metadata
license: llama2
datasets:
- theblackcat102/evol-codealpaca-v1
- Pasta009/Instruction-Fusion-Code-v1
language:
- en
metrics:
- accuracy
pipeline_tag: text-generation
tags:
- code
Trained with data generated by the Instruction Fusion method.
- Base Model: CodeLlama-Python-34b
- Seed Instructions: theblackcat102/evol-codealpaca-v1
- Repository: https://github.com/XpastaX/Instruction-Fusion
- Paper: https://arxiv.org/abs/2312.15692
Dataset:
- theblackcat102/evol-codealpaca-v1: Python instructions Only
- Pasta009/Instruction-Fusion-Code-v1: Python instructions Only (first 50k)
BibTeX:
@misc{guo2024instruction,
title={Instruction Fusion: Advancing Prompt Evolution through Hybridization},
author={Weidong Guo and Jiuding Yang and Kaitong Yang and Xiangyang Li and Zhuwei Rao and Yu Xu and Di Niu},
year={2024},
eprint={2312.15692},
archivePrefix={arXiv},
primaryClass={cs.AI}
}