File size: 3,529 Bytes
b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb b7ba939 0c324fb |
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 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 |
---
license: apache-2.0
datasets:
- openbmb/UltraFeedback
language:
- en
pipeline_tag: text-generation
---
Self-Play Preference Optimization for Language Model Alignment (https://arxiv.org/abs/2405.00675)
# Llama-3-Instruct-8B-SPPO-Iter1
This model was developed using [Self-Play Preference Optimization](https://arxiv.org/abs/2405.00675) at iteration 1, based on the [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) architecture as starting point. We utilized the prompt sets from the [openbmb/UltraFeedback](https://huggingface.co/datasets/openbmb/UltraFeedback) dataset, splited to 3 parts for 3 iterations by [snorkelai/Snorkel-Mistral-PairRM-DPO-Dataset](https://huggingface.co/datasets/snorkelai/Snorkel-Mistral-PairRM-DPO-Dataset). All responses used are synthetic.
## Links to Other Models
- [Llama-3-Instruct-8B-SPPO-Iter1](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter1)
- [Llama-3-Instruct-8B-SPPO-Iter2](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter2)
- [Llama-3-Instruct-8B-SPPO-Iter3](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3)
### Model Description
- Model type: A 8B parameter GPT-like model fine-tuned on synthetic datasets.
- Language(s) (NLP): Primarily English
- License: Apache-2.0
- Finetuned from model: meta-llama/Meta-Llama-3-8B-Instruct
## [AlpacaEval Leaderboard Evaluation Results](https://tatsu-lab.github.io/alpaca_eval/)
| Model | LC. Win Rate | Win Rate | Avg. Length |
|-------------------------------------------|:------------:|:--------:|:-----------:|
|[Llama-3-8B-SPPO Iter1](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter1) |31.73 |31.74 | 1962
|[Llama-3-8B-SPPO Iter2](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter2) |35.15 |35.98 | 2021
|[Llama-3-8B-SPPO Iter3](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3) |38.77 |39.85 | 2066
## [Open LLM Leaderboard Evaluation Results](https://github.com/EleutherAI/lm-evaluation-harness)
Results are reported by using [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) v0.4.1
| | arc_challenge | truthfulqa_mc2 | winogrande | gsm8k | hellaswag | mmlu | average |
|--------|---------------|----------------|------------|-------|-----------|-------|---------|
|[Llama-3-8B-SPPO Iter1](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter1) | 63.82 | 54.96 | 76.40 | 75.44 | 79.80 | 65.65 | 69.35
|[Llama-3-8B-SPPO Iter2](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter2) | 64.93 | 56.48 | 76.87 | 75.13 | 80.39 | 65.67 | 69.91
|[Llama-3-8B-SPPO Iter3](https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3) | 65.19 | 58.04 | 77.11 | 74.91 | 80.86 | 65.60 | 70.29
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-07
- eta: 1000
- per_device_train_batch_size: 8
- gradient_accumulation_steps: 1
- seed: 42
- distributed_type: deepspeed_zero3
- num_devices: 8
- optimizer: RMSProp
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_train_epochs: 6.0 (stop at epoch=1.0)
## Citation
```
@misc{wu2024self,
title={Self-Play Preference Optimization for Language Model Alignment},
author={Wu, Yue and Sun, Zhiqing and Yuan, Huizhuo and Ji, Kaixuan and Yang, Yiming and Gu, Quanquan},
year={2024},
eprint={2405.00675},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
``` |