trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.687006413936615,
  "best_model_checkpoint": "tiny_bert_rand_100_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.41298094391822815,
      "learning_rate": 4.9e-05,
      "loss": 0.7006,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5523465703971119,
      "eval_loss": 0.6890371441841125,
      "eval_runtime": 0.0992,
      "eval_samples_per_second": 2791.916,
      "eval_steps_per_second": 20.158,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.400938481092453,
      "learning_rate": 4.8e-05,
      "loss": 0.6901,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5523465703971119,
      "eval_loss": 0.687006413936615,
      "eval_runtime": 0.1033,
      "eval_samples_per_second": 2682.189,
      "eval_steps_per_second": 19.366,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5469049215316772,
      "learning_rate": 4.7e-05,
      "loss": 0.6767,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5631768953068592,
      "eval_loss": 0.6920408606529236,
      "eval_runtime": 0.1118,
      "eval_samples_per_second": 2477.751,
      "eval_steps_per_second": 17.89,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9113770127296448,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6364,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7453216910362244,
      "eval_runtime": 0.124,
      "eval_samples_per_second": 2234.591,
      "eval_steps_per_second": 16.134,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.277000069618225,
      "learning_rate": 4.5e-05,
      "loss": 0.5824,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.8142063617706299,
      "eval_runtime": 0.1129,
      "eval_samples_per_second": 2452.939,
      "eval_steps_per_second": 17.711,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.0929064750671387,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4991,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.9067699313163757,
      "eval_runtime": 0.104,
      "eval_samples_per_second": 2663.991,
      "eval_steps_per_second": 19.235,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.7687212228775024,
      "learning_rate": 4.3e-05,
      "loss": 0.3752,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 1.1385396718978882,
      "eval_runtime": 0.0997,
      "eval_samples_per_second": 2779.093,
      "eval_steps_per_second": 20.066,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 457076966492160.0,
      "train_loss": 0.5943562643868583,
      "train_runtime": 16.3443,
      "train_samples_per_second": 7617.351,
      "train_steps_per_second": 30.592
    }
  ],
  "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": 457076966492160.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}