trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.450101763010025,
  "best_model_checkpoint": "tiny_bert_km_20_v1_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": 5.4918437004089355,
      "learning_rate": 4.9e-05,
      "loss": 0.4405,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.450101763010025,
      "eval_runtime": 0.2736,
      "eval_samples_per_second": 3187.253,
      "eval_steps_per_second": 14.62,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.100834369659424,
      "learning_rate": 4.8e-05,
      "loss": 0.2442,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7935779816513762,
      "eval_loss": 0.5186178684234619,
      "eval_runtime": 0.2677,
      "eval_samples_per_second": 3257.694,
      "eval_steps_per_second": 14.944,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.591289043426514,
      "learning_rate": 4.7e-05,
      "loss": 0.1889,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7935779816513762,
      "eval_loss": 0.5366886258125305,
      "eval_runtime": 0.2801,
      "eval_samples_per_second": 3112.736,
      "eval_steps_per_second": 14.279,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.194523811340332,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1546,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7821100917431193,
      "eval_loss": 0.5473640561103821,
      "eval_runtime": 0.2739,
      "eval_samples_per_second": 3183.61,
      "eval_steps_per_second": 14.604,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.071108818054199,
      "learning_rate": 4.5e-05,
      "loss": 0.1278,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7786697247706422,
      "eval_loss": 0.7206570506095886,
      "eval_runtime": 0.2738,
      "eval_samples_per_second": 3184.877,
      "eval_steps_per_second": 14.61,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.966401100158691,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1066,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7740825688073395,
      "eval_loss": 0.7687958478927612,
      "eval_runtime": 0.2656,
      "eval_samples_per_second": 3283.047,
      "eval_steps_per_second": 15.06,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.21043807328349412,
      "train_runtime": 222.7353,
      "train_samples_per_second": 15118.615,
      "train_steps_per_second": 59.263
    }
  ],
  "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
}