{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.080321285140563,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.10040160642570281,
      "grad_norm": 16.52980613708496,
      "learning_rate": 4.800000000000001e-07,
      "loss": 1.5503,
      "step": 25
    },
    {
      "epoch": 0.20080321285140562,
      "grad_norm": 11.904419898986816,
      "learning_rate": 9.800000000000001e-07,
      "loss": 0.7686,
      "step": 50
    },
    {
      "epoch": 0.30120481927710846,
      "grad_norm": 10.730138778686523,
      "learning_rate": 1.48e-06,
      "loss": 0.5875,
      "step": 75
    },
    {
      "epoch": 0.40160642570281124,
      "grad_norm": 9.80690860748291,
      "learning_rate": 1.98e-06,
      "loss": 0.5653,
      "step": 100
    },
    {
      "epoch": 0.5020080321285141,
      "grad_norm": 9.938033103942871,
      "learning_rate": 2.4800000000000004e-06,
      "loss": 0.5157,
      "step": 125
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 10.42677116394043,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.5222,
      "step": 150
    },
    {
      "epoch": 0.7028112449799196,
      "grad_norm": 10.295225143432617,
      "learning_rate": 3.46e-06,
      "loss": 0.4931,
      "step": 175
    },
    {
      "epoch": 0.8032128514056225,
      "grad_norm": 11.053289413452148,
      "learning_rate": 3.96e-06,
      "loss": 0.4682,
      "step": 200
    },
    {
      "epoch": 0.9036144578313253,
      "grad_norm": 10.42480754852295,
      "learning_rate": 4.4600000000000005e-06,
      "loss": 0.4781,
      "step": 225
    },
    {
      "epoch": 1.0040160642570282,
      "grad_norm": 8.67918586730957,
      "learning_rate": 4.960000000000001e-06,
      "loss": 0.4537,
      "step": 250
    },
    {
      "epoch": 1.104417670682731,
      "grad_norm": 10.315282821655273,
      "learning_rate": 5.460000000000001e-06,
      "loss": 0.3555,
      "step": 275
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 9.206184387207031,
      "learning_rate": 5.9600000000000005e-06,
      "loss": 0.3416,
      "step": 300
    },
    {
      "epoch": 1.3052208835341366,
      "grad_norm": 9.105973243713379,
      "learning_rate": 6.460000000000001e-06,
      "loss": 0.341,
      "step": 325
    },
    {
      "epoch": 1.4056224899598393,
      "grad_norm": 9.135680198669434,
      "learning_rate": 6.96e-06,
      "loss": 0.3486,
      "step": 350
    },
    {
      "epoch": 1.5060240963855422,
      "grad_norm": 8.966490745544434,
      "learning_rate": 7.4600000000000006e-06,
      "loss": 0.3703,
      "step": 375
    },
    {
      "epoch": 1.606425702811245,
      "grad_norm": 9.549636840820312,
      "learning_rate": 7.960000000000002e-06,
      "loss": 0.3481,
      "step": 400
    },
    {
      "epoch": 1.7068273092369477,
      "grad_norm": 8.577873229980469,
      "learning_rate": 8.46e-06,
      "loss": 0.3451,
      "step": 425
    },
    {
      "epoch": 1.8072289156626506,
      "grad_norm": 7.0459980964660645,
      "learning_rate": 8.96e-06,
      "loss": 0.3492,
      "step": 450
    },
    {
      "epoch": 1.9076305220883534,
      "grad_norm": 8.674671173095703,
      "learning_rate": 9.460000000000001e-06,
      "loss": 0.3398,
      "step": 475
    },
    {
      "epoch": 2.0080321285140563,
      "grad_norm": 6.492681980133057,
      "learning_rate": 9.960000000000001e-06,
      "loss": 0.3226,
      "step": 500
    },
    {
      "epoch": 2.108433734939759,
      "grad_norm": 8.931197166442871,
      "learning_rate": 9.94888888888889e-06,
      "loss": 0.2213,
      "step": 525
    },
    {
      "epoch": 2.208835341365462,
      "grad_norm": 8.153618812561035,
      "learning_rate": 9.893333333333334e-06,
      "loss": 0.199,
      "step": 550
    },
    {
      "epoch": 2.3092369477911645,
      "grad_norm": 6.562689304351807,
      "learning_rate": 9.837777777777778e-06,
      "loss": 0.2146,
      "step": 575
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 6.018052577972412,
      "learning_rate": 9.782222222222222e-06,
      "loss": 0.2127,
      "step": 600
    },
    {
      "epoch": 2.5100401606425704,
      "grad_norm": 6.111701011657715,
      "learning_rate": 9.726666666666668e-06,
      "loss": 0.2061,
      "step": 625
    },
    {
      "epoch": 2.610441767068273,
      "grad_norm": 6.124913215637207,
      "learning_rate": 9.671111111111112e-06,
      "loss": 0.2056,
      "step": 650
    },
    {
      "epoch": 2.710843373493976,
      "grad_norm": 5.902347087860107,
