{'loss': 2.4436, 'learning_rate': 8.535533905932739e-06, 'epoch': 2.0}
{'loss': 1.7417, 'learning_rate': 5e-06, 'epoch': 2.67}          
{'loss': 0.6206, 'learning_rate': 1.4644660940672628e-06, 'epoch': 3.0}
{'loss': 2.3251, 'learning_rate': 0.0, 'epoch': 4.0}             
{'train_runtime': 642.8468, 'train_samples_per_second': 0.023, 'train_steps_per_second': 0.008, 'train_loss': 1.9149054765701294, 'epoch': 4.0}