Commit
·
1ea7aae
1
Parent(s):
80c1224
Update README.md
Browse files
README.md
CHANGED
|
@@ -53,6 +53,7 @@ The following hyperparameters were used during training:
|
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
- lr_scheduler_warmup_steps: 18
|
| 55 |
- num_epochs: 1
|
|
|
|
| 56 |
|
| 57 |
### Training results
|
| 58 |
|
|
|
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
- lr_scheduler_warmup_steps: 18
|
| 55 |
- num_epochs: 1
|
| 56 |
+
- dpo_beta: .1
|
| 57 |
|
| 58 |
### Training results
|
| 59 |
|