checkpoint-312/trainer_state.json

{
  "best_metric": 0.6127167630057804,
  "best_model_checkpoint": "/home/ubuntu/utah/babylm-24/src/evaluation/results/finetune/DebertaV2-Base-10M_babylm-A/rte/checkpoint-312",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5179855823516846,
      "eval_f1": 0.6127167630057804,
      "eval_loss": 0.6988222599029541,
      "eval_mcc": 0.09869207097117276,
      "eval_runtime": 0.2603,
      "eval_samples_per_second": 534.027,
      "eval_steps_per_second": 69.155,
      "step": 312
    }
  ],
  "logging_steps": 500,
  "max_steps": 6240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.001
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 146904368855040.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}