 0.0218,
      "step": 3320
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.1491100788116455,
      "learning_rate": 7.422222222222223e-06,
      "loss": 0.1215,
      "step": 3330
    },
    {
      "epoch": 1.05,
      "grad_norm": 3.6713123321533203,
      "learning_rate": 7.377777777777778e-06,
      "loss": 0.0429,
      "step": 3340
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.03133914992213249,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.027,
      "step": 3350
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.002692360896617174,
      "learning_rate": 7.28888888888889e-06,
      "loss": 0.0216,
      "step": 3360
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.0034821536391973495,
      "learning_rate": 7.244444444444445e-06,
      "loss": 0.043,
      "step": 3370
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.00455117505043745,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.0688,
      "step": 3380
    },
    {
      "epoch": 1.06,
      "grad_norm": 3.5100715160369873,
      "learning_rate": 7.155555555555556e-06,
      "loss": 0.0136,
      "step": 3390
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.0031513776630163193,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.0155,
      "step": 3400
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3787556886672974,
      "learning_rate": 7.066666666666667e-06,
      "loss": 0.0254,
      "step": 3410
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.000773483538068831,
      "learning_rate": 7.022222222222222e-06,
      "loss": 0.0031,
      "step": 3420
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.0014327614335343242,
      "learning_rate": 6.977777777777779e-06,
      "loss": 0.0376,
      "step": 3430
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0031394639518111944,
      "learning_rate": 6.9333333333333344e-06,
      "loss": 0.0596,
      "step": 3440
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0024705962277948856,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.0095,
      "step": 3450
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0005248151137493551,
      "learning_rate": 6.844444444444445e-06,
      "loss": 0.0199,
      "step": 3460
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.49637269973754883,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0154,
      "step": 3470
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3430148363113403,
      "learning_rate": 6.755555555555556e-06,
      "loss": 0.0395,
      "step": 3480
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.885108232498169,
      "learning_rate": 6.711111111111111e-06,
      "loss": 0.0472,
      "step": 3490
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.0007279856945388019,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0369,
      "step": 3500
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.022176509723067284,
      "eval_runtime": 61.8847,
      "eval_samples_per_second": 16.159,
      "eval_steps_per_second": 16.159,
      "step": 3500
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.003986152354627848,
      "learning_rate": 6.6222222222222236e-06,
      "loss": 0.0583,
      "step": 3510
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.0014766375534236431,
      "learning_rate": 6.577777777777779e-06,
      "loss": 0.0153,
      "step": 3520
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.791298508644104,
      "learning_rate": 6.533333333333334e-06,
      "loss": 0.0508,
      "step": 3530
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.02302105911076069,
      "learning_rate": 6.488888888888889e-06,
      "loss": 0.0275,
      "step": 3540
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.00240346509963274,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.0062,
      "step": 3550
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0011688686208799481,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0458,
      "step": 3560
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0007077022455632687,
      "learning_rate": 6.355555555555556e-06,
      "loss": 0.0274,
      "step": 3570
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0014391880249604583,
      "learning_rate": 6.311111111111111e-06,
      "loss": 0.0178,
      "step": 3580
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.002971325535327196,
      "learning_rate": 6.266666666666668e-06,
      "loss": 0.0099,
      "step": 3590
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5885258913040161,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.0277,
      "step": 3600
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.0023686839267611504,
      "learning_rate": 6.177777777777778e-06,
      "loss": 0.0001,
      "step": 3610
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.0018330584280192852,
      "learning_rate": 6.133333333333334e-06,
      "loss": 0.034,
      "step": 3620
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.5174291133880615,
      "learning_rate": 6.08888888888889e-06,
      "loss": 0.0095,
      "step": 3630
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.6314389705657959,
      "learning_rate": 6.044444444444445e-06,
      "loss": 0.035,
      "step": 3640
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.273723125457764,
      "learning_rate": 6e-06,
      "loss": 0.0332,
      "step": 3650
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.590868592262268,
