trainer_state.json

{
  "best_metric": 0.6198700070381165,
  "best_model_checkpoint": "bert_base_lda_20_v1_cola/checkpoint-68",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 238,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.7223057150840759,
      "learning_rate": 4.9e-05,
      "loss": 0.6136,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6265467405319214,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7187,
      "eval_samples_per_second": 1451.19,
      "eval_steps_per_second": 6.957,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.90400230884552,
      "learning_rate": 4.8e-05,
      "loss": 0.6112,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6198700070381165,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.684,
      "eval_samples_per_second": 1524.743,
      "eval_steps_per_second": 7.309,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.334337592124939,
      "learning_rate": 4.7e-05,
      "loss": 0.5873,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.641418993473053,
      "eval_loss": 0.6348920464515686,
      "eval_matthews_correlation": 0.06986894691574659,
      "eval_runtime": 0.6727,
      "eval_samples_per_second": 1550.524,
      "eval_steps_per_second": 7.433,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9891555309295654,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5381,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5963566899299622,
      "eval_loss": 0.6501876711845398,
      "eval_matthews_correlation": 0.10535705260984664,
      "eval_runtime": 0.682,
      "eval_samples_per_second": 1529.263,
      "eval_steps_per_second": 7.331,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.578742027282715,
      "learning_rate": 4.5e-05,
      "loss": 0.4951,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6442952752113342,
      "eval_loss": 0.6613349318504333,
      "eval_matthews_correlation": 0.07683234381179825,
      "eval_runtime": 0.6814,
      "eval_samples_per_second": 1530.617,
      "eval_steps_per_second": 7.338,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.737532615661621,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4506,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5810163021087646,
      "eval_loss": 0.7528904676437378,
      "eval_matthews_correlation": 0.12820404799222795,
      "eval_runtime": 0.6701,
      "eval_samples_per_second": 1556.391,
      "eval_steps_per_second": 7.461,
      "step": 204
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.000977516174316,
      "learning_rate": 4.3e-05,
      "loss": 0.4248,
      "step": 238
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6395014524459839,
      "eval_loss": 0.7049092650413513,
      "eval_matthews_correlation": 0.053910422269610486,
      "eval_runtime": 0.6819,
      "eval_samples_per_second": 1529.505,
      "eval_steps_per_second": 7.332,
      "step": 238
    },
    {
      "epoch": 7.0,
      "step": 238,
      "total_flos": 7874519220341760.0,
      "train_loss": 0.5315281884009097,
      "train_runtime": 120.102,
      "train_samples_per_second": 3559.89,
      "train_steps_per_second": 14.155
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 7874519220341760.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}