trainer_state.json

{
  "best_metric": 0.6929468512535095,
  "best_model_checkpoint": "bert_tiny_lda_100_v1_book_wnli/checkpoint-3",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.5578425526618958,
      "learning_rate": 4.9e-05,
      "loss": 0.7022,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.6929468512535095,
      "eval_runtime": 0.0284,
      "eval_samples_per_second": 2495.689,
      "eval_steps_per_second": 35.151,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.49783381819725037,
      "learning_rate": 4.8e-05,
      "loss": 0.6965,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4084507042253521,
      "eval_loss": 0.7018594145774841,
      "eval_runtime": 0.0308,
      "eval_samples_per_second": 2308.654,
      "eval_steps_per_second": 32.516,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5858719348907471,
      "learning_rate": 4.7e-05,
      "loss": 0.6972,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4788732394366197,
      "eval_loss": 0.706371009349823,
      "eval_runtime": 0.031,
      "eval_samples_per_second": 2289.221,
      "eval_steps_per_second": 32.243,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7486657500267029,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6917,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4788732394366197,
      "eval_loss": 0.7016945481300354,
      "eval_runtime": 0.0398,
      "eval_samples_per_second": 1785.142,
      "eval_steps_per_second": 25.143,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8318929076194763,
      "learning_rate": 4.5e-05,
      "loss": 0.6919,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.36619718309859156,
      "eval_loss": 0.7100571393966675,
      "eval_runtime": 0.0309,
      "eval_samples_per_second": 2296.211,
      "eval_steps_per_second": 32.341,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8059946894645691,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6925,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.323943661971831,
      "eval_loss": 0.7246920466423035,
      "eval_runtime": 0.0306,
      "eval_samples_per_second": 2321.577,
      "eval_steps_per_second": 32.698,
      "step": 18
    },
    {
      "epoch": 6.0,
      "step": 18,
      "total_flos": 99911832606720.0,
      "train_loss": 0.695326010386149,
      "train_runtime": 8.2224,
      "train_samples_per_second": 3861.423,
      "train_steps_per_second": 18.243
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "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": 99911832606720.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}