Writing logs to ./outputs/2024-02-28-10-23-08-068616/train_log.txt.
Wrote original training args to ./outputs/2024-02-28-10-23-08-068616/training_args.json.
***** Running training *****
  Num examples = 9600
  Num epochs = 5
  Num clean epochs = 5
  Instantaneous batch size per device = 8
  Total train batch size (w. parallel, distributed & accumulation) = 8
  Gradient accumulation steps = 1
  Total optimization steps = 6000
==========================================================
Epoch 1
Running clean epoch 1/5
Train accuracy: 88.40%
Eval accuracy: 91.67%
Best score found. Saved model to ./outputs/2024-02-28-10-23-08-068616/best_model/
==========================================================
Epoch 2
Running clean epoch 2/5
Train accuracy: 94.80%
Eval accuracy: 93.67%
Best score found. Saved model to ./outputs/2024-02-28-10-23-08-068616/best_model/
==========================================================
Epoch 3
Running clean epoch 3/5
Train accuracy: 97.62%
Eval accuracy: 94.67%
Best score found. Saved model to ./outputs/2024-02-28-10-23-08-068616/best_model/
==========================================================
Epoch 4
Running clean epoch 4/5
Train accuracy: 98.95%
Eval accuracy: 94.33%
==========================================================
Epoch 5
Running clean epoch 5/5
Train accuracy: 99.34%
Eval accuracy: 94.58%
Wrote README to ./outputs/2024-02-28-10-23-08-068616/README.md.