      "learning_rate": 5.955555555555555e-06,
      "loss": 0.0398,
      "step": 3660
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.3291077613830566,
      "learning_rate": 5.911111111111112e-06,
      "loss": 0.0306,
      "step": 3670
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.0011222765315324068,
      "learning_rate": 5.8666666666666675e-06,
      "loss": 0.0194,
      "step": 3680
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.7731515169143677,
      "learning_rate": 5.822222222222223e-06,
      "loss": 0.0747,
      "step": 3690
    },
    {
      "epoch": 1.16,
      "grad_norm": 3.05898380279541,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.0239,
      "step": 3700
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.0017392480513080955,
      "learning_rate": 5.733333333333334e-06,
      "loss": 0.0415,
      "step": 3710
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.03126866742968559,
      "learning_rate": 5.688888888888889e-06,
      "loss": 0.0814,
      "step": 3720
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.0008881581015884876,
      "learning_rate": 5.6444444444444445e-06,
      "loss": 0.0211,
      "step": 3730
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.0007252222276292741,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0441,
      "step": 3740
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.002276132581755519,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.0433,
      "step": 3750
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0014605377800762653,
      "learning_rate": 5.511111111111112e-06,
      "loss": 0.0204,
      "step": 3760
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0017992451321333647,
      "learning_rate": 5.466666666666667e-06,
      "loss": 0.0992,
      "step": 3770
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0018204891821369529,
      "learning_rate": 5.422222222222223e-06,
      "loss": 0.0132,
      "step": 3780
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.16287018358707428,
      "learning_rate": 5.3777777777777784e-06,
      "loss": 0.0291,
      "step": 3790
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.21604575216770172,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0158,
      "step": 3800
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.0013267812319099903,
      "learning_rate": 5.288888888888889e-06,
      "loss": 0.0053,
      "step": 3810
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.0006841329741291702,
      "learning_rate": 5.244444444444445e-06,
      "loss": 0.0212,
      "step": 3820
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.19196736812591553,
      "learning_rate": 5.2e-06,
      "loss": 0.0206,
      "step": 3830
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.001533058239147067,
      "learning_rate": 5.155555555555556e-06,
      "loss": 0.0338,
      "step": 3840
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.001141805900260806,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.0006,
      "step": 3850
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.453786849975586,
      "learning_rate": 5.0666666666666676e-06,
      "loss": 0.0449,
      "step": 3860
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.0016863587079569697,
      "learning_rate": 5.022222222222223e-06,
      "loss": 0.0379,
      "step": 3870
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.0005151178920641541,
      "learning_rate": 4.977777777777778e-06,
      "loss": 0.0318,
      "step": 3880
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.0469911098480225,
      "learning_rate": 4.933333333333334e-06,
      "loss": 0.0185,
      "step": 3890
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.7001366019248962,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.0566,
      "step": 3900
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.0007412733975797892,
      "learning_rate": 4.8444444444444446e-06,
      "loss": 0.0003,
      "step": 3910
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.0008465189021080732,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0493,
      "step": 3920
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.3275091648101807,
      "learning_rate": 4.755555555555556e-06,
      "loss": 0.0109,
      "step": 3930
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.0014587095938622952,
      "learning_rate": 4.711111111111111e-06,
      "loss": 0.0,
      "step": 3940
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0013576536439359188,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0255,
      "step": 3950
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.008543136529624462,
      "learning_rate": 4.622222222222222e-06,
      "loss": 0.0352,
      "step": 3960
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0009798618266358972,
      "learning_rate": 4.5777777777777785e-06,
      "loss": 0.0556,
      "step": 3970
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.0005635471316054463,
      "learning_rate": 4.533333333333334e-06,
      "loss": 0.0325,
      "step": 3980
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.8276817202568054,
      "learning_rate": 4.488888888888889e-06,
      "loss": 0.0129,
      "step": 3990
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.013259605504572392,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0193,
      "step": 4000
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.022383807227015495,