      "learning_rate": 9.615555555555558e-06,
      "loss": 0.203,
      "step": 675
    },
    {
      "epoch": 2.8112449799196786,
      "grad_norm": 6.789490699768066,
      "learning_rate": 9.56e-06,
      "loss": 0.2187,
      "step": 700
    },
    {
      "epoch": 2.9116465863453813,
      "grad_norm": 5.213627338409424,
      "learning_rate": 9.504444444444446e-06,
      "loss": 0.2085,
      "step": 725
    },
    {
      "epoch": 3.0120481927710845,
      "grad_norm": 4.021098613739014,
      "learning_rate": 9.44888888888889e-06,
      "loss": 0.2026,
      "step": 750
    },
    {
      "epoch": 3.112449799196787,
      "grad_norm": 5.916001319885254,
      "learning_rate": 9.393333333333334e-06,
      "loss": 0.1129,
      "step": 775
    },
    {
      "epoch": 3.21285140562249,
      "grad_norm": 5.063960552215576,
      "learning_rate": 9.33777777777778e-06,
      "loss": 0.116,
      "step": 800
    },
    {
      "epoch": 3.3132530120481927,
      "grad_norm": 5.580726146697998,
      "learning_rate": 9.282222222222222e-06,
      "loss": 0.1109,
      "step": 825
    },
    {
      "epoch": 3.4136546184738954,
      "grad_norm": 4.686521053314209,
      "learning_rate": 9.226666666666668e-06,
      "loss": 0.1187,
      "step": 850
    },
    {
      "epoch": 3.5140562248995986,
      "grad_norm": 5.128270149230957,
      "learning_rate": 9.171111111111112e-06,
      "loss": 0.1133,
      "step": 875
    },
    {
      "epoch": 3.6144578313253013,
      "grad_norm": 4.7233662605285645,
      "learning_rate": 9.115555555555556e-06,
      "loss": 0.1241,
      "step": 900
    },
    {
      "epoch": 3.714859437751004,
      "grad_norm": 5.689490795135498,
      "learning_rate": 9.060000000000001e-06,
      "loss": 0.1297,
      "step": 925
    },
    {
      "epoch": 3.8152610441767068,
      "grad_norm": 5.967415809631348,
      "learning_rate": 9.004444444444445e-06,
      "loss": 0.1251,
      "step": 950
    },
    {
      "epoch": 3.9156626506024095,
      "grad_norm": 4.376407146453857,
      "learning_rate": 8.94888888888889e-06,
      "loss": 0.1207,
      "step": 975
    },
    {
      "epoch": 4.016064257028113,
      "grad_norm": 4.3209662437438965,
      "learning_rate": 8.893333333333333e-06,
      "loss": 0.1125,
      "step": 1000
    },
    {
      "epoch": 4.016064257028113,
      "eval_loss": 0.3742842376232147,
      "eval_runtime": 1816.0709,
      "eval_samples_per_second": 2.963,
      "eval_steps_per_second": 0.186,
      "eval_wer": 0.2641258101713786,
      "step": 1000
    },
    {
      "epoch": 4.116465863453815,
      "grad_norm": 4.307711124420166,
      "learning_rate": 8.83777777777778e-06,
      "loss": 0.0669,
      "step": 1025
    },
    {
      "epoch": 4.216867469879518,
      "grad_norm": 3.720428228378296,
      "learning_rate": 8.782222222222223e-06,
      "loss": 0.0672,
      "step": 1050
    },
    {
      "epoch": 4.317269076305221,
      "grad_norm": 4.094714641571045,
      "learning_rate": 8.726666666666667e-06,
      "loss": 0.0644,
      "step": 1075
    },
    {
      "epoch": 4.417670682730924,
      "grad_norm": 4.306257247924805,
      "learning_rate": 8.671111111111113e-06,
      "loss": 0.0701,
      "step": 1100
    },
    {
      "epoch": 4.518072289156627,
      "grad_norm": 3.7728941440582275,
      "learning_rate": 8.615555555555555e-06,
      "loss": 0.0707,
      "step": 1125
    },
    {
      "epoch": 4.618473895582329,
      "grad_norm": 4.781138896942139,
      "learning_rate": 8.560000000000001e-06,
      "loss": 0.0722,
      "step": 1150
    },
    {
      "epoch": 4.718875502008032,
      "grad_norm": 3.8744442462921143,
      "learning_rate": 8.504444444444445e-06,
      "loss": 0.0708,
      "step": 1175
    },
    {
      "epoch": 4.8192771084337345,
      "grad_norm": 4.639457702636719,
      "learning_rate": 8.448888888888889e-06,
      "loss": 0.0712,
      "step": 1200
    },
    {
      "epoch": 4.919678714859438,
      "grad_norm": 2.9719793796539307,
      "learning_rate": 8.393333333333335e-06,
      "loss": 0.0748,
      "step": 1225
    },
    {
      "epoch": 5.020080321285141,
      "grad_norm": 2.4250712394714355,
      "learning_rate": 8.337777777777777e-06,
      "loss": 0.0601,
      "step": 1250
    },
    {
      "epoch": 5.120481927710843,
      "grad_norm": 3.8074028491973877,
      "learning_rate": 8.282222222222223e-06,
      "loss": 0.0396,
      "step": 1275
    },
    {
      "epoch": 5.220883534136546,
      "grad_norm": 2.799952507019043,
