Update README.md
Browse files
README.md
CHANGED
|
@@ -43,14 +43,14 @@ ORPO
|
|
| 43 |
### Training Parameters
|
| 44 |
## Training Arguments:
|
| 45 |
|
| 46 |
-
Learning Rate: 1e-5
|
| 47 |
-
Batch Size: 1
|
| 48 |
-
max_steps: 1
|
| 49 |
-
Block Size: 512
|
| 50 |
-
Warmup Ratio: 0.1
|
| 51 |
-
Weight Decay: 0.01
|
| 52 |
-
Gradient Accumulation: 4
|
| 53 |
-
Mixed Precision: bf16
|
| 54 |
|
| 55 |
|
| 56 |
#### Training Hyperparameters
|
|
|
|
| 43 |
### Training Parameters
|
| 44 |
## Training Arguments:
|
| 45 |
|
| 46 |
+
- Learning Rate: 1e-5
|
| 47 |
+
- Batch Size: 1
|
| 48 |
+
- max_steps: 1
|
| 49 |
+
- Block Size: 512
|
| 50 |
+
- Warmup Ratio: 0.1
|
| 51 |
+
- Weight Decay: 0.01
|
| 52 |
+
- Gradient Accumulation: 4
|
| 53 |
+
- Mixed Precision: bf16
|
| 54 |
|
| 55 |
|
| 56 |
#### Training Hyperparameters
|