Update README.md
Browse files
README.md
CHANGED
|
@@ -5,6 +5,7 @@ license: cc-by-nc-2.0
|
|
| 5 |
This is a merge of [LongAlpaca-70B-lora](https://huggingface.co/Yukang/LongAlpaca-70B-lora) into lizpreciatior's [lzlv_70b_fp16_hf](https://huggingface.co/lizpreciatior/lzlv_70b_fp16_hf), and removing the extra row and pad token so that the vocabularies match.
|
| 6 |
|
| 7 |
There is no additional fine-tuning. The resulting model seems to not be broken... you can test whether it is truly the original model + 32K capability (use linear rope scaling 8).
|
|
|
|
| 8 |
|
| 9 |
You could also try merging this with other models of longLORA descendency (like [Aurelian](https://huggingface.co/grimulkan/aurelian-v0.5-70b-rope8-32K-fp16)).
|
| 10 |
|
|
|
|
| 5 |
This is a merge of [LongAlpaca-70B-lora](https://huggingface.co/Yukang/LongAlpaca-70B-lora) into lizpreciatior's [lzlv_70b_fp16_hf](https://huggingface.co/lizpreciatior/lzlv_70b_fp16_hf), and removing the extra row and pad token so that the vocabularies match.
|
| 6 |
|
| 7 |
There is no additional fine-tuning. The resulting model seems to not be broken... you can test whether it is truly the original model + 32K capability (use linear rope scaling 8).
|
| 8 |
+
[ChuckMcSneed](https://huggingface.co/ChuckMcSneed) did a benchmark [here](https://huggingface.co/grimulkan/lzlv-longLORA-70b-rope8-32k-fp16/discussions/2), indicating 30% degradation with 8x the context length.
|
| 9 |
|
| 10 |
You could also try merging this with other models of longLORA descendency (like [Aurelian](https://huggingface.co/grimulkan/aurelian-v0.5-70b-rope8-32K-fp16)).
|
| 11 |
|