ThomasBaruzier
commited on
Commit
•
2585838
1
Parent(s):
38dc4bd
Update README.md
Browse files
README.md
CHANGED
@@ -195,7 +195,7 @@ extra_gated_button_content: Submit
|
|
195 |
<!-- Better pic but I would like to talk about my quants on Linkedin so yeah <img src="https://cdn-uploads.huggingface.co/production/uploads/646410e04bf9122922289dc7/xlkSJli8IQ9KoTAuTKOF2.png" alt="llama" width="30%"/> -->
|
196 |
<img src="https://cdn-uploads.huggingface.co/production/uploads/646410e04bf9122922289dc7/LQUL7YII8okA8CG54mQSI.jpeg" alt="llama" width="60%"/>
|
197 |
|
198 |
-
Using llama.cpp commit [
|
199 |
|
200 |
Original model: [https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct)
|
201 |
|
|
|
195 |
<!-- Better pic but I would like to talk about my quants on Linkedin so yeah <img src="https://cdn-uploads.huggingface.co/production/uploads/646410e04bf9122922289dc7/xlkSJli8IQ9KoTAuTKOF2.png" alt="llama" width="30%"/> -->
|
196 |
<img src="https://cdn-uploads.huggingface.co/production/uploads/646410e04bf9122922289dc7/LQUL7YII8okA8CG54mQSI.jpeg" alt="llama" width="60%"/>
|
197 |
|
198 |
+
Using llama.cpp commit [268c566](https://github.com/ggerganov/llama.cpp/commit/268c5660062270a2c19a36fc655168aa287aaec2) for quantization, featuring llama 3.1 rope scaling factors. This fixes low-quality issues when using 8-128k context lengths.
|
199 |
|
200 |
Original model: [https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct)
|
201 |
|