Hugging Face's logo

Yhyu13
/
phi-2-sft-dpo-gpt4_en-ep1

Text Generation
Transformers
Safetensors
phi-msft
custom_code
Inference Endpoints
Model card Files Community
phi-2-sft-dpo-gpt4_en-ep1
File size: 2,076 Bytes 
7e0c8e9
 
 
 
 
570fea4
 
 
505dc7d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
---
license: other
license_name: microsoft-research-license
license_link: https://huggingface.co/microsoft/phi-2/resolve/main/LICENSE
---

This is the merged model for LoRA https://huggingface.co/Yhyu13/phi-2-sft-dpo-gpt4_en-ep1-lora

This model is a dpo improvement to this base model https://huggingface.co/Yhyu13/phi-2-sft-alpaca_gpt4_en-ep1 who achieve better than text-davinci-003 on AlpcaEval judged by ChatGPT.

# AlpacaEval

Quote from this discussion
https://huggingface.co/microsoft/phi-2/discussions/38

Since phi2 requires remote code which HF open llm leaderboard would not accept at this moment,

I ran phi2 and my dpo to the AlpcaEval benchmark

https://tatsu-lab.github.io/alpaca_eval/

Here is result evaluated by chatpgpt https://github.com/tatsu-lab/alpaca_eval/pull/183

```
                       win_rate  standard_error  n_total  avg_length
gpt4                      73.79            1.54      805        1365
claude                    70.37            1.60      805        1082
chatgpt                   66.09            1.66      805         811
wizardlm-13b              65.16            1.67      805         985
vicuna-13b                64.10            1.69      805        1037
guanaco-65b               62.36            1.71      805        1249
oasst-rlhf-llama-33b      62.05            1.71      805        1079
alpaca-farm-ppo-human     60.25            1.72      805         803
falcon-40b-instruct       56.52            1.74      805         662
phi-2-alpaca-gpt4-dpo(new)55.60            1.75      804        4532
phi-2-alpaca-gpt4(new)    54.23            1.75      804        1138
text_davinci_003          50.00            0.00      805         307
alpaca-7b                 45.22            1.74      805         396
phi-2(new)                43.79            1.74      805         924
text_davinci_001          28.07            1.56      805         296
```

phi-2-alpaca-gpt4-dpo is only slightly better than my previous sft phi-2-alpaca-gpt4, when evaluted by chatgpt, but the dpo tuned model outputs significantly longer result!