Update README.md
Browse files
README.md
CHANGED
|
@@ -51,6 +51,8 @@ The following hyperparameters were used during training:
|
|
| 51 |
- lr_scheduler_type: linear
|
| 52 |
- lr_scheduler_warmup_steps: 200
|
| 53 |
- num_epochs: 8
|
|
|
|
|
|
|
| 54 |
|
| 55 |
### Training results
|
| 56 |
|
|
|
|
| 51 |
- lr_scheduler_type: linear
|
| 52 |
- lr_scheduler_warmup_steps: 200
|
| 53 |
- num_epochs: 8
|
| 54 |
+
- weight_decay: 0.001
|
| 55 |
+
- gradient_acumulation_steps: 1
|
| 56 |
|
| 57 |
### Training results
|
| 58 |
|