File size: 924 Bytes
0576ba4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
730b57b
0576ba4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
730b57b
0576ba4
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
---
# For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/modelcard.md?plain=1
# Doc / guide: https://huggingface.co/docs/hub/model-cards
{}
---
# dataset Intruction
---
**datasets:** \
- mxz/CValues_DPO \
**language:** \
- zh \
- en \
**metrics:** \
- perplexity \
**pipeline_tag:** \
- text-generation \
**tags:** \
- DPO \
- fintune \
- alignment \
- LoRA \
- Llama-3
--- 

# About mxz-llama-3-8B-sft

This model trained by SFT and PPO.

It's have coding, reasoing, chinese QA .

# evaluation

Result:

| Model               | MMLU  | C-EVAL | C-MMLU |
| ------------------- | ----- | ------ | ------ |
| Llama-3-8B          | 55.5  | 47.0   | 48.0   |
| Llama-3-8B-Instruct | 60.1  | 49.7   | 49.3   |
| Llama-3-8B-dpo      | 62.2  | 49.9   | 49.4   |

- Llama-3-8B evaluation result from [ymcui/Chinese-LLaMA-Alpaca-3](https://github.com/ymcui/Chinese-LLaMA-Alpaca-3)