Model save

README.md

@@ -34,15 +34,15 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 5e-05
-- train_batch_size: 32
+- train_batch_size: 64
 - eval_batch_size: 64
 - seed: 42
-- gradient_accumulation_steps: 8
+- gradient_accumulation_steps: 4
 - total_train_batch_size: 256
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_steps: 500
-- num_epochs: 3
+- num_epochs: 20
 
 ### Training results