---
language:
- en
license: apache-2.0
datasets:
- openbmb/UltraFeedback
pipeline_tag: text-generation
model-index:
- name: GPO-Llama-3-8B-Instruct-GPM-2B
  results: []
---

General Preference Modeling with Preference Representations for Aligning Language Models (https://arxiv.org/abs/2410.02197) 

# GPO-Llama-3-8B-Instruct-GPM-2B

This model was developed using [General Preference Optimization (GPO)](https://arxiv.org/abs/2405.00675) at iteration 3 and the [General Preference representation Model (GPM)](https://arxiv.org/abs/2410.02197) (specifically, using [GPM-Gemma-2B](https://huggingface.co/general-preference/GPM-Gemma-2B)), 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
- [SPPO-Llama-3-8B-Instruct-GPM-2B](https://huggingface.co/general-preference/SPPO-Llama-3-8B-Instruct-GPM-2B)
- [GPO-Llama-3-8B-Instruct-GPM-2B](https://huggingface.co/general-preference/GPO-Llama-3-8B-Instruct-GPM-2B)

### 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 |
|-------------------------------------------|:------------:|:--------:|:-----------:|
|[GPO-Llama-3-8B-Instruct-GPM-2B](https://huggingface.co/general-preference/GPO-Llama-3-8B-Instruct-GPM-2B) | 38.43 | 48.87 | 2613 



## [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 |
|--------|---------------|----------------|------------|-------|-----------|-------|---------|
|[GPO-Llama-3-8B-Instruct-GPM-2B](https://huggingface.co/general-preference/GPO-Llama-3-8B-Instruct-GPM-2B) | 61.43 | 53.54 | 75.22 | 76.12 | 78.06 | 65.65 | 68.34  


### Training hyperparameters
The following hyperparameters were used during training:

- learning_rate: 5e-07
- beta: 0.001
- 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
```
@article{zhang2024general,
  title={General Preference Modeling with Preference Representations for Aligning Language Models},
  author={Zhang, Yifan and Zhang, Ge and Wu, Yue and Xu, Kangping and Gu, Quanquan},
  journal={arXiv preprint arXiv:2410.02197},
  year={2024}
}
```