CompilerGym is a library of reinforcement learning environments focused on compiler optimization tasks. It enables machine learning researchers to engage with compiler optimization issues in a familiar language, bridging the gap between compiler development and machine learning research. By providing easy-to-use APIs and an open-source environment, it aims to foster advancements in compiler performance using AI techniques. The tool aims to simplify experimentation with compiler optimizations and improve their implementation across various programming contexts.
• reinforcement learning environments for compiler tasks
• support for multiple compilers such as llvm and gcc
• experimentation toolkit for ai researchers
• integration with openai gym interface
• apis for python and c++
CompilerGym lets you control compiler decisions during program optimization, steering it towards the best compiled program based on reward signals.
No, you can utilize any technique such as genetic algorithms, supervised learning, or deep reinforcement learning.
CompilerGym uses specific directories for caching datasets and storing additional files, which can be overridden with environment variables.
Yes, you must call env.close() to clean up resources after using an environment.
Produce a minimal reproducible example and check CompilerGym’s logging output for debugging.
Average Rating: 0.0
5 Stars:
0 Ratings
4 Stars:
0 Ratings
3 Stars:
0 Ratings
2 Stars:
0 Ratings
1 Star:
0 Ratings
No ratings available.
A National AI Research Institute focused on learning-enabled optimization in various domains.
View DetailsA federated AI framework that integrates decentralized data sources for AI development.
View Details