      "eval_runtime": 61.879,
      "eval_samples_per_second": 16.161,
      "eval_steps_per_second": 16.161,
      "step": 4000
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.0017350054113194346,
      "learning_rate": 4.4e-06,
      "loss": 0.0464,
      "step": 4010
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.8484432697296143,
      "learning_rate": 4.3555555555555555e-06,
      "loss": 0.0215,
      "step": 4020
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.376138687133789,
      "learning_rate": 4.3111111111111115e-06,
      "loss": 0.0781,
      "step": 4030
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.0006440070574171841,
      "learning_rate": 4.266666666666668e-06,
      "loss": 0.01,
      "step": 4040
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.001261880504898727,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.1061,
      "step": 4050
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.9878569841384888,
      "learning_rate": 4.177777777777778e-06,
      "loss": 0.0254,
      "step": 4060
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.05665569379925728,
      "learning_rate": 4.133333333333333e-06,
      "loss": 0.0044,
      "step": 4070
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6482369303703308,
      "learning_rate": 4.088888888888889e-06,
      "loss": 0.0259,
      "step": 4080
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.0012170624686405063,
      "learning_rate": 4.044444444444445e-06,
      "loss": 0.0438,
      "step": 4090
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.062455616891384125,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0284,
      "step": 4100
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.0014791043940931559,
      "learning_rate": 3.955555555555556e-06,
      "loss": 0.0212,
      "step": 4110
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.445824384689331,
      "learning_rate": 3.911111111111112e-06,
      "loss": 0.0511,
      "step": 4120
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.981981039047241,
      "learning_rate": 3.866666666666667e-06,
      "loss": 0.0482,
      "step": 4130
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.04528600350022316,
      "learning_rate": 3.8222222222222224e-06,
      "loss": 0.0224,
      "step": 4140
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.0005888556479476392,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.0517,
      "step": 4150
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.5039355754852295,
      "learning_rate": 3.7333333333333337e-06,
      "loss": 0.037,
      "step": 4160
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.8256263732910156,
      "learning_rate": 3.688888888888889e-06,
      "loss": 0.0261,
      "step": 4170
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.0010332756210118532,
      "learning_rate": 3.644444444444445e-06,
      "loss": 0.0269,
      "step": 4180
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.0005628996295854449,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0053,
      "step": 4190
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.0006191189750097692,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.0056,
      "step": 4200
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.178837537765503,
      "learning_rate": 3.511111111111111e-06,
      "loss": 0.0136,
      "step": 4210
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.00813736766576767,
      "learning_rate": 3.4666666666666672e-06,
      "loss": 0.0249,
      "step": 4220
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0015987426741048694,
      "learning_rate": 3.4222222222222224e-06,
      "loss": 0.0424,
      "step": 4230
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0005763465305790305,
      "learning_rate": 3.377777777777778e-06,
      "loss": 0.0149,
      "step": 4240
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0005146890762262046,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0207,
      "step": 4250
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0012806435115635395,
      "learning_rate": 3.2888888888888894e-06,
      "loss": 0.0157,
      "step": 4260
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0015921818558126688,
      "learning_rate": 3.2444444444444446e-06,
      "loss": 0.029,
      "step": 4270
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.0711112022399902,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.1302,
      "step": 4280
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0007020393386483192,
      "learning_rate": 3.1555555555555555e-06,
      "loss": 0.0076,
      "step": 4290
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.0005048629245720804,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.0086,
      "step": 4300
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.0023996131494641304,
      "learning_rate": 3.066666666666667e-06,
      "loss": 0.0196,
      "step": 4310
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.0025332309305667877,
      "learning_rate": 3.0222222222222225e-06,
      "loss": 0.0461,
      "step": 4320
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0007468872354365885,
      "learning_rate": 2.9777777777777777e-06,
      "loss": 0.0282,
      "step": 4330
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0005446248687803745,
      "learning_rate": 2.9333333333333338e-06,
