trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6892063021659851,
  "best_model_checkpoint": "tiny_bert_km_20_v2_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.5985978841781616,
      "learning_rate": 4.9e-05,
      "loss": 0.6958,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.6892063021659851,
      "eval_runtime": 0.0927,
      "eval_samples_per_second": 2987.434,
      "eval_steps_per_second": 21.57,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7859639525413513,
      "learning_rate": 4.8e-05,
      "loss": 0.6782,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 0.704993486404419,
      "eval_runtime": 0.0937,
      "eval_samples_per_second": 2956.622,
      "eval_steps_per_second": 21.347,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6229118704795837,
      "learning_rate": 4.7e-05,
      "loss": 0.66,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.7001389265060425,
      "eval_runtime": 0.0957,
      "eval_samples_per_second": 2895.464,
      "eval_steps_per_second": 20.906,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6870622038841248,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6341,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.7196595668792725,
      "eval_runtime": 0.1139,
      "eval_samples_per_second": 2430.921,
      "eval_steps_per_second": 17.552,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9636316299438477,
      "learning_rate": 4.5e-05,
      "loss": 0.6062,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.7338955402374268,
      "eval_runtime": 0.0955,
      "eval_samples_per_second": 2900.046,
      "eval_steps_per_second": 20.939,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9693595767021179,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5654,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.7659070491790771,
      "eval_runtime": 0.0922,
      "eval_samples_per_second": 3004.438,
      "eval_steps_per_second": 21.693,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 391780256993280.0,
      "train_loss": 0.6399651845296224,
      "train_runtime": 12.7067,
      "train_samples_per_second": 9797.982,
      "train_steps_per_second": 39.349
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 391780256993280.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}