      "learning_rate": 8.226666666666667e-06,
      "loss": 0.0366,
      "step": 1300
    },
    {
      "epoch": 5.321285140562249,
      "grad_norm": 2.802058219909668,
      "learning_rate": 8.171111111111113e-06,
      "loss": 0.0452,
      "step": 1325
    },
    {
      "epoch": 5.421686746987952,
      "grad_norm": 2.9185872077941895,
      "learning_rate": 8.115555555555557e-06,
      "loss": 0.0429,
      "step": 1350
    },
    {
      "epoch": 5.522088353413655,
      "grad_norm": 4.0047526359558105,
      "learning_rate": 8.06e-06,
      "loss": 0.045,
      "step": 1375
    },
    {
      "epoch": 5.622489959839357,
      "grad_norm": 2.992342710494995,
      "learning_rate": 8.004444444444445e-06,
      "loss": 0.0466,
      "step": 1400
    },
    {
      "epoch": 5.72289156626506,
      "grad_norm": 4.288572311401367,
      "learning_rate": 7.948888888888889e-06,
      "loss": 0.0471,
      "step": 1425
    },
    {
      "epoch": 5.823293172690763,
      "grad_norm": 2.9833133220672607,
      "learning_rate": 7.893333333333335e-06,
      "loss": 0.0427,
      "step": 1450
    },
    {
      "epoch": 5.923694779116466,
      "grad_norm": 3.362567663192749,
      "learning_rate": 7.837777777777779e-06,
      "loss": 0.0503,
      "step": 1475
    },
    {
      "epoch": 6.024096385542169,
      "grad_norm": 3.4157121181488037,
      "learning_rate": 7.782222222222223e-06,
      "loss": 0.04,
      "step": 1500
    },
    {
      "epoch": 6.124497991967871,
      "grad_norm": 2.642629623413086,
      "learning_rate": 7.726666666666667e-06,
      "loss": 0.0266,
      "step": 1525
    },
    {
      "epoch": 6.224899598393574,
      "grad_norm": 2.336653470993042,
      "learning_rate": 7.67111111111111e-06,
      "loss": 0.028,
      "step": 1550
    },
    {
      "epoch": 6.325301204819277,
      "grad_norm": 2.498267889022827,
      "learning_rate": 7.6155555555555564e-06,
      "loss": 0.0268,
      "step": 1575
    },
    {
      "epoch": 6.42570281124498,
      "grad_norm": 2.4469289779663086,
      "learning_rate": 7.5600000000000005e-06,
      "loss": 0.0261,
      "step": 1600
    },
    {
      "epoch": 6.526104417670683,
      "grad_norm": 2.872375726699829,
      "learning_rate": 7.504444444444445e-06,
      "loss": 0.03,
      "step": 1625
    },
    {
      "epoch": 6.626506024096385,
      "grad_norm": 4.681485652923584,
      "learning_rate": 7.44888888888889e-06,
      "loss": 0.0279,
      "step": 1650
    },
    {
      "epoch": 6.7269076305220885,
      "grad_norm": 3.043527841567993,
      "learning_rate": 7.393333333333333e-06,
      "loss": 0.0305,
      "step": 1675
    },
    {
      "epoch": 6.827309236947791,
      "grad_norm": 3.1976921558380127,
      "learning_rate": 7.337777777777778e-06,
      "loss": 0.0339,
      "step": 1700
    },
    {
      "epoch": 6.927710843373494,
      "grad_norm": 2.7758069038391113,
      "learning_rate": 7.282222222222222e-06,
      "loss": 0.0299,
      "step": 1725
    },
    {
      "epoch": 7.028112449799197,
      "grad_norm": 2.653137445449829,
      "learning_rate": 7.226666666666667e-06,
      "loss": 0.0289,
      "step": 1750
    },
    {
      "epoch": 7.128514056224899,
      "grad_norm": 2.6387505531311035,
      "learning_rate": 7.171111111111112e-06,
      "loss": 0.0188,
      "step": 1775
    },
    {
      "epoch": 7.228915662650603,
      "grad_norm": 2.1044726371765137,
      "learning_rate": 7.115555555555557e-06,
      "loss": 0.0198,
      "step": 1800
    },
    {
      "epoch": 7.329317269076305,
      "grad_norm": 2.4335100650787354,
      "learning_rate": 7.06e-06,
      "loss": 0.0187,
      "step": 1825
    },
    {
      "epoch": 7.429718875502008,
      "grad_norm": 2.240412473678589,
      "learning_rate": 7.004444444444445e-06,
      "loss": 0.0209,
      "step": 1850
    },
    {
      "epoch": 7.530120481927711,
      "grad_norm": 3.036201238632202,
      "learning_rate": 6.948888888888889e-06,
      "loss": 0.0209,
      "step": 1875
    },
    {
      "epoch": 7.6305220883534135,
      "grad_norm": 1.8761370182037354,
      "learning_rate": 6.893333333333334e-06,
      "loss": 0.0204,
      "step": 1900
    },
    {
      "epoch": 7.730923694779117,
      "grad_norm": 3.35725474357605,
      "learning_rate": 6.837777777777779e-06,
      "loss": 0.0193,
      "step": 1925
    },
    {
      "epoch": 7.831325301204819,
      "grad_norm": 2.9908254146575928,
      "learning_rate": 6.782222222222222e-06,
      "loss": 0.0204,
      "step": 1950
