google/ml-compiler-opt
View on GitHub ↗ ★ 807 Python
Infrastructure for Machine Learning Guided Optimization (MLGO) in LLVM.
6 mentions on Hacker News by 4 people: 0 stories, 6 comments · first 2022-07-06 · latest 2023-09-18
20082026
-
…an optimal set, and this is sometimes done as a baseline in pieces of ML-in-compiler research. 1. https://github.com/google/ml-compiler-opt/pull/109 2. https://youtu.be/0uUKDQyn1Z4?si=PHrx9RICJIiA3E6C
comment by boomanaiden154 · 2023-09-18 · in Large Language Models for Compiler Optimization -
…definitely feasible if the problem is important enough to you to throw some compute at it [2]. 1. https://github.com/google/ml-compiler-opt 2. https://doi.org/10.1145/3503222.3507744
comment by boomanaiden154 · 2023-05-07 · in How to make smaller C and C++ binaries -
Looks like they do have a pretrained model: https://github.com/google/ml-compiler-opt/releases/tag/inlin... The code will by default auto-download it during the build process. It's about 800 kbytes, which seems very reasonable for…
comment by londons_explore · 2022-07-07 · in Google ML Compiler Inlining Achieves 3-7% Reduction in Size -
Based on https://github.com/google/ml-compiler-opt#pretrained-models , it seems like the models are versioned, so I guess they'd update their models when updating LLVM or something like that.
comment by jamesfinlayson · 2022-07-07 · in Google ML Compiler Inlining Achieves 3-7% Reduction in Size -
I'm not sure - https://github.com/google/ml-compiler-opt#pretrained-models says the they sometimes get released on GitHub but I couldn't see anything obvious in that repo.
comment by jamesfinlayson · 2022-07-07 · in Google ML Compiler Inlining Achieves 3-7% Reduction in Size -
…solution on github and a demo that uses policy gradient to train an inlining-for-size policy. https://github.com/google/ml-compiler-opt https://github.com/google/ml-compiler-opt/blob/main/docs/demo/demo.md With code, that's awesome—what I…
comment by mcint · 2022-07-06 · in Google ML Compiler Inlining Achieves 3-7% Reduction in Size