AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data

TL;DR

AlchemistCoder enhances code generation and generalization in LLMs by fine-tuning on multi-source data with a novel harmonization method using hindsight relabeling and data construction processes, outperforming comparable models.

Contribution

The paper introduces AlchemistCoder, a new fine-tuning approach that harmonizes multi-source code data and incorporates data construction tasks to improve code LLM performance.

Findings

Outperforms models of similar size (6.7B/7B)

Rivals or surpasses larger models (15B/33B/70B)

Demonstrates improved instruction-following and code understanding

Abstract

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.