    },
    {
      "epoch": 7.931726907630522,
      "grad_norm": 4.191911220550537,
      "learning_rate": 6.726666666666667e-06,
      "loss": 0.022,
      "step": 1975
    },
    {
      "epoch": 8.032128514056225,
      "grad_norm": 1.2732858657836914,
      "learning_rate": 6.671111111111112e-06,
      "loss": 0.0178,
      "step": 2000
    },
    {
      "epoch": 8.032128514056225,
      "eval_loss": 0.42306816577911377,
      "eval_runtime": 1886.6147,
      "eval_samples_per_second": 2.852,
      "eval_steps_per_second": 0.179,
      "eval_wer": 0.249890651715774,
      "step": 2000
    },
    {
      "epoch": 8.132530120481928,
      "grad_norm": 4.009426593780518,
      "learning_rate": 6.615555555555556e-06,
      "loss": 0.0132,
      "step": 2025
    },
    {
      "epoch": 8.23293172690763,
      "grad_norm": 3.8768014907836914,
      "learning_rate": 6.560000000000001e-06,
      "loss": 0.0148,
      "step": 2050
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 2.1475863456726074,
      "learning_rate": 6.504444444444446e-06,
      "loss": 0.0145,
      "step": 2075
    },
    {
      "epoch": 8.433734939759036,
      "grad_norm": 1.8194453716278076,
      "learning_rate": 6.448888888888889e-06,
      "loss": 0.0151,
      "step": 2100
    },
    {
      "epoch": 8.534136546184738,
      "grad_norm": 2.893296241760254,
      "learning_rate": 6.393333333333334e-06,
      "loss": 0.015,
      "step": 2125
    },
    {
      "epoch": 8.634538152610443,
      "grad_norm": 1.6417754888534546,
      "learning_rate": 6.3377777777777786e-06,
      "loss": 0.0164,
      "step": 2150
    },
    {
      "epoch": 8.734939759036145,
      "grad_norm": 2.3964133262634277,
      "learning_rate": 6.282222222222223e-06,
      "loss": 0.0167,
      "step": 2175
    },
    {
      "epoch": 8.835341365461847,
      "grad_norm": 2.9204821586608887,
      "learning_rate": 6.2266666666666675e-06,
      "loss": 0.0135,
      "step": 2200
    },
    {
      "epoch": 8.93574297188755,
      "grad_norm": 3.2966790199279785,
      "learning_rate": 6.171111111111112e-06,
      "loss": 0.0133,
      "step": 2225
    },
    {
      "epoch": 9.036144578313253,
      "grad_norm": 1.6279815435409546,
      "learning_rate": 6.1155555555555555e-06,
      "loss": 0.0146,
      "step": 2250
    },
    {
      "epoch": 9.136546184738956,
      "grad_norm": 3.7027101516723633,
      "learning_rate": 6.0600000000000004e-06,
      "loss": 0.0104,
      "step": 2275
    },
    {
      "epoch": 9.236947791164658,
      "grad_norm": 2.011039972305298,
      "learning_rate": 6.006666666666667e-06,
      "loss": 0.0108,
      "step": 2300
    },
    {
      "epoch": 9.337349397590362,
      "grad_norm": 2.4486405849456787,
      "learning_rate": 5.951111111111112e-06,
      "loss": 0.01,
      "step": 2325
    },
    {
      "epoch": 9.437751004016064,
      "grad_norm": 1.047418475151062,
      "learning_rate": 5.895555555555557e-06,
      "loss": 0.012,
      "step": 2350
    },
    {
      "epoch": 9.538152610441767,
      "grad_norm": 2.6720094680786133,
      "learning_rate": 5.84e-06,
      "loss": 0.011,
      "step": 2375
    },
    {
      "epoch": 9.638554216867469,
      "grad_norm": 1.3718459606170654,
      "learning_rate": 5.784444444444445e-06,
      "loss": 0.0122,
      "step": 2400
    },
    {
      "epoch": 9.738955823293173,
      "grad_norm": 1.6287081241607666,
      "learning_rate": 5.72888888888889e-06,
      "loss": 0.009,
      "step": 2425
    },
    {
      "epoch": 9.839357429718875,
      "grad_norm": 2.190797805786133,
      "learning_rate": 5.673333333333334e-06,
      "loss": 0.0132,
      "step": 2450
    },
    {
      "epoch": 9.939759036144578,
      "grad_norm": 2.275399684906006,
      "learning_rate": 5.617777777777779e-06,
      "loss": 0.0105,
      "step": 2475
    },
    {
      "epoch": 10.040160642570282,
      "grad_norm": 1.9050276279449463,
      "learning_rate": 5.562222222222222e-06,
      "loss": 0.0107,
      "step": 2500
    },
    {
      "epoch": 10.140562248995984,
      "grad_norm": 0.9361718893051147,
      "learning_rate": 5.506666666666667e-06,
      "loss": 0.0074,
      "step": 2525
    },
    {
      "epoch": 10.240963855421686,
      "grad_norm": 2.5661396980285645,
      "learning_rate": 5.451111111111112e-06,
      "loss": 0.0081,
      "step": 2550
    },
    {
      "epoch": 10.34136546184739,
      "grad_norm": 1.1169323921203613,
      "learning_rate": 5.3955555555555565e-06,
      "loss": 0.0088,
