trainer_state.json

{
  "best_metric": 0.7066460847854614,
  "best_model_checkpoint": "bert_base_lda_20_v1_book_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.5193098783493042,
      "learning_rate": 4.9e-05,
      "loss": 0.701,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4225352112676056,
      "eval_loss": 0.7117077708244324,
      "eval_runtime": 0.0538,
      "eval_samples_per_second": 1318.479,
      "eval_steps_per_second": 18.57,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.910641670227051,
      "learning_rate": 4.8e-05,
      "loss": 0.6963,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4647887323943662,
      "eval_loss": 0.7066460847854614,
      "eval_runtime": 0.0621,
      "eval_samples_per_second": 1143.389,
      "eval_steps_per_second": 16.104,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.596242904663086,
      "learning_rate": 4.7e-05,
      "loss": 0.6902,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.43661971830985913,
      "eval_loss": 0.7150639295578003,
      "eval_runtime": 0.0538,
      "eval_samples_per_second": 1318.969,
      "eval_steps_per_second": 18.577,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7775526642799377,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6908,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3380281690140845,
      "eval_loss": 0.73046875,
      "eval_runtime": 0.0552,
      "eval_samples_per_second": 1287.208,
      "eval_steps_per_second": 18.13,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1994614601135254,
      "learning_rate": 4.5e-05,
      "loss": 0.6894,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.23943661971830985,
      "eval_loss": 0.736135721206665,
      "eval_runtime": 0.0539,
      "eval_samples_per_second": 1316.433,
      "eval_steps_per_second": 18.541,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7005854845046997,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.689,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.19718309859154928,
      "eval_loss": 0.7537961006164551,
      "eval_runtime": 0.0537,
      "eval_samples_per_second": 1322.184,
      "eval_steps_per_second": 18.622,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.6941657066345215,
      "learning_rate": 4.3e-05,
      "loss": 0.6867,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.19718309859154928,
      "eval_loss": 0.7771235704421997,
      "eval_runtime": 0.0538,
      "eval_samples_per_second": 1320.455,
      "eval_steps_per_second": 18.598,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 584764320537600.0,
      "train_loss": 0.6919111751374745,
      "train_runtime": 23.6666,
      "train_samples_per_second": 1341.553,
      "train_steps_per_second": 6.338
    }
  ],
  "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": 584764320537600.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}