Update README.md
Browse files
README.md
CHANGED
|
@@ -63,9 +63,7 @@ A 36k+ expanded dataset is planned for v2.0.
|
|
| 63 |
|
| 64 |
### **Hyperparameters**
|
| 65 |
- Epochs: 6
|
| 66 |
-
- Batch size:
|
| 67 |
-
0.35b + 0.7b = 4
|
| 68 |
-
1.2b + 2.6b = 16
|
| 69 |
- Learning rate: cosine schedule, peak ~4e‑5
|
| 70 |
- Optimizer: AdamW
|
| 71 |
- Gradient clipping: 1.0
|
|
|
|
| 63 |
|
| 64 |
### **Hyperparameters**
|
| 65 |
- Epochs: 6
|
| 66 |
+
- Batch size: 4
|
|
|
|
|
|
|
| 67 |
- Learning rate: cosine schedule, peak ~4e‑5
|
| 68 |
- Optimizer: AdamW
|
| 69 |
- Gradient clipping: 1.0
|