trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.622419536113739,
  "best_model_checkpoint": "tiny_bert_km_10_v2_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6918684244155884,
      "learning_rate": 4.9e-05,
      "loss": 0.6217,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.622419536113739,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3489,
      "eval_samples_per_second": 2989.274,
      "eval_steps_per_second": 14.33,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9699868559837341,
      "learning_rate": 4.8e-05,
      "loss": 0.6029,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6245683431625366,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.361,
      "eval_samples_per_second": 2889.026,
      "eval_steps_per_second": 13.85,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.235334038734436,
      "learning_rate": 4.7e-05,
      "loss": 0.5867,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6864812970161438,
      "eval_loss": 0.6235543489456177,
      "eval_matthews_correlation": 0.006434621036303265,
      "eval_runtime": 0.33,
      "eval_samples_per_second": 3160.216,
      "eval_steps_per_second": 15.15,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.5698190927505493,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5567,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6634707450866699,
      "eval_loss": 0.642329752445221,
      "eval_matthews_correlation": 0.053932562872819126,
      "eval_runtime": 0.3392,
      "eval_samples_per_second": 3074.445,
      "eval_steps_per_second": 14.738,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.029982089996338,
      "learning_rate": 4.5e-05,
      "loss": 0.5108,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.665388286113739,
      "eval_loss": 0.6568330526351929,
      "eval_matthews_correlation": 0.07044477018230917,
      "eval_runtime": 0.3458,
      "eval_samples_per_second": 3015.902,
      "eval_steps_per_second": 14.458,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.075061798095703,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4644,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6558005809783936,
      "eval_loss": 0.7074902653694153,
      "eval_matthews_correlation": 0.07964656167895688,
      "eval_runtime": 0.3314,
      "eval_samples_per_second": 3147.486,
      "eval_steps_per_second": 15.089,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.557197608199774,
      "train_runtime": 33.5799,
      "train_samples_per_second": 12732.306,
      "train_steps_per_second": 50.625
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 1345426898614272.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}