Date:  01_20, 22:30
Batch Size: 32
Eval Strategy: epoch
Early Stopping: epoch
Early Stopping Patience: 5
Best Model Metric: loss
{'eval_loss': 0.19123336672782898, 'eval_accuracy': 0.9474048442906574, 'eval_f1': 0.9477535633238166, 'eval_runtime': 295.531, 'eval_samples_per_second': 4.89, 'eval_steps_per_second': 0.156, 'epoch': 8.0}