{
  "best_metric": 14.785271867071973,
  "best_model_checkpoint": "C:/song/final project/model\\checkpoint-3200",
  "epoch": 0.512,
  "eval_steps": 200,
  "global_step": 3200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 3.1421,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.7048,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.1335,
      "step": 75
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.4821,
      "step": 100
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.1068,
      "step": 125
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.9e-06,
      "loss": 0.9643,
      "step": 150
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.804,
      "step": 175
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.7481,
      "step": 200
    },
    {
      "epoch": 0.03,
      "eval_cer": 44.84835325577674,
      "eval_loss": 0.7554882764816284,
      "eval_runtime": 870.4091,
      "eval_samples_per_second": 2.924,
      "eval_steps_per_second": 0.366,
      "step": 200
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.4e-06,
      "loss": 0.7255,
      "step": 225
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.6376,
      "step": 250
    },
    {
      "epoch": 0.04,
      "learning_rate": 5.400000000000001e-06,
      "loss": 0.6077,
      "step": 275
    },
    {
      "epoch": 0.05,
      "learning_rate": 5.9e-06,
      "loss": 0.6158,
      "step": 300
    },
    {
      "epoch": 0.05,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.5763,
      "step": 325
    },
    {
      "epoch": 0.06,
      "learning_rate": 6.9e-06,
      "loss": 0.5964,
      "step": 350
    },
    {
      "epoch": 0.06,
      "learning_rate": 7.4e-06,
      "loss": 0.5576,
      "step": 375
    },
    {
      "epoch": 0.06,
      "learning_rate": 7.9e-06,
      "loss": 0.5393,
      "step": 400
    },
    {
      "epoch": 0.06,
      "eval_cer": 20.696866500582818,
      "eval_loss": 0.5607804656028748,
      "eval_runtime": 766.1672,
      "eval_samples_per_second": 3.322,
      "eval_steps_per_second": 0.416,
      "step": 400
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.57,
      "step": 425
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.5365,
      "step": 450
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.4e-06,
      "loss": 0.5101,
      "step": 475
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.9e-06,
      "loss": 0.5088,
      "step": 500
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.955555555555556e-06,
      "loss": 0.5239,
      "step": 525
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.9e-06,
      "loss": 0.5373,
      "step": 550
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.844444444444446e-06,
      "loss": 0.5209,
      "step": 575
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.78888888888889e-06,
      "loss": 0.5012,
      "step": 600
    },
    {
      "epoch": 0.1,
      "eval_cer": 17.3679519118689,
      "eval_loss": 0.5292569994926453,
      "eval_runtime": 756.1706,
      "eval_samples_per_second": 3.366,
      "eval_steps_per_second": 0.422,
      "step": 600
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.733333333333334e-06,
      "loss": 0.5047,
      "step": 625
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.677777777777778e-06,
      "loss": 0.5145,
      "step": 650
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.622222222222222e-06,
      "loss": 0.4845,
      "step": 675
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.566666666666668e-06,
      "loss": 0.5201,
      "step": 700
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.511111111111112e-06,
      "loss": 0.5152,
      "step": 725
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.455555555555557e-06,
      "loss": 0.4835,
      "step": 750
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.4e-06,
      "loss": 0.4793,
      "step": 775
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.344444444444446e-06,
      "loss": 0.504,
      "step": 800
    },
    {
      "epoch": 0.13,
      "eval_cer": 17.52908371997349,
      "eval_loss": 0.5074815154075623,
      "eval_runtime": 754.4777,
      "eval_samples_per_second": 3.373,
      "eval_steps_per_second": 0.423,
      "step": 800
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.28888888888889e-06,
      "loss": 0.5113,
      "step": 825
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.233333333333334e-06,
      "loss": 0.4858,
      "step": 850
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.17777777777778e-06,
      "loss": 0.4845,
      "step": 875
