Update README.md
Browse files
README.md
CHANGED
|
@@ -16,6 +16,7 @@ We simply take the average of every two experts.weight.
|
|
| 16 |
The same goes for gate.weight.
|
| 17 |
|
| 18 |
# How To Convert
|
|
|
|
| 19 |
[convert_mixtral_8x7b_to_4x7b.ipynb](https://huggingface.co/mmnga/Mixtral-Fusion-4x7B-Instruct-v0.1/blob/main/notebook/convert_mixtral_8x7b_to_4x7b.ipynb)
|
| 20 |
|
| 21 |
# Usage
|
|
|
|
| 16 |
The same goes for gate.weight.
|
| 17 |
|
| 18 |
# How To Convert
|
| 19 |
+
use colab cpu-high-memory.
|
| 20 |
[convert_mixtral_8x7b_to_4x7b.ipynb](https://huggingface.co/mmnga/Mixtral-Fusion-4x7B-Instruct-v0.1/blob/main/notebook/convert_mixtral_8x7b_to_4x7b.ipynb)
|
| 21 |
|
| 22 |
# Usage
|