Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.15701 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915350515286016 |
|---|---|
| author | Pan, Haolin Lin, Hongyu Luo, Haoran Liu, Yang Yao, Kaichun Zhang, Libo Xing, Mingjie Wu, Yanjun |
| author_facet | Pan, Haolin Lin, Hongyu Luo, Haoran Liu, Yang Yao, Kaichun Zhang, Libo Xing, Mingjie Wu, Yanjun |
| contents | Compiler auto-tuning optimizes pass sequences to improve performance metrics such as Intermediate Representation (IR) instruction count. Although recent advances leveraging Large Language Models (LLMs) have shown promise in automating compiler tuning, two significant challenges still remain: the absence of high-quality reasoning datasets for agents training, and limited effective interactions with the compilation environment. In this work, we introduce Compiler-R1, the first reinforcement learning (RL)-driven framework specifically augmenting LLM capabilities for compiler auto-tuning. Compiler-R1 features a curated, high-quality reasoning dataset and a novel two-stage end-to-end RL training pipeline, enabling efficient environment exploration and learning through an outcome-based reward. Extensive experiments across seven datasets demonstrate Compiler-R1 achieving an average 8.46% IR instruction count reduction compared to opt -Oz, showcasing the strong potential of RL-trained LLMs for compiler optimization. Our code and datasets are publicly available at https://github.com/Panhaolin2001/Compiler-R1. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_15701 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Compiler-R1: Towards Agentic Compiler Auto-tuning with Reinforcement Learning Pan, Haolin Lin, Hongyu Luo, Haoran Liu, Yang Yao, Kaichun Zhang, Libo Xing, Mingjie Wu, Yanjun Machine Learning Artificial Intelligence Compiler auto-tuning optimizes pass sequences to improve performance metrics such as Intermediate Representation (IR) instruction count. Although recent advances leveraging Large Language Models (LLMs) have shown promise in automating compiler tuning, two significant challenges still remain: the absence of high-quality reasoning datasets for agents training, and limited effective interactions with the compilation environment. In this work, we introduce Compiler-R1, the first reinforcement learning (RL)-driven framework specifically augmenting LLM capabilities for compiler auto-tuning. Compiler-R1 features a curated, high-quality reasoning dataset and a novel two-stage end-to-end RL training pipeline, enabling efficient environment exploration and learning through an outcome-based reward. Extensive experiments across seven datasets demonstrate Compiler-R1 achieving an average 8.46% IR instruction count reduction compared to opt -Oz, showcasing the strong potential of RL-trained LLMs for compiler optimization. Our code and datasets are publicly available at https://github.com/Panhaolin2001/Compiler-R1. |
| title | Compiler-R1: Towards Agentic Compiler Auto-tuning with Reinforcement Learning |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2506.15701 |