      "loss": 0.0469,
      "step": 4340
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0011565398890525103,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.0335,
      "step": 4350
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.8544560670852661,
      "learning_rate": 2.8444444444444446e-06,
      "loss": 0.0703,
      "step": 4360
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.215923547744751,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0396,
      "step": 4370
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.535111427307129,
      "learning_rate": 2.755555555555556e-06,
      "loss": 0.0719,
      "step": 4380
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.01463981345295906,
      "learning_rate": 2.7111111111111116e-06,
      "loss": 0.0246,
      "step": 4390
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.059357643127441,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.0367,
      "step": 4400
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.0006797238602302969,
      "learning_rate": 2.6222222222222225e-06,
      "loss": 0.0489,
      "step": 4410
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5824676752090454,
      "learning_rate": 2.577777777777778e-06,
      "loss": 0.0285,
      "step": 4420
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.00045088992919772863,
      "learning_rate": 2.5333333333333338e-06,
      "loss": 0.0328,
      "step": 4430
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.129044771194458,
      "learning_rate": 2.488888888888889e-06,
      "loss": 0.058,
      "step": 4440
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.00038907071575522423,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.0504,
      "step": 4450
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.0011422157986089587,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.022,
      "step": 4460
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.4815700054168701,
      "learning_rate": 2.3555555555555555e-06,
      "loss": 0.0416,
      "step": 4470
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.4049506187438965,
      "learning_rate": 2.311111111111111e-06,
      "loss": 0.0286,
      "step": 4480
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.32649314403533936,
      "learning_rate": 2.266666666666667e-06,
      "loss": 0.0191,
      "step": 4490
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.001270271954126656,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.0208,
      "step": 4500
    },
    {
      "epoch": 1.41,
      "eval_loss": 0.018216300755739212,
      "eval_runtime": 61.8518,
      "eval_samples_per_second": 16.168,
      "eval_steps_per_second": 16.168,
      "step": 4500
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.0010293573141098022,
      "learning_rate": 2.1777777777777777e-06,
      "loss": 0.0186,
      "step": 4510
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.47490406036376953,
      "learning_rate": 2.133333333333334e-06,
      "loss": 0.0181,
      "step": 4520
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.000406662467867136,
      "learning_rate": 2.088888888888889e-06,
      "loss": 0.0383,
      "step": 4530
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.0009539751335978508,
      "learning_rate": 2.0444444444444447e-06,
      "loss": 0.0139,
      "step": 4540
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0005080964183434844,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0084,
      "step": 4550
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0007780414307489991,
      "learning_rate": 1.955555555555556e-06,
      "loss": 0.0133,
      "step": 4560
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0003762560954783112,
      "learning_rate": 1.9111111111111112e-06,
      "loss": 0.0149,
      "step": 4570
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0010716378455981612,
      "learning_rate": 1.8666666666666669e-06,
      "loss": 0.0096,
      "step": 4580
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0012602531351149082,
      "learning_rate": 1.8222222222222225e-06,
      "loss": 0.0501,
      "step": 4590
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0011576686520129442,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.0367,
      "step": 4600
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.00044630825868807733,
      "learning_rate": 1.7333333333333336e-06,
      "loss": 0.0335,
      "step": 4610
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.0005398432840593159,
      "learning_rate": 1.688888888888889e-06,
      "loss": 0.0359,
      "step": 4620
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.7338486909866333,
      "learning_rate": 1.6444444444444447e-06,
      "loss": 0.0557,
      "step": 4630
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.0004819166788365692,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0,
      "step": 4640
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.6127287149429321,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.0276,
      "step": 4650
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.07772258669137955,
      "learning_rate": 1.5111111111111112e-06,
      "loss": 0.088,
      "step": 4660
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.0005863177357241511,
      "learning_rate": 1.4666666666666669e-06,
      "loss": 0.0132,
      "step": 4670
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.0006563079077750444,