      "step": 2575
    },
    {
      "epoch": 10.441767068273093,
      "grad_norm": 1.0962170362472534,
      "learning_rate": 5.3400000000000005e-06,
      "loss": 0.0077,
      "step": 2600
    },
    {
      "epoch": 10.542168674698795,
      "grad_norm": 1.0353940725326538,
      "learning_rate": 5.2844444444444454e-06,
      "loss": 0.0096,
      "step": 2625
    },
    {
      "epoch": 10.642570281124499,
      "grad_norm": 3.781886339187622,
      "learning_rate": 5.228888888888889e-06,
      "loss": 0.0082,
      "step": 2650
    },
    {
      "epoch": 10.742971887550201,
      "grad_norm": 3.071993112564087,
      "learning_rate": 5.1733333333333335e-06,
      "loss": 0.0076,
      "step": 2675
    },
    {
      "epoch": 10.843373493975903,
      "grad_norm": 1.5968672037124634,
      "learning_rate": 5.117777777777778e-06,
      "loss": 0.0066,
      "step": 2700
    },
    {
      "epoch": 10.943775100401606,
      "grad_norm": 1.1753120422363281,
      "learning_rate": 5.062222222222222e-06,
      "loss": 0.0079,
      "step": 2725
    },
    {
      "epoch": 11.04417670682731,
      "grad_norm": 0.6646733283996582,
      "learning_rate": 5.006666666666667e-06,
      "loss": 0.0068,
      "step": 2750
    },
    {
      "epoch": 11.144578313253012,
      "grad_norm": 1.3602428436279297,
      "learning_rate": 4.951111111111111e-06,
      "loss": 0.0066,
      "step": 2775
    },
    {
      "epoch": 11.244979919678714,
      "grad_norm": 0.6632346510887146,
      "learning_rate": 4.895555555555556e-06,
      "loss": 0.0042,
      "step": 2800
    },
    {
      "epoch": 11.345381526104418,
      "grad_norm": 2.823923349380493,
      "learning_rate": 4.84e-06,
      "loss": 0.0044,
      "step": 2825
    },
    {
      "epoch": 11.44578313253012,
      "grad_norm": 1.5665075778961182,
      "learning_rate": 4.784444444444445e-06,
      "loss": 0.0048,
      "step": 2850
    },
    {
      "epoch": 11.546184738955823,
      "grad_norm": 2.2108278274536133,
      "learning_rate": 4.728888888888889e-06,
      "loss": 0.006,
      "step": 2875
    },
    {
      "epoch": 11.646586345381525,
      "grad_norm": 0.9662758111953735,
      "learning_rate": 4.673333333333333e-06,
      "loss": 0.0049,
      "step": 2900
    },
    {
      "epoch": 11.74698795180723,
      "grad_norm": 1.4958486557006836,
      "learning_rate": 4.617777777777778e-06,
      "loss": 0.0062,
      "step": 2925
    },
    {
      "epoch": 11.847389558232932,
      "grad_norm": 2.0174615383148193,
      "learning_rate": 4.562222222222222e-06,
      "loss": 0.0058,
      "step": 2950
    },
    {
      "epoch": 11.947791164658634,
      "grad_norm": 2.672804594039917,
      "learning_rate": 4.506666666666667e-06,
      "loss": 0.0056,
      "step": 2975
    },
    {
      "epoch": 12.048192771084338,
      "grad_norm": 0.4241655170917511,
      "learning_rate": 4.451111111111112e-06,
      "loss": 0.0064,
      "step": 3000
    },
    {
      "epoch": 12.048192771084338,
      "eval_loss": 0.46265721321105957,
      "eval_runtime": 1822.5129,
      "eval_samples_per_second": 2.953,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.2341246172810052,
      "step": 3000
    },
    {
      "epoch": 12.14859437751004,
      "grad_norm": 0.3871651589870453,
      "learning_rate": 4.395555555555556e-06,
      "loss": 0.0035,
      "step": 3025
    },
    {
      "epoch": 12.248995983935743,
      "grad_norm": 0.2253613919019699,
      "learning_rate": 4.34e-06,
      "loss": 0.0029,
      "step": 3050
    },
    {
      "epoch": 12.349397590361447,
      "grad_norm": 0.38276419043540955,
      "learning_rate": 4.284444444444445e-06,
      "loss": 0.0037,
      "step": 3075
    },
    {
      "epoch": 12.449799196787149,
      "grad_norm": 0.8364894390106201,
      "learning_rate": 4.228888888888889e-06,
      "loss": 0.0036,
      "step": 3100
    },
    {
      "epoch": 12.550200803212851,
      "grad_norm": 0.37833768129348755,
      "learning_rate": 4.173333333333334e-06,
      "loss": 0.0027,
      "step": 3125
    },
    {
      "epoch": 12.650602409638553,
      "grad_norm": 0.246841162443161,
      "learning_rate": 4.117777777777779e-06,
      "loss": 0.0044,
      "step": 3150
    },
    {
      "epoch": 12.751004016064257,
      "grad_norm": 0.3080684542655945,
      "learning_rate": 4.062222222222223e-06,
      "loss": 0.0039,
      "step": 3175
    },
    {
      "epoch": 12.85140562248996,
      "grad_norm": 0.14490999281406403,
      "learning_rate": 4.006666666666667e-06,
