CommitLLM: A Fine-Tuned Pipeline for Git Commit Message Generation
2026-07-20 • Software Engineering
Software EngineeringArtificial Intelligence
AI summaryⓘ
The authors created CommitLLM, a tool that writes short, clear git commit messages from code changes. It uses a small language model fine-tuned on a special dataset, plus rules to keep messages brief and properly formatted. Their tests show CommitLLM makes much better and shorter messages than the original model. They also found that cleaning up the output with fixed rules helps more than just training the model better. The system is efficient enough to run on a regular consumer GPU.
git commit messageslanguage modelfine-tuningQLoRAcode diffsConventional Commitsconstrained decodingdeterministic post-processingMistral-7BLLM-as-a-Judge
Authors
Md Rafid Haque, Poojan Narendrabhai Patel, Meetkumar Vijaybhai Raychura
Abstract
Developers frequently write uninformative git commit messages such as "fix" or "update stuff", degrading the value of version-control history for code review, debugging, and onboarding. We present CommitLLM, a three-stage pipeline that generates concise, Conventional Commits-compliant messages from code diffs using a fine-tuned small language model. The system combines (1) QLoRA fine-tuning of Mistral-7B-Instruct-v0.2 on the CommitPackFT dataset, (2) constrained decoding to enforce brevity, and (3) deterministic post-processing to strip conversational artifacts and enforce format. On a 50-sample evaluation, CommitLLM achieves 98% format compliance (vs. 22% for vanilla Mistral), reduces average output length from 154.8 to 37.9 characters, and improves LLM-as-a-Judge scores from 1.97 to 3.68 out of 5. Notably, the post-processing layers contribute more to quality improvement than the fine-tuning itself, suggesting that for structured-output tasks, treating the LLM as a component in a deterministic pipeline is more effective than optimizing the model alone. The entire system runs on a single consumer GPU (NVIDIA T4, 16 GB VRAM).