Commit
•
ba6ca96
1
Parent(s):
1d82f9f
Update README.md
Browse files
README.md
CHANGED
|
@@ -25,6 +25,7 @@ datasets:
|
|
| 25 |
## 相关模型
|
| 26 |
|
| 27 |
- exl2 4bpw量化版: [Orion-zhen/Llama3-70B-Orion-Chinese-4bpw-exl2](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-4bpw-exl2)
|
|
|
|
| 28 |
- [Orion-zhen/Llama3-70B-Orion-Chinese-SE](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-SE): 精简DPO数据集以换取更高概率的emoji输出
|
| 29 |
- [Orion-zhen/Llama3-70B-Orion-Chinese-Plus](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-Plus): 重复融合了emoji对应的lora模型, emoji触发频率非常高, 语气活泼, 但是有些时候输出不稳定
|
| 30 |
|
|
|
|
| 25 |
## 相关模型
|
| 26 |
|
| 27 |
- exl2 4bpw量化版: [Orion-zhen/Llama3-70B-Orion-Chinese-4bpw-exl2](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-4bpw-exl2)
|
| 28 |
+
- [Orion-zhen/Llama3-70B-Orion-Chinese-Ultra](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-Ultra): 最终之作, 经过中文SFT训练和多个方面的emoji DPO训练, 意在呈现最优秀的模型
|
| 29 |
- [Orion-zhen/Llama3-70B-Orion-Chinese-SE](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-SE): 精简DPO数据集以换取更高概率的emoji输出
|
| 30 |
- [Orion-zhen/Llama3-70B-Orion-Chinese-Plus](https://huggingface.co/Orion-zhen/Llama3-70B-Orion-Chinese-Plus): 重复融合了emoji对应的lora模型, emoji触发频率非常高, 语气活泼, 但是有些时候输出不稳定
|
| 31 |
|