      "loss": 0.0045,
      "step": 3200
    },
    {
      "epoch": 12.951807228915662,
      "grad_norm": 0.6069326996803284,
      "learning_rate": 3.951111111111112e-06,
      "loss": 0.0031,
      "step": 3225
    },
    {
      "epoch": 13.052208835341366,
      "grad_norm": 0.8227677941322327,
      "learning_rate": 3.895555555555556e-06,
      "loss": 0.0023,
      "step": 3250
    },
    {
      "epoch": 13.152610441767068,
      "grad_norm": 1.3870292901992798,
      "learning_rate": 3.8400000000000005e-06,
      "loss": 0.0031,
      "step": 3275
    },
    {
      "epoch": 13.25301204819277,
      "grad_norm": 2.032243490219116,
      "learning_rate": 3.784444444444445e-06,
      "loss": 0.0021,
      "step": 3300
    },
    {
      "epoch": 13.353413654618475,
      "grad_norm": 3.4899916648864746,
      "learning_rate": 3.728888888888889e-06,
      "loss": 0.0022,
      "step": 3325
    },
    {
      "epoch": 13.453815261044177,
      "grad_norm": 1.3607572317123413,
      "learning_rate": 3.673333333333334e-06,
      "loss": 0.0017,
      "step": 3350
    },
    {
      "epoch": 13.55421686746988,
      "grad_norm": 0.21469266712665558,
      "learning_rate": 3.617777777777778e-06,
      "loss": 0.0018,
      "step": 3375
    },
    {
      "epoch": 13.654618473895582,
      "grad_norm": 0.11461154371500015,
      "learning_rate": 3.5622222222222224e-06,
      "loss": 0.0011,
      "step": 3400
    },
    {
      "epoch": 13.755020080321286,
      "grad_norm": 3.34506893157959,
      "learning_rate": 3.5066666666666673e-06,
      "loss": 0.0018,
      "step": 3425
    },
    {
      "epoch": 13.855421686746988,
      "grad_norm": 0.4698576331138611,
      "learning_rate": 3.4511111111111113e-06,
      "loss": 0.0022,
      "step": 3450
    },
    {
      "epoch": 13.95582329317269,
      "grad_norm": 0.8212801814079285,
      "learning_rate": 3.3955555555555558e-06,
      "loss": 0.0014,
      "step": 3475
    },
    {
      "epoch": 14.056224899598394,
      "grad_norm": 0.26110076904296875,
      "learning_rate": 3.3400000000000006e-06,
      "loss": 0.0014,
      "step": 3500
    },
    {
      "epoch": 14.156626506024097,
      "grad_norm": 0.08139172196388245,
      "learning_rate": 3.2844444444444447e-06,
      "loss": 0.0009,
      "step": 3525
    },
    {
      "epoch": 14.257028112449799,
      "grad_norm": 0.040998633950948715,
      "learning_rate": 3.228888888888889e-06,
      "loss": 0.0009,
      "step": 3550
    },
    {
      "epoch": 14.357429718875501,
      "grad_norm": 0.1332196742296219,
      "learning_rate": 3.173333333333334e-06,
      "loss": 0.001,
      "step": 3575
    },
    {
      "epoch": 14.457831325301205,
      "grad_norm": 0.0915873721241951,
      "learning_rate": 3.117777777777778e-06,
      "loss": 0.0019,
      "step": 3600
    },
    {
      "epoch": 14.558232931726907,
      "grad_norm": 0.11250325292348862,
      "learning_rate": 3.0622222222222225e-06,
      "loss": 0.0013,
      "step": 3625
    },
    {
      "epoch": 14.65863453815261,
      "grad_norm": 0.06163531169295311,
      "learning_rate": 3.0066666666666674e-06,
      "loss": 0.0009,
      "step": 3650
    },
    {
      "epoch": 14.759036144578314,
      "grad_norm": 0.2607724368572235,
      "learning_rate": 2.9511111111111114e-06,
      "loss": 0.0009,
      "step": 3675
    },
    {
      "epoch": 14.859437751004016,
      "grad_norm": 0.0904630720615387,
      "learning_rate": 2.895555555555556e-06,
      "loss": 0.001,
      "step": 3700
    },
    {
      "epoch": 14.959839357429718,
      "grad_norm": 0.04550072178244591,
      "learning_rate": 2.84e-06,
      "loss": 0.001,
      "step": 3725
    },
    {
      "epoch": 15.060240963855422,
      "grad_norm": 0.48310792446136475,
      "learning_rate": 2.784444444444445e-06,
      "loss": 0.0008,
      "step": 3750
    },
    {
      "epoch": 15.160642570281125,
      "grad_norm": 0.037569478154182434,
      "learning_rate": 2.7288888888888893e-06,
      "loss": 0.0005,
      "step": 3775
    },
    {
      "epoch": 15.261044176706827,
      "grad_norm": 0.03933248668909073,
      "learning_rate": 2.6733333333333333e-06,
      "loss": 0.0006,
      "step": 3800
    },
    {
      "epoch": 15.36144578313253,
      "grad_norm": 0.03279777243733406,
      "learning_rate": 2.617777777777778e-06,
      "loss": 0.0014,
      "step": 3825
    },
    {
      "epoch": 15.461847389558233,
      "grad_norm": 0.0519753098487854,
      "learning_rate": 2.5622222222222226e-06,
