File size: 2,141 Bytes
cdd4cb8
 
0b0a813
 
 
 
 
cdd4cb8
 
5ae3103
 
4f82cba
5ae3103
4f82cba
5ae3103
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
44
45
46
---
library_name: transformers
license: mit
datasets:
- HuggingFaceH4/ultrafeedback_binarized
language:
- en
---

<!-- This is a model released from the preprint: *[Bootstrapping Language Models with DPO Implicit Rewards](https://arxiv.org/abs/2406.09760)*. Please refer to our [repository](https://github.com/sail-sg/dice) for more details. -->

# Llama-3-Base-8B-DICE-Iter1

This model was developed using [Bootstrapping Language Models with DPO Implicit Rewards](https://arxiv.org/abs/2406.09760) (DICE) at iteration 1, based on the [princeton-nlp/Llama-3-Base-8B-SFT-DPO](https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-DPO) architecture as the starting point. 

<!-- We utilized the prompt sets extracted from [HuggingFaceH4/ultrafeedback_binarized](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). -->

## Links to Other Models
- [Llama-3-Base-8B-DICE-Iter1](https://huggingface.co/sail/Llama-3-Base-8B-DICE-Iter1)
- [Llama-3-Base-8B-DICE-Iter2](https://huggingface.co/sail/Llama-3-Base-8B-DICE-Iter2)

## Model Description

- Model type: An 8B parameter GPT-like model fine-tuned on synthetic datasets.
- Language(s) (NLP): Primarily English
- License: MIT
- Fine-tuned from model: princeton-nlp/Llama-3-Base-8B-SFT-DPO

## [AlpacaEval Leaderboard Evaluation Results](https://tatsu-lab.github.io/alpaca_eval/)

|                Model                           | LC. Win Rate | Win Rate | 
|-------------------------------------------|:------------:|:--------:|
|[Llama-3-Base-8B-SFT-DPO](https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-DPO) |18.20 |15.50
|[Llama-3-Base-8B-DICE-Iter1](https://huggingface.co/sail/Llama-3-Base-8B-DICE-Iter1) |25.08 |25.77
|[Llama-3-Base-8B-DICE-Iter2](https://huggingface.co/sail/Llama-3-Base-8B-DICE-Iter2) |**27.55** |**30.99**

## Citation

```bibtex
@article{chen2024bootstrapping,
  title={Bootstrapping Language Models with DPO Implicit Rewards},
  author={Chen, Changyu and Liu, Zichen and Du, Chao and Pang, Tianyu and Liu, Qian and Sinha, Arunesh and Varakantham, Pradeep and Lin, Min},
  journal={arXiv preprint arXiv:2406.09760},
  year={2024}
}
```