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.122222222222223e-06,
      "loss": 0.5021,
      "step": 900
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.066666666666667e-06,
      "loss": 0.4751,
      "step": 925
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.011111111111111e-06,
      "loss": 0.4758,
      "step": 950
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.955555555555555e-06,
      "loss": 0.4813,
      "step": 975
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.4934,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "eval_cer": 15.684638767627362,
      "eval_loss": 0.49266406893730164,
      "eval_runtime": 737.1456,
      "eval_samples_per_second": 3.453,
      "eval_steps_per_second": 0.433,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.844444444444445e-06,
      "loss": 0.4536,
      "step": 1025
    },
    {
      "epoch": 0.17,
      "learning_rate": 8.788888888888891e-06,
      "loss": 0.4474,
      "step": 1050
    },
    {
      "epoch": 0.17,
      "learning_rate": 8.733333333333333e-06,
      "loss": 0.4615,
      "step": 1075
    },
    {
      "epoch": 0.18,
      "learning_rate": 8.677777777777779e-06,
      "loss": 0.4744,
      "step": 1100
    },
    {
      "epoch": 0.18,
      "learning_rate": 8.622222222222223e-06,
      "loss": 0.4405,
      "step": 1125
    },
    {
      "epoch": 0.18,
      "learning_rate": 8.566666666666667e-06,
      "loss": 0.4805,
      "step": 1150
    },
    {
      "epoch": 0.19,
      "learning_rate": 8.511111111111113e-06,
      "loss": 0.4828,
      "step": 1175
    },
    {
      "epoch": 0.19,
      "learning_rate": 8.455555555555555e-06,
      "loss": 0.4958,
      "step": 1200
    },
    {
      "epoch": 0.19,
      "eval_cer": 17.708499988572214,
      "eval_loss": 0.4837556481361389,
      "eval_runtime": 759.8692,
      "eval_samples_per_second": 3.349,
      "eval_steps_per_second": 0.42,
      "step": 1200
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.4797,
      "step": 1225
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.344444444444445e-06,
      "loss": 0.4835,
      "step": 1250
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.288888888888889e-06,
      "loss": 0.4749,
      "step": 1275
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.233333333333335e-06,
      "loss": 0.4627,
      "step": 1300
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.177777777777779e-06,
      "loss": 0.4733,
      "step": 1325
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.122222222222223e-06,
      "loss": 0.432,
      "step": 1350
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.066666666666667e-06,
      "loss": 0.4586,
      "step": 1375
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.011111111111113e-06,
      "loss": 0.4697,
      "step": 1400
    },
    {
      "epoch": 0.22,
      "eval_cer": 18.58843964985258,
      "eval_loss": 0.47684401273727417,
      "eval_runtime": 763.4509,
      "eval_samples_per_second": 3.334,
      "eval_steps_per_second": 0.418,
      "step": 1400
    },
    {
      "epoch": 0.23,
      "learning_rate": 7.955555555555557e-06,
      "loss": 0.441,
      "step": 1425
    },
    {
      "epoch": 0.23,
      "learning_rate": 7.9e-06,
      "loss": 0.4356,
      "step": 1450
    },
    {
      "epoch": 0.24,
      "learning_rate": 7.844444444444446e-06,
      "loss": 0.4899,
      "step": 1475
    },
    {
      "epoch": 0.24,
      "learning_rate": 7.788888888888889e-06,
      "loss": 0.4471,
      "step": 1500
    },
    {
      "epoch": 0.24,
      "learning_rate": 7.733333333333334e-06,
      "loss": 0.4387,
      "step": 1525
    },
    {
      "epoch": 0.25,
      "learning_rate": 7.677777777777778e-06,
      "loss": 0.5033,
      "step": 1550
    },
    {
      "epoch": 0.25,
      "learning_rate": 7.622222222222223e-06,
      "loss": 0.4565,
      "step": 1575
    },
    {
      "epoch": 0.26,
      "learning_rate": 7.566666666666667e-06,
      "loss": 0.4501,
      "step": 1600
    },
    {
      "epoch": 0.26,
      "eval_cer": 16.55086508353713,
      "eval_loss": 0.4688284993171692,
      "eval_runtime": 741.3239,
      "eval_samples_per_second": 3.433,
      "eval_steps_per_second": 0.43,
      "step": 1600
    },
    {
      "epoch": 0.26,
      "learning_rate": 7.511111111111111e-06,
      "loss": 0.4534,
      "step": 1625
    },
    {
      "epoch": 0.26,
      "learning_rate": 7.455555555555556e-06,
      "loss": 0.4376,
      "step": 1650
    },
    {
      "epoch": 0.27,
      "learning_rate": 7.4e-06,
      "loss": 0.4286,
