GraphCanon updated 2w · GitHub synced 2w
Decision brief
Optimized Triton kernels for accelerating LLM training, especially on ROCm PyTorch installations.
Good fit when
- When enhancing training speed of large language models with ROCm-compatible hardware.
- For projects requiring customization through install flags, like enabling cuTile with tileiras.
Avoid when
- Avoid if only CUDA environments are supported, as Liger-Kernel emphasizes ROCm compatibility.
- Skip for simple setup requirements; prefer more streamlined tools without extensive customization options.
Observed Jul 12, 2026 · Source: enrich:decision_facts
Verify the decision
Maintenance and security
Full trust report- Maintenance
- Very active (0d since push)
- As of 2w
- Provenance
- Not a fork · Organization account
- As of 2w
- Security (OSV)
- No lockfile
- As of 1mo
Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.
Install
pip install Liger-KernelSimilar tools
Same-category neighbours. No typed graph edges are catalogued for this tool yet.
Evidence and technical details
Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.
Overview
Provides optimized Triton kernels designed to speed up the training process of large language models (LLMs). Supports ROCm PyTorch installation and offers customization options through installation flags.
Capability facts
- Languages
- python
Source: github.language+pyproject.toml · Aug 7, 2026
Categories
Compatibility
Sourced claims from the README excerpt - not unsourced marketing copy.
Tags
README
On ROCm, install ROCm PyTorch first from the PyTorch ROCm index.
pip install -e .
Or install cuTile with the optional tileiras compiler
pip install -e ".[cutile-tileiras]"
Getting Started
There are a couple of ways to apply Liger kernels, depending on the level of customization required.
Contributing, Acknowledgements, and License
For agents
This page has a .md twin and JSON over the API.