Writing logs to ./outputs/2024-04-29-22-36-28-742223/train_log.txt.
Wrote original training args to ./outputs/2024-04-29-22-36-28-742223/training_args.json.
***** Running training *****
  Num examples = 4800
  Num epochs = 3
  Num clean epochs = 3
  Instantaneous batch size per device = 8
  Total train batch size (w. parallel, distributed & accumulation) = 8
  Gradient accumulation steps = 1
  Total optimization steps = 1800
==========================================================
Epoch 1
Running clean epoch 1/3
Train accuracy: 88.60%
Eval accuracy: 94.27%
Best score found. Saved model to ./outputs/2024-04-29-22-36-28-742223/best_model/
==========================================================
Epoch 2
Running clean epoch 2/3
Train accuracy: 95.29%
Eval accuracy: 95.37%
Best score found. Saved model to ./outputs/2024-04-29-22-36-28-742223/best_model/
==========================================================
Epoch 3
Running clean epoch 3/3
Train accuracy: 98.08%
Eval accuracy: 95.27%
Wrote README to ./outputs/2024-04-29-22-36-28-742223/README.md.