Update README.md
Browse files
README.md
CHANGED
|
@@ -145,8 +145,7 @@ For training data details, please see the [Dolma](https://huggingface.co/dataset
|
|
| 145 |
|
| 146 |
### Hyperparameters
|
| 147 |
|
| 148 |
-
The hyperparameters for the two phases of training are below
|
| 149 |
-
Certainly! Here's the table with SFT and DPO as rows:
|
| 150 |
|
| 151 |
| | Learning Rate | Beta | Epochs | Warmup | Weight Decay | Gradient Clipping | Maximum Sequence Length |
|
| 152 |
|-------------------------|---------------|------|--------|------------------------------------------------------------------------|--------------|-------------------|-------------------------|
|
|
|
|
| 145 |
|
| 146 |
### Hyperparameters
|
| 147 |
|
| 148 |
+
The hyperparameters for the two phases of training are below:
|
|
|
|
| 149 |
|
| 150 |
| | Learning Rate | Beta | Epochs | Warmup | Weight Decay | Gradient Clipping | Maximum Sequence Length |
|
| 151 |
|-------------------------|---------------|------|--------|------------------------------------------------------------------------|--------------|-------------------|-------------------------|
|