trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.46620461344718933,
  "best_model_checkpoint": "tiny_bert_km_100_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 6.078668594360352,
      "learning_rate": 4.9e-05,
      "loss": 0.458,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.46620461344718933,
      "eval_runtime": 0.2849,
      "eval_samples_per_second": 3061.241,
      "eval_steps_per_second": 14.042,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.059361934661865,
      "learning_rate": 4.8e-05,
      "loss": 0.2519,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.4957926571369171,
      "eval_runtime": 0.2588,
      "eval_samples_per_second": 3370.039,
      "eval_steps_per_second": 15.459,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.3963823318481445,
      "learning_rate": 4.7e-05,
      "loss": 0.193,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8073394495412844,
      "eval_loss": 0.5082025527954102,
      "eval_runtime": 0.2587,
      "eval_samples_per_second": 3370.369,
      "eval_steps_per_second": 15.46,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.676812171936035,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1579,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5986669659614563,
      "eval_runtime": 0.2658,
      "eval_samples_per_second": 3280.88,
      "eval_steps_per_second": 15.05,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.239998817443848,
      "learning_rate": 4.5e-05,
      "loss": 0.1297,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.6561070680618286,
      "eval_runtime": 0.2612,
      "eval_samples_per_second": 3338.466,
      "eval_steps_per_second": 15.314,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.036471366882324,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1087,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.786697247706422,
      "eval_loss": 0.7549962997436523,
      "eval_runtime": 0.2581,
      "eval_samples_per_second": 3378.794,
      "eval_steps_per_second": 15.499,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.21653618354990026,
      "train_runtime": 217.2953,
      "train_samples_per_second": 15497.11,
      "train_steps_per_second": 60.747
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}