language:
- en
license: apache-2.0
size_categories: - 10K<n<100K
dataset_info:
features:- name: prompt
dtype: string - name: messages
list:- name: content
dtype: string - name: role
dtype: string
- name: content
- name: type
dtype: string
splits: - name: train_sft
num_bytes: 64859889
num_examples: 13606
download_size: 31036314
dataset_size: 64859889
configs:
- name: prompt
- config_name: default
data_files:- split: train_sft
path: data/train_sft-*
- split: train_sft
Dataset Card for Conifer
Conifer is an open-sourced dataset aiming to improve the instruction-following ability of large language models (LLM).
We recommend integrating Conifer with additional SFT datasets such as ShareGPT or Deita to enhance overall performance.
Performance
Supervised Fine-tuned (SFT) Models
| - | Final Stage | IFEval | FollowBench Avg | FollowBench Hard (L4-L5) | InFoBench | AlpacaEval LC Win Rate | MT-Bench |
|---|---|---|---|---|---|---|---|
| Deita-7B-v1.0-SFT | SFT | 45.1 | 42.0 | 31.6 | 78.6 | - | 7.22 |
| Evol-Instruct-7B-SFT | SFT | 44.0 | 40.7 | 27.6 | 75.6 | 9.4% | 6.51 |
| ShareGPT-7B-SFT | SFT | 43.3 | 42.9 | 32.3 | 78.5 | 11.6% | 6.86 |
| Conifer-7B-SFT | SFT | 50.8 | 44.9 | 35.7 | 79.5 | 12.5% | 7.08 |
DPO/RLHF Models
| - | Final Stage | IFEval | FollowBench Avg | FollowBench Hard (L4-L5) | InFoBench | AlpacaEval LC Win Rate | MT-Bench |
|---|---|---|---|---|---|---|---|
| LLaMA-2-70B-Chat | RLHF | - | 47.5 | 39.0 | 84.4 | 14.7% | 6.86 |
| Zephyr-7B-beta | DPO | 44.9 | 44.8 | 36.4 | 78.0 | 13.2% | 7.34 |
| Deita-7B-v1.0 | DPO | 51.9 | 45.7 | 38.5 | 80.9 | 16.1% | 7.55 |
| ShareGPT-7B-DPO | DPO | 48.2 | 47.7 | 38.9 | 82.0 | 15.1% | 7.10 |
| Conifer-7B-DPO | DPO | 52.3 | 50.0 | 44.1 | 82.3 | 17.1% | 7.25 |
Citation
If you find the content of this project helpful, please cite our paper as follows:
@article{
coniferlm,
title={Conifer: Improving Complex Constrained Instruction-Following Ability of Large Language Models},
author={Haoran Sun and Lixin Liu and Junjie Li and Fengyu Wang and Baohua Dong and Ran Lin and Ruohui Huang},
journal={arxiv preprint arXiv:2404.02823},
year={2024},
url={https://arxiv.org/abs/2404.02823}
}