trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 200,
  "global_step": 5010,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3992015968063872,
      "eval_loss": 1.943410873413086,
      "eval_runtime": 41.2309,
      "eval_samples_per_second": 11.52,
      "eval_steps_per_second": 1.455,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 0.7984031936127745,
      "eval_loss": 0.3648405969142914,
      "eval_runtime": 41.3464,
      "eval_samples_per_second": 11.488,
      "eval_steps_per_second": 1.451,
      "eval_wer": 0.5411255411255411,
      "step": 400
    },
    {
      "epoch": 0.998003992015968,
      "grad_norm": 0.7569882273674011,
      "learning_rate": 0.00027568228105906314,
      "loss": 2.5692,
      "step": 500
    },
    {
      "epoch": 1.1976047904191618,
      "eval_loss": 0.3371403217315674,
      "eval_runtime": 40.9227,
      "eval_samples_per_second": 11.607,
      "eval_steps_per_second": 1.466,
      "eval_wer": 0.5174433409727527,
      "step": 600
    },
    {
      "epoch": 1.596806387225549,
      "eval_loss": 0.32294145226478577,
      "eval_runtime": 41.2686,
      "eval_samples_per_second": 11.51,
      "eval_steps_per_second": 1.454,
      "eval_wer": 0.5212630506748154,
      "step": 800
    },
    {
      "epoch": 1.996007984031936,
      "grad_norm": 0.6301750540733337,
      "learning_rate": 0.000245132382892057,
      "loss": 0.3941,
      "step": 1000
    },
    {
      "epoch": 1.996007984031936,
      "eval_loss": 0.31834524869918823,
      "eval_runtime": 40.8769,
      "eval_samples_per_second": 11.62,
      "eval_steps_per_second": 1.468,
      "eval_wer": 0.49146931499872676,
      "step": 1000
    },
    {
      "epoch": 2.3952095808383236,
      "eval_loss": 0.3067522943019867,
      "eval_runtime": 40.1904,
      "eval_samples_per_second": 11.819,
      "eval_steps_per_second": 1.493,
      "eval_wer": 0.5072574484339191,
      "step": 1200
    },
    {
      "epoch": 2.7944111776447107,
      "eval_loss": 0.30573877692222595,
      "eval_runtime": 40.3076,
      "eval_samples_per_second": 11.784,
      "eval_steps_per_second": 1.489,
      "eval_wer": 0.46880570409982175,
      "step": 1400
    },
    {
      "epoch": 2.9940119760479043,
      "grad_norm": 0.6246519088745117,
      "learning_rate": 0.0002145824847250509,
      "loss": 0.3502,
      "step": 1500
    },
    {
      "epoch": 3.193612774451098,
      "eval_loss": 0.30173251032829285,
      "eval_runtime": 40.844,
      "eval_samples_per_second": 11.63,
      "eval_steps_per_second": 1.469,
      "eval_wer": 0.47771836007130125,
      "step": 1600
    },
    {
      "epoch": 3.592814371257485,
      "eval_loss": 0.2904900014400482,
      "eval_runtime": 40.4163,
      "eval_samples_per_second": 11.753,
      "eval_steps_per_second": 1.485,
      "eval_wer": 0.46473134708428826,
      "step": 1800
    },
    {
      "epoch": 3.992015968063872,
      "grad_norm": 0.9133301973342896,
      "learning_rate": 0.00018403258655804477,
      "loss": 0.3253,
      "step": 2000
    },
    {
      "epoch": 3.992015968063872,
      "eval_loss": 0.2856718897819519,
      "eval_runtime": 40.3664,
      "eval_samples_per_second": 11.767,
      "eval_steps_per_second": 1.486,
      "eval_wer": 0.4685510567863509,
      "step": 2000
    },
    {
      "epoch": 4.391217564870259,
      "eval_loss": 0.28921443223953247,
      "eval_runtime": 40.4989,
      "eval_samples_per_second": 11.729,
      "eval_steps_per_second": 1.482,
      "eval_wer": 0.4601476954418131,
      "step": 2200
    },
    {
      "epoch": 4.790419161676647,
      "eval_loss": 0.28484851121902466,
      "eval_runtime": 40.2649,
      "eval_samples_per_second": 11.797,
      "eval_steps_per_second": 1.49,
      "eval_wer": 0.47593582887700536,
      "step": 2400
    },
    {
      "epoch": 4.99001996007984,
      "grad_norm": 0.7654047012329102,
      "learning_rate": 0.00015348268839103868,
      "loss": 0.3066,
      "step": 2500
    },
    {
      "epoch": 5.189620758483034,
      "eval_loss": 0.2800922989845276,
      "eval_runtime": 40.3508,
      "eval_samples_per_second": 11.772,
      "eval_steps_per_second": 1.487,
      "eval_wer": 0.44435956200662086,
      "step": 2600
    },
    {
      "epoch": 5.588822355289421,