      "step": 1675
    },
    {
      "epoch": 0.27,
      "learning_rate": 7.344444444444445e-06,
      "loss": 0.4511,
      "step": 1700
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.28888888888889e-06,
      "loss": 0.46,
      "step": 1725
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.233333333333334e-06,
      "loss": 0.4669,
      "step": 1750
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.177777777777778e-06,
      "loss": 0.4614,
      "step": 1775
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.122222222222222e-06,
      "loss": 0.4505,
      "step": 1800
    },
    {
      "epoch": 0.29,
      "eval_cer": 17.254816812561423,
      "eval_loss": 0.46170613169670105,
      "eval_runtime": 743.5392,
      "eval_samples_per_second": 3.423,
      "eval_steps_per_second": 0.429,
      "step": 1800
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.066666666666667e-06,
      "loss": 0.4498,
      "step": 1825
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.011111111111112e-06,
      "loss": 0.417,
      "step": 1850
    },
    {
      "epoch": 0.3,
      "learning_rate": 6.955555555555557e-06,
      "loss": 0.4599,
      "step": 1875
    },
    {
      "epoch": 0.3,
      "learning_rate": 6.9e-06,
      "loss": 0.4335,
      "step": 1900
    },
    {
      "epoch": 0.31,
      "learning_rate": 6.844444444444445e-06,
      "loss": 0.4544,
      "step": 1925
    },
    {
      "epoch": 0.31,
      "learning_rate": 6.788888888888889e-06,
      "loss": 0.4567,
      "step": 1950
    },
    {
      "epoch": 0.32,
      "learning_rate": 6.733333333333334e-06,
      "loss": 0.4323,
      "step": 1975
    },
    {
      "epoch": 0.32,
      "learning_rate": 6.677777777777779e-06,
      "loss": 0.4434,
      "step": 2000
    },
    {
      "epoch": 0.32,
      "eval_cer": 15.714351015930337,
      "eval_loss": 0.45656266808509827,
      "eval_runtime": 734.9359,
      "eval_samples_per_second": 3.463,
      "eval_steps_per_second": 0.434,
      "step": 2000
    },
    {
      "epoch": 0.32,
      "learning_rate": 6.6222222222222236e-06,
      "loss": 0.4203,
      "step": 2025
    },
    {
      "epoch": 0.33,
      "learning_rate": 6.566666666666667e-06,
      "loss": 0.4322,
      "step": 2050
    },
    {
      "epoch": 0.33,
      "learning_rate": 6.511111111111112e-06,
      "loss": 0.4352,
      "step": 2075
    },
    {
      "epoch": 0.34,
      "learning_rate": 6.455555555555556e-06,
      "loss": 0.4508,
      "step": 2100
    },
    {
      "epoch": 0.34,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.448,
      "step": 2125
    },
    {
      "epoch": 0.34,
      "learning_rate": 6.3444444444444454e-06,
      "loss": 0.4432,
      "step": 2150
    },
    {
      "epoch": 0.35,
      "learning_rate": 6.28888888888889e-06,
      "loss": 0.4404,
      "step": 2175
    },
    {
      "epoch": 0.35,
      "learning_rate": 6.2333333333333335e-06,
      "loss": 0.4482,
      "step": 2200
    },
    {
      "epoch": 0.35,
      "eval_cer": 16.002331268713004,
      "eval_loss": 0.4514978229999542,
      "eval_runtime": 749.3273,
      "eval_samples_per_second": 3.396,
      "eval_steps_per_second": 0.426,
      "step": 2200
    },
    {
      "epoch": 0.36,
      "learning_rate": 6.177777777777778e-06,
      "loss": 0.4295,
      "step": 2225
    },
    {
      "epoch": 0.36,
      "learning_rate": 6.1222222222222224e-06,
      "loss": 0.4231,
      "step": 2250
    },
    {
      "epoch": 0.36,
      "learning_rate": 6.066666666666667e-06,
      "loss": 0.4215,
      "step": 2275
    },
    {
      "epoch": 0.37,
      "learning_rate": 6.011111111111112e-06,
      "loss": 0.4435,
      "step": 2300
    },
    {
      "epoch": 0.37,
      "learning_rate": 5.955555555555555e-06,
      "loss": 0.4317,
      "step": 2325
    },
    {
      "epoch": 0.38,
      "learning_rate": 5.9e-06,
      "loss": 0.4458,
      "step": 2350
    },
    {
      "epoch": 0.38,
      "learning_rate": 5.844444444444445e-06,
      "loss": 0.3967,
      "step": 2375
    },
    {
      "epoch": 0.38,
      "learning_rate": 5.788888888888889e-06,
      "loss": 0.4333,
      "step": 2400
    },
    {
      "epoch": 0.38,
      "eval_cer": 14.82641190318378,
      "eval_loss": 0.4464516043663025,
      "eval_runtime": 744.46,
      "eval_samples_per_second": 3.419,
      "eval_steps_per_second": 0.428,
      "step": 2400
    },
    {
      "epoch": 0.39,
      "learning_rate": 5.733333333333334e-06,
      "loss": 0.4227,
      "step": 2425
    },
    {
      "epoch": 0.39,
      "learning_rate": 5.677777777777779e-06,
