Date:  11-08, 15:05
Batch Size: 64
Eval Strategy: no
Early Stopping: steps
Early Stopping Patience: 5
Best Model Metric: None
{'eval_loss': 0.3047332465648651, 'eval_accuracy': 0.8676975945017182, 'eval_f1': 0.8651488616462346, 'eval_runtime': 9.0782, 'eval_samples_per_second': 64.109, 'eval_steps_per_second': 1.102, 'epoch': 20.19}