AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913673553903616 |
|---|---|
| author | Song, Zifan Wang, Yudong Zhang, Wenwei Liu, Kuikun Lyu, Chengqi Song, Demin Guo, Qipeng Yan, Hang Lin, Dahua Chen, Kai Zhao, Cairong |
| author_facet | Song, Zifan Wang, Yudong Zhang, Wenwei Liu, Kuikun Lyu, Chengqi Song, Demin Guo, Qipeng Yan, Hang Lin, Dahua Chen, Kai Zhao, Cairong |
| contents | Open-source Large Language Models (LLMs) and their specialized variants, particularly Code LLMs, have recently delivered impressive performance. However, previous Code LLMs are typically fine-tuned on single-source data with limited quality and diversity, which may insufficiently elicit the potential of pre-trained Code LLMs. In this paper, we present AlchemistCoder, a series of Code LLMs with enhanced code generation and generalization capabilities fine-tuned on multi-source data. To achieve this, we pioneer to unveil inherent conflicts among the various styles and qualities in multi-source code corpora and introduce data-specific prompts with hindsight relabeling, termed AlchemistPrompts, to harmonize different data sources and instruction-response pairs. Additionally, we propose incorporating the data construction process into the fine-tuning data as code comprehension tasks, including instruction evolution, data filtering, and code review. Extensive experiments demonstrate that AlchemistCoder holds a clear lead among all models of the same size (6.7B/7B) and rivals or even surpasses larger models (15B/33B/70B), showcasing the efficacy of our method in refining instruction-following capabilities and advancing the boundaries of code intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_19265 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data Song, Zifan Wang, Yudong Zhang, Wenwei Liu, Kuikun Lyu, Chengqi Song, Demin Guo, Qipeng Yan, Hang Lin, Dahua Chen, Kai Zhao, Cairong Computation and Language Open-source Large Language Models (LLMs) and their specialized variants, particularly Code LLMs, have recently delivered impressive performance. However, previous Code LLMs are typically fine-tuned on single-source data with limited quality and diversity, which may insufficiently elicit the potential of pre-trained Code LLMs. In this paper, we present AlchemistCoder, a series of Code LLMs with enhanced code generation and generalization capabilities fine-tuned on multi-source data. To achieve this, we pioneer to unveil inherent conflicts among the various styles and qualities in multi-source code corpora and introduce data-specific prompts with hindsight relabeling, termed AlchemistPrompts, to harmonize different data sources and instruction-response pairs. Additionally, we propose incorporating the data construction process into the fine-tuning data as code comprehension tasks, including instruction evolution, data filtering, and code review. Extensive experiments demonstrate that AlchemistCoder holds a clear lead among all models of the same size (6.7B/7B) and rivals or even surpasses larger models (15B/33B/70B), showcasing the efficacy of our method in refining instruction-following capabilities and advancing the boundaries of code intelligence. |
| title | AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2405.19265 |