      "loss": 0.0007,
      "step": 3850
    },
    {
      "epoch": 15.562248995983936,
      "grad_norm": 0.06775977462530136,
      "learning_rate": 2.5066666666666667e-06,
      "loss": 0.001,
      "step": 3875
    },
    {
      "epoch": 15.662650602409638,
      "grad_norm": 0.962402880191803,
      "learning_rate": 2.451111111111111e-06,
      "loss": 0.0011,
      "step": 3900
    },
    {
      "epoch": 15.763052208835342,
      "grad_norm": 0.03904138505458832,
      "learning_rate": 2.3955555555555556e-06,
      "loss": 0.0007,
      "step": 3925
    },
    {
      "epoch": 15.863453815261044,
      "grad_norm": 0.041153594851493835,
      "learning_rate": 2.3400000000000005e-06,
      "loss": 0.0005,
      "step": 3950
    },
    {
      "epoch": 15.963855421686747,
      "grad_norm": 0.0350770466029644,
      "learning_rate": 2.2844444444444445e-06,
      "loss": 0.0005,
      "step": 3975
    },
    {
      "epoch": 16.06425702811245,
      "grad_norm": 0.027113359421491623,
      "learning_rate": 2.228888888888889e-06,
      "loss": 0.0004,
      "step": 4000
    },
    {
      "epoch": 16.06425702811245,
      "eval_loss": 0.4913391172885895,
      "eval_runtime": 2057.5949,
      "eval_samples_per_second": 2.615,
      "eval_steps_per_second": 0.164,
      "eval_wer": 0.22008827388762972,
      "step": 4000
    },
    {
      "epoch": 16.164658634538153,
      "grad_norm": 0.027138447389006615,
      "learning_rate": 2.1733333333333334e-06,
      "loss": 0.0004,
      "step": 4025
    },
    {
      "epoch": 16.265060240963855,
      "grad_norm": 0.02426302060484886,
      "learning_rate": 2.117777777777778e-06,
      "loss": 0.0004,
      "step": 4050
    },
    {
      "epoch": 16.365461847389557,
      "grad_norm": 0.04138059541583061,
      "learning_rate": 2.0622222222222223e-06,
      "loss": 0.0004,
      "step": 4075
    },
    {
      "epoch": 16.46586345381526,
      "grad_norm": 0.056056153029203415,
      "learning_rate": 2.006666666666667e-06,
      "loss": 0.0004,
      "step": 4100
    },
    {
      "epoch": 16.566265060240966,
      "grad_norm": 0.024884460493922234,
      "learning_rate": 1.9511111111111113e-06,
      "loss": 0.0004,
      "step": 4125
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.023256096988916397,
      "learning_rate": 1.8955555555555557e-06,
      "loss": 0.0003,
      "step": 4150
    },
    {
      "epoch": 16.76706827309237,
      "grad_norm": 0.030147407203912735,
      "learning_rate": 1.8400000000000002e-06,
      "loss": 0.0005,
      "step": 4175
    },
    {
      "epoch": 16.867469879518072,
      "grad_norm": 0.02672114223241806,
      "learning_rate": 1.7844444444444444e-06,
      "loss": 0.0003,
      "step": 4200
    },
    {
      "epoch": 16.967871485943775,
      "grad_norm": 0.02835569903254509,
      "learning_rate": 1.728888888888889e-06,
      "loss": 0.0003,
      "step": 4225
    },
    {
      "epoch": 17.068273092369477,
      "grad_norm": 0.02380695752799511,
      "learning_rate": 1.6733333333333335e-06,
      "loss": 0.0005,
      "step": 4250
    },
    {
      "epoch": 17.16867469879518,
      "grad_norm": 0.0286069605499506,
      "learning_rate": 1.6177777777777778e-06,
      "loss": 0.0004,
      "step": 4275
    },
    {
      "epoch": 17.269076305220885,
      "grad_norm": 0.030239233747124672,
      "learning_rate": 1.5622222222222225e-06,
      "loss": 0.0003,
      "step": 4300
    },
    {
      "epoch": 17.369477911646587,
      "grad_norm": 0.02350962907075882,
      "learning_rate": 1.506666666666667e-06,
      "loss": 0.0003,
      "step": 4325
    },
    {
      "epoch": 17.46987951807229,
      "grad_norm": 0.021099906414747238,
      "learning_rate": 1.4511111111111112e-06,
      "loss": 0.0003,
      "step": 4350
    },
    {
      "epoch": 17.570281124497992,
      "grad_norm": 0.020377565175294876,
      "learning_rate": 1.3955555555555556e-06,
      "loss": 0.0004,
      "step": 4375
    },
    {
      "epoch": 17.670682730923694,
      "grad_norm": 0.025648899376392365,
      "learning_rate": 1.34e-06,
      "loss": 0.0003,
      "step": 4400
    },
    {
      "epoch": 17.771084337349397,
      "grad_norm": 0.024144701659679413,
      "learning_rate": 1.2844444444444445e-06,
      "loss": 0.0003,
      "step": 4425
    },
    {
      "epoch": 17.8714859437751,
      "grad_norm": 0.02264849655330181,
      "learning_rate": 1.228888888888889e-06,
      "loss": 0.0003,
      "step": 4450
    },
    {
      "epoch": 17.971887550200805,
