Update README.md

README.md

@@ -44,8 +44,9 @@ print(tokenizer.decode(outputs[0], skip_special_tokens=True))
 ## Training Details
 
  - Trained for one epoch on SlimOrca-Dedup
- - Learning rate: 1e-5
+ - Learning rate: 2e-5
+ - Cosine learning rate decay to 0
  - Optimizer: AdamW
- - Effective batch size: 64
- - Gradient accumulation steps (mini batch size): 16 (4)
+ - Effective batch size: 256
+ - Gradient accumulation steps (mini batch size): 64 (4)
  - Trained with FP32