shibing624
commited on
Commit
•
f17a52b
1
Parent(s):
95db440
Update README.md
Browse files
README.md
CHANGED
@@ -30,7 +30,7 @@ widget:
|
|
30 |
|
31 |
本模型是 [decapoda-research/llama-13b-hf](https://huggingface.co/decapoda-research/llama-13b-hf)
|
32 |
底座模型 合并 [ziqingyang/chinese-llama-plus-lora-13b](https://huggingface.co/ziqingyang/chinese-llama-plus-lora-13b) LoRA权重,
|
33 |
-
并转化为HuggingFace版本权重(.bin文件),可以在此中文LLaMA
|
34 |
|
35 |
|
36 |
test case:
|
|
|
30 |
|
31 |
本模型是 [decapoda-research/llama-13b-hf](https://huggingface.co/decapoda-research/llama-13b-hf)
|
32 |
底座模型 合并 [ziqingyang/chinese-llama-plus-lora-13b](https://huggingface.co/ziqingyang/chinese-llama-plus-lora-13b) LoRA权重,
|
33 |
+
并转化为HuggingFace版本权重(.bin文件),可以在此中文LLaMA模型上继续指令微调训练,LLaMA模型为底座模型,直接调用会效果不佳。
|
34 |
|
35 |
|
36 |
test case:
|