      "learning_rate": 1.4222222222222223e-06,
      "loss": 0.0028,
      "step": 4680
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.803096294403076,
      "learning_rate": 1.377777777777778e-06,
      "loss": 0.0694,
      "step": 4690
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.7182486653327942,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.0065,
      "step": 4700
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.0009489316144026816,
      "learning_rate": 1.288888888888889e-06,
      "loss": 0.0136,
      "step": 4710
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.001954195322468877,
      "learning_rate": 1.2444444444444445e-06,
      "loss": 0.0142,
      "step": 4720
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.5916552543640137,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0469,
      "step": 4730
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.119339942932129,
      "learning_rate": 1.1555555555555556e-06,
      "loss": 0.0551,
      "step": 4740
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.0009017220581881702,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.0486,
      "step": 4750
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.0012501273304224014,
      "learning_rate": 1.066666666666667e-06,
      "loss": 0.0048,
      "step": 4760
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.8751838803291321,
      "learning_rate": 1.0222222222222223e-06,
      "loss": 0.0409,
      "step": 4770
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0013982664095237851,
      "learning_rate": 9.77777777777778e-07,
      "loss": 0.0577,
      "step": 4780
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0014040243113413453,
      "learning_rate": 9.333333333333334e-07,
      "loss": 0.031,
      "step": 4790
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.977519989013672,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.1057,
      "step": 4800
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.0011434502666816115,
      "learning_rate": 8.444444444444445e-07,
      "loss": 0.0066,
      "step": 4810
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.0013947113184258342,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.0337,
      "step": 4820
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.3012702167034149,
      "learning_rate": 7.555555555555556e-07,
      "loss": 0.0554,
      "step": 4830
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.1272149384021759,
      "learning_rate": 7.111111111111112e-07,
      "loss": 0.0208,
      "step": 4840
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.004117014352232218,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.0357,
      "step": 4850
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.0005114726372994483,
      "learning_rate": 6.222222222222223e-07,
      "loss": 0.0189,
      "step": 4860
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.000449322717031464,
      "learning_rate": 5.777777777777778e-07,
      "loss": 0.079,
      "step": 4870
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.3229221105575562,
      "learning_rate": 5.333333333333335e-07,
      "loss": 0.0197,
      "step": 4880
    },
    {
      "epoch": 1.53,
      "grad_norm": 4.149044990539551,
      "learning_rate": 4.88888888888889e-07,
      "loss": 0.0383,
      "step": 4890
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.00048808971769176424,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.0473,
      "step": 4900
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0010669564362615347,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.0838,
      "step": 4910
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.030236389487981796,
      "learning_rate": 3.555555555555556e-07,
      "loss": 0.0108,
      "step": 4920
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0004108142456971109,
      "learning_rate": 3.111111111111111e-07,
      "loss": 0.0301,
      "step": 4930
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.0008436466450802982,
      "learning_rate": 2.666666666666667e-07,
      "loss": 0.0101,
      "step": 4940
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.0007567039574496448,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.0237,
      "step": 4950
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.001360149122774601,
      "learning_rate": 1.777777777777778e-07,
      "loss": 0.0352,
      "step": 4960
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.0012200737837702036,
      "learning_rate": 1.3333333333333336e-07,
      "loss": 0.0482,
      "step": 4970
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.0006987242959439754,
      "learning_rate": 8.88888888888889e-08,
      "loss": 0.0164,
      "step": 4980
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.005137372761964798,
      "learning_rate": 4.444444444444445e-08,
      "loss": 0.0424,
      "step": 4990
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.0009189796401187778,
      "learning_rate": 0.0,
      "loss": 0.0353,
      "step": 5000
    },
    {
      "epoch": 1.57,
      "eval_loss": 0.021896515041589737,
      "eval_runtime": 61.7867,
      "eval_samples_per_second": 16.185,
      "eval_steps_per_second": 16.185,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 2500,
  "total_flos": 8.051062996992e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}