trainer_state.json

{
  "best_metric": 0.6925343871116638,
  "best_model_checkpoint": "bert_tiny_lda_100_v1_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": 1.2701410055160522,
      "learning_rate": 4.9e-05,
      "loss": 0.7159,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.7091183662414551,
      "eval_runtime": 0.0972,
      "eval_samples_per_second": 2849.909,
      "eval_steps_per_second": 20.577,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6685553789138794,
      "learning_rate": 4.8e-05,
      "loss": 0.6926,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 0.6925343871116638,
      "eval_runtime": 0.1007,
      "eval_samples_per_second": 2750.35,
      "eval_steps_per_second": 19.858,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5383151769638062,
      "learning_rate": 4.7e-05,
      "loss": 0.6814,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.6944310665130615,
      "eval_runtime": 0.0969,
      "eval_samples_per_second": 2858.407,
      "eval_steps_per_second": 20.638,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6841632723808289,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6663,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.697773277759552,
      "eval_runtime": 0.0972,
      "eval_samples_per_second": 2850.811,
      "eval_steps_per_second": 20.583,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.6859617233276367,
      "learning_rate": 4.5e-05,
      "loss": 0.6472,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.7425259351730347,
      "eval_runtime": 0.0971,
      "eval_samples_per_second": 2851.944,
      "eval_steps_per_second": 20.592,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.9587653875350952,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6276,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7315176129341125,
      "eval_runtime": 0.098,
      "eval_samples_per_second": 2825.759,
      "eval_steps_per_second": 20.403,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.4208569526672363,
      "learning_rate": 4.3e-05,
      "loss": 0.5534,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5018050541516246,
      "eval_loss": 0.816455602645874,
      "eval_runtime": 0.1092,
      "eval_samples_per_second": 2537.616,
      "eval_steps_per_second": 18.322,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 457076966492160.0,
      "train_loss": 0.6549058369227818,
      "train_runtime": 15.9605,
      "train_samples_per_second": 7800.521,
      "train_steps_per_second": 31.327
    }
  ],
  "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
}