trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0547945205479454,
  "eval_steps": 30,
  "global_step": 150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.410958904109589,
      "grad_norm": 10.404386520385742,
      "learning_rate": 9.666666666666667e-06,
      "loss": 3.5162,
      "step": 30
    },
    {
      "epoch": 0.410958904109589,
      "eval_loss": 2.604057788848877,
      "eval_runtime": 186.0476,
      "eval_samples_per_second": 6.262,
      "eval_steps_per_second": 0.392,
      "eval_wer": 0.9534165328061287,
      "step": 30
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 6.850046634674072,
      "learning_rate": 7.583333333333333e-06,
      "loss": 2.1333,
      "step": 60
    },
    {
      "epoch": 0.821917808219178,
      "eval_loss": 1.7795313596725464,
      "eval_runtime": 191.9598,
      "eval_samples_per_second": 6.069,
      "eval_steps_per_second": 0.38,
      "eval_wer": 0.7878413443716792,
      "step": 60
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 6.052253723144531,
      "learning_rate": 5.0833333333333335e-06,
      "loss": 1.6574,
      "step": 90
    },
    {
      "epoch": 1.2328767123287672,
      "eval_loss": 1.5370458364486694,
      "eval_runtime": 189.5245,
      "eval_samples_per_second": 6.147,
      "eval_steps_per_second": 0.385,
      "eval_wer": 0.7172865439268503,
      "step": 90
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 5.911397457122803,
      "learning_rate": 2.5833333333333337e-06,
      "loss": 1.4484,
      "step": 120
    },
    {
      "epoch": 1.643835616438356,
      "eval_loss": 1.435214877128601,
      "eval_runtime": 189.6046,
      "eval_samples_per_second": 6.144,
      "eval_steps_per_second": 0.385,
      "eval_wer": 0.6735450389225256,
      "step": 120
    },
    {
      "epoch": 2.0547945205479454,
      "grad_norm": 5.800022602081299,
      "learning_rate": 8.333333333333334e-08,
      "loss": 1.3908,
      "step": 150
    },
    {
      "epoch": 2.0547945205479454,
      "eval_loss": 1.4041714668273926,
      "eval_runtime": 190.107,
      "eval_samples_per_second": 6.128,
      "eval_steps_per_second": 0.384,
      "eval_wer": 0.667737550969974,
      "step": 150
    },
    {
      "epoch": 2.0547945205479454,
      "step": 150,
      "total_flos": 4.7164860112896e+17,
      "train_loss": 2.0292181396484374,
      "train_runtime": 3829.3257,
      "train_samples_per_second": 5.014,
      "train_steps_per_second": 0.039
    }
  ],
  "logging_steps": 30,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.7164860112896e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}