trainer_state.json

{
  "best_metric": 1.0962358713150024,
  "best_model_checkpoint": "bert_tiny_lda_20_v1_mnli/checkpoint-3068",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 10738,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.02029508352279663,
      "learning_rate": 0.00098,
      "loss": 1.0993,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.0992883443832397,
      "eval_runtime": 2.9964,
      "eval_samples_per_second": 3275.549,
      "eval_steps_per_second": 13.015,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.11836767196655273,
      "learning_rate": 0.00096,
      "loss": 1.0986,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.0962358713150024,
      "eval_runtime": 3.0153,
      "eval_samples_per_second": 3255.045,
      "eval_steps_per_second": 12.934,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.03678739443421364,
      "learning_rate": 0.00094,
      "loss": 1.0987,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.0964475870132446,
      "eval_runtime": 3.0346,
      "eval_samples_per_second": 3234.397,
      "eval_steps_per_second": 12.852,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.062206000089645386,
      "learning_rate": 0.00092,
      "loss": 1.0985,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.10163414478302,
      "eval_runtime": 3.0235,
      "eval_samples_per_second": 3246.215,
      "eval_steps_per_second": 12.899,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.07343439757823944,
      "learning_rate": 0.0009000000000000001,
      "loss": 1.0985,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.0988649129867554,
      "eval_runtime": 3.0456,
      "eval_samples_per_second": 3222.725,
      "eval_steps_per_second": 12.806,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.047856856137514114,
      "learning_rate": 0.00088,
      "loss": 1.0986,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.0988649129867554,
      "eval_runtime": 3.0238,
      "eval_samples_per_second": 3245.866,
      "eval_steps_per_second": 12.897,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0775880515575409,
      "learning_rate": 0.00086,
      "loss": 1.0986,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.0992883443832397,
      "eval_runtime": 3.0565,
      "eval_samples_per_second": 3211.238,
      "eval_steps_per_second": 12.76,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "step": 10738,
      "total_flos": 7.208852705843712e+16,
      "train_loss": 1.0986731009324362,
      "train_runtime": 1497.1474,
      "train_samples_per_second": 13115.008,
      "train_steps_per_second": 51.231
    }
  ],
  "logging_steps": 1,
  "max_steps": 76700,
  "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": 7.208852705843712e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}