      "grad_norm": 0.02645409293472767,
      "learning_rate": 1.1733333333333335e-06,
      "loss": 0.0003,
      "step": 4475
    },
    {
      "epoch": 18.072289156626507,
      "grad_norm": 0.018070507794618607,
      "learning_rate": 1.117777777777778e-06,
      "loss": 0.0003,
      "step": 4500
    },
    {
      "epoch": 18.17269076305221,
      "grad_norm": 0.021406978368759155,
      "learning_rate": 1.0622222222222222e-06,
      "loss": 0.0003,
      "step": 4525
    },
    {
      "epoch": 18.27309236947791,
      "grad_norm": 0.01881973072886467,
      "learning_rate": 1.0066666666666668e-06,
      "loss": 0.0003,
      "step": 4550
    },
    {
      "epoch": 18.373493975903614,
      "grad_norm": 0.018671726807951927,
      "learning_rate": 9.511111111111111e-07,
      "loss": 0.0003,
      "step": 4575
    },
    {
      "epoch": 18.473895582329316,
      "grad_norm": 0.014478115364909172,
      "learning_rate": 8.955555555555557e-07,
      "loss": 0.0004,
      "step": 4600
    },
    {
      "epoch": 18.57429718875502,
      "grad_norm": 0.02397969737648964,
      "learning_rate": 8.400000000000001e-07,
      "loss": 0.0003,
      "step": 4625
    },
    {
      "epoch": 18.674698795180724,
      "grad_norm": 0.020082278177142143,
      "learning_rate": 7.844444444444445e-07,
      "loss": 0.0003,
      "step": 4650
    },
    {
      "epoch": 18.775100401606426,
      "grad_norm": 0.02104588970541954,
      "learning_rate": 7.28888888888889e-07,
      "loss": 0.0003,
      "step": 4675
    },
    {
      "epoch": 18.87550200803213,
      "grad_norm": 0.018279341980814934,
      "learning_rate": 6.733333333333334e-07,
      "loss": 0.0003,
      "step": 4700
    },
    {
      "epoch": 18.97590361445783,
      "grad_norm": 0.021302426233887672,
      "learning_rate": 6.177777777777778e-07,
      "loss": 0.0003,
      "step": 4725
    },
    {
      "epoch": 19.076305220883533,
      "grad_norm": 0.020490070804953575,
      "learning_rate": 5.622222222222223e-07,
      "loss": 0.0003,
      "step": 4750
    },
    {
      "epoch": 19.176706827309236,
      "grad_norm": 0.020895950496196747,
      "learning_rate": 5.066666666666667e-07,
      "loss": 0.0003,
      "step": 4775
    },
    {
      "epoch": 19.27710843373494,
      "grad_norm": 0.01935204304754734,
      "learning_rate": 4.511111111111111e-07,
      "loss": 0.0003,
      "step": 4800
    },
    {
      "epoch": 19.377510040160644,
      "grad_norm": 0.016865380108356476,
      "learning_rate": 3.9555555555555557e-07,
      "loss": 0.0003,
      "step": 4825
    },
    {
      "epoch": 19.477911646586346,
      "grad_norm": 0.01524381060153246,
      "learning_rate": 3.4000000000000003e-07,
      "loss": 0.0003,
      "step": 4850
    },
    {
      "epoch": 19.57831325301205,
      "grad_norm": 0.014133242890238762,
      "learning_rate": 2.844444444444445e-07,
      "loss": 0.0003,
      "step": 4875
    },
    {
      "epoch": 19.67871485943775,
      "grad_norm": 0.014944775030016899,
      "learning_rate": 2.2888888888888892e-07,
      "loss": 0.0002,
      "step": 4900
    },
    {
      "epoch": 19.779116465863453,
      "grad_norm": 0.016863036900758743,
      "learning_rate": 1.7333333333333335e-07,
      "loss": 0.0003,
      "step": 4925
    },
    {
      "epoch": 19.879518072289155,
      "grad_norm": 0.022053442895412445,
      "learning_rate": 1.1777777777777778e-07,
      "loss": 0.0003,
      "step": 4950
    },
    {
      "epoch": 19.97991967871486,
      "grad_norm": 0.01967422477900982,
      "learning_rate": 6.222222222222223e-08,
      "loss": 0.0003,
      "step": 4975
    },
    {
      "epoch": 20.080321285140563,
      "grad_norm": 0.017572658136487007,
      "learning_rate": 6.666666666666667e-09,
      "loss": 0.0003,
      "step": 5000
    },
    {
      "epoch": 20.080321285140563,
      "eval_loss": 0.4997508227825165,
      "eval_runtime": 2038.971,
      "eval_samples_per_second": 2.639,
      "eval_steps_per_second": 0.165,
      "eval_wer": 0.2181001232653386,
      "step": 5000
    },
    {
      "epoch": 20.080321285140563,
      "step": 5000,
      "total_flos": 2.727921844224e+20,
      "train_loss": 0.07613397375550121,
      "train_runtime": 53449.9289,
      "train_samples_per_second": 2.993,
      "train_steps_per_second": 0.094
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 21,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.727921844224e+20,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}