run-22/checkpoint-1054/trainer_state.json

{
  "best_metric": 0.9048165137614679,
  "best_model_checkpoint": "electra-distilled-sst\\run-22\\checkpoint-1054",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1054,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 13.01644229888916,
      "learning_rate": 2.7002954057123424e-05,
      "loss": 0.6671,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9036697247706422,
      "eval_loss": 0.428321897983551,
      "eval_runtime": 1.451,
      "eval_samples_per_second": 600.984,
      "eval_steps_per_second": 4.824,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 25.425676345825195,
      "learning_rate": 2.4002625828554156e-05,
      "loss": 0.3779,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9048165137614679,
      "eval_loss": 0.40753328800201416,
      "eval_runtime": 1.4531,
      "eval_samples_per_second": 600.091,
      "eval_steps_per_second": 4.817,
      "step": 1054
    }
  ],
  "logging_steps": 500,
  "max_steps": 5270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 373776249469380.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9315044934322999,
    "learning_rate": 3.0003282285692692e-05,
    "num_train_epochs": 10,
    "temperature": 20
  }
}