      "eval_loss": 0.27517372369766235,
      "eval_runtime": 40.2103,
      "eval_samples_per_second": 11.813,
      "eval_steps_per_second": 1.492,
      "eval_wer": 0.4626941685765215,
      "step": 2800
    },
    {
      "epoch": 5.9880239520958085,
      "grad_norm": 0.9405556917190552,
      "learning_rate": 0.00012293279022403258,
      "loss": 0.2988,
      "step": 3000
    },
    {
      "epoch": 5.9880239520958085,
      "eval_loss": 0.28181877732276917,
      "eval_runtime": 40.5684,
      "eval_samples_per_second": 11.709,
      "eval_steps_per_second": 1.479,
      "eval_wer": 0.4614209320091673,
      "step": 3000
    },
    {
      "epoch": 6.387225548902196,
      "eval_loss": 0.27585282921791077,
      "eval_runtime": 41.0192,
      "eval_samples_per_second": 11.58,
      "eval_steps_per_second": 1.463,
      "eval_wer": 0.44435956200662086,
      "step": 3200
    },
    {
      "epoch": 6.786427145708583,
      "eval_loss": 0.27509135007858276,
      "eval_runtime": 40.7557,
      "eval_samples_per_second": 11.655,
      "eval_steps_per_second": 1.472,
      "eval_wer": 0.4382480264833206,
      "step": 3400
    },
    {
      "epoch": 6.986027944111776,
      "grad_norm": 0.8755282163619995,
      "learning_rate": 9.238289205702647e-05,
      "loss": 0.2877,
      "step": 3500
    },
    {
      "epoch": 7.18562874251497,
      "eval_loss": 0.2725882828235626,
      "eval_runtime": 40.5214,
      "eval_samples_per_second": 11.722,
      "eval_steps_per_second": 1.481,
      "eval_wer": 0.4471606824548001,
      "step": 3600
    },
    {
      "epoch": 7.584830339321357,
      "eval_loss": 0.27224990725517273,
      "eval_runtime": 40.7203,
      "eval_samples_per_second": 11.665,
      "eval_steps_per_second": 1.473,
      "eval_wer": 0.4484339190221543,
      "step": 3800
    },
    {
      "epoch": 7.984031936127744,
      "grad_norm": 0.6233875155448914,
      "learning_rate": 6.183299389002036e-05,
      "loss": 0.2812,
      "step": 4000
    },
    {
      "epoch": 7.984031936127744,
      "eval_loss": 0.2709678113460541,
      "eval_runtime": 40.5287,
      "eval_samples_per_second": 11.72,
      "eval_steps_per_second": 1.48,
      "eval_wer": 0.434428316781258,
      "step": 4000
    },
    {
      "epoch": 8.383233532934131,
      "eval_loss": 0.2734103798866272,
      "eval_runtime": 40.5,
      "eval_samples_per_second": 11.728,
      "eval_steps_per_second": 1.481,
      "eval_wer": 0.4410491469314999,
      "step": 4200
    },
    {
      "epoch": 8.782435129740518,
      "eval_loss": 0.27336403727531433,
      "eval_runtime": 40.7879,
      "eval_samples_per_second": 11.646,
      "eval_steps_per_second": 1.471,
      "eval_wer": 0.435956200662083,
      "step": 4400
    },
    {
      "epoch": 8.982035928143713,
      "grad_norm": 0.49290069937705994,
      "learning_rate": 3.128309572301426e-05,
      "loss": 0.2742,
      "step": 4500
    },
    {
      "epoch": 9.181636726546905,
      "eval_loss": 0.2758755087852478,
      "eval_runtime": 40.5476,
      "eval_samples_per_second": 11.715,
      "eval_steps_per_second": 1.48,
      "eval_wer": 0.43977591036414565,
      "step": 4600
    },
    {
      "epoch": 9.580838323353294,
      "eval_loss": 0.27402016520500183,
      "eval_runtime": 40.6331,
      "eval_samples_per_second": 11.69,
      "eval_steps_per_second": 1.477,
      "eval_wer": 0.4336643748408454,
      "step": 4800
    },
    {
      "epoch": 9.980039920159681,
      "grad_norm": 0.8148034811019897,
      "learning_rate": 7.331975560081466e-07,
      "loss": 0.2731,
      "step": 5000
    },
    {
      "epoch": 9.980039920159681,
      "eval_loss": 0.27219507098197937,
      "eval_runtime": 40.3217,
      "eval_samples_per_second": 11.78,
      "eval_steps_per_second": 1.488,
      "eval_wer": 0.4382480264833206,
      "step": 5000
    },
    {
      "epoch": 10.0,
      "step": 5010,
      "total_flos": 7.851078607918333e+18,
      "train_loss": 0.535597919037718,
      "train_runtime": 6681.5115,
      "train_samples_per_second": 5.997,
      "train_steps_per_second": 0.75
    }
  ],
  "logging_steps": 500,
  "max_steps": 5010,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.851078607918333e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}