      "loss": 0.4463,
      "step": 2450
    },
    {
      "epoch": 0.4,
      "learning_rate": 5.622222222222222e-06,
      "loss": 0.4242,
      "step": 2475
    },
    {
      "epoch": 0.4,
      "learning_rate": 5.566666666666667e-06,
      "loss": 0.4211,
      "step": 2500
    },
    {
      "epoch": 0.4,
      "learning_rate": 5.511111111111112e-06,
      "loss": 0.417,
      "step": 2525
    },
    {
      "epoch": 0.41,
      "learning_rate": 5.455555555555556e-06,
      "loss": 0.4255,
      "step": 2550
    },
    {
      "epoch": 0.41,
      "learning_rate": 5.400000000000001e-06,
      "loss": 0.4312,
      "step": 2575
    },
    {
      "epoch": 0.42,
      "learning_rate": 5.344444444444446e-06,
      "loss": 0.433,
      "step": 2600
    },
    {
      "epoch": 0.42,
      "eval_cer": 15.294951203346056,
      "eval_loss": 0.4443516135215759,
      "eval_runtime": 749.6689,
      "eval_samples_per_second": 3.395,
      "eval_steps_per_second": 0.426,
      "step": 2600
    },
    {
      "epoch": 0.42,
      "learning_rate": 5.288888888888889e-06,
      "loss": 0.4413,
      "step": 2625
    },
    {
      "epoch": 0.42,
      "learning_rate": 5.233333333333334e-06,
      "loss": 0.4135,
      "step": 2650
    },
    {
      "epoch": 0.43,
      "learning_rate": 5.177777777777779e-06,
      "loss": 0.4225,
      "step": 2675
    },
    {
      "epoch": 0.43,
      "learning_rate": 5.122222222222223e-06,
      "loss": 0.4419,
      "step": 2700
    },
    {
      "epoch": 0.44,
      "learning_rate": 5.0666666666666676e-06,
      "loss": 0.4167,
      "step": 2725
    },
    {
      "epoch": 0.44,
      "learning_rate": 5.011111111111111e-06,
      "loss": 0.4263,
      "step": 2750
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.9555555555555565e-06,
      "loss": 0.4475,
      "step": 2775
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.4027,
      "step": 2800
    },
    {
      "epoch": 0.45,
      "eval_cer": 15.371517381665257,
      "eval_loss": 0.4405870735645294,
      "eval_runtime": 734.4792,
      "eval_samples_per_second": 3.465,
      "eval_steps_per_second": 0.434,
      "step": 2800
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.8444444444444446e-06,
      "loss": 0.4253,
      "step": 2825
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.7888888888888894e-06,
      "loss": 0.3856,
      "step": 2850
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.7333333333333335e-06,
      "loss": 0.4163,
      "step": 2875
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.677777777777778e-06,
      "loss": 0.4451,
      "step": 2900
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.622222222222222e-06,
      "loss": 0.4345,
      "step": 2925
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.566666666666667e-06,
      "loss": 0.396,
      "step": 2950
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.511111111111111e-06,
      "loss": 0.4313,
      "step": 2975
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.455555555555555e-06,
      "loss": 0.4181,
      "step": 3000
    },
    {
      "epoch": 0.48,
      "eval_cer": 14.890407514913262,
      "eval_loss": 0.4375590682029724,
      "eval_runtime": 729.2459,
      "eval_samples_per_second": 3.49,
      "eval_steps_per_second": 0.437,
      "step": 3000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.4e-06,
      "loss": 0.3966,
      "step": 3025
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.344444444444445e-06,
      "loss": 0.4107,
      "step": 3050
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.288888888888889e-06,
      "loss": 0.4266,
      "step": 3075
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.233333333333334e-06,
      "loss": 0.4181,
      "step": 3100
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.177777777777778e-06,
      "loss": 0.4189,
      "step": 3125
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.122222222222222e-06,
      "loss": 0.406,
      "step": 3150
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.066666666666667e-06,
      "loss": 0.4378,
      "step": 3175
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.011111111111111e-06,
      "loss": 0.4209,
      "step": 3200
    },
    {
      "epoch": 0.51,
      "eval_cer": 14.785271867071973,
      "eval_loss": 0.4335635304450989,
      "eval_runtime": 729.2992,
      "eval_samples_per_second": 3.49,
      "eval_steps_per_second": 0.437,
      "step": 3200
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "total_flos": 3.320831803392e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}