Skip to main content

Custom Triton/CUDA kernels that complement Liger Kernel for LLM post-training

Project description

Chalk

Custom Triton/CUDA kernels that complement Liger Kernel.

pip install freesolo-chalk

Liger fuses the cross-entropy, activation, and RMSNorm paths. Chalk fills the gaps that matter for Freesolo's Flash post-training stack — fused GEMMs, the LoRA-delta matmuls, the QKV norm+RoPE epilogue, embedding gather, and FP8 frozen-base GEMMs — each behind a documented, benchmarked, opt-in entry point.

Chalk's repository layout and conventions intentionally mirror Liger Kernel.

Layout

src/chalk/
  ops/           # raw Triton/CUDA kernels + autograd.Function wrappers
  transformers/  # model-level installers that monkeypatch kernels into HF modules
  utils.py       # device detection helpers
test/            # correctness + gating tests (mirrors test/ops, test/transformers)

Design principles

  • Worker-side kernel library. Like Liger, chalk depends on torch + triton — it is meant to be installed where kernels actually run (the GPU worker), so consumers should depend on it from a gpu extra rather than their base install. Importing the top-level chalk package is still cheap (kernels lazy-load), so probing chalk.utils.infer_device() never forces a heavy import.
  • Complements, not replaces, Liger. Liger fuses CE / activation / RMSNorm; chalk fuses the GEMMs, LoRA delta, QKV epilogue, embedding, and FP8 base.
  • Safe fallback. Every installer is arch-gated, runs a numeric self-test on install, patches only frozen nn.Linear layers (never trainable / PEFT-wrapped layers), and silently falls back to the eager / Liger path on any import / compile / self-test failure.
  • Opt-in & evidence-based. Kernels are off unless explicitly enabled, and every kept kernel has end-to-end loss-curve evidence — not just a microbenchmark.

Development

pip install -e '.[dev]'
make checkstyle   # ruff check + format
make test         # pytest with coverage

Status

Intentionally minimal to start — kernels are landed one at a time under chalk/ops + chalk/transformers, each with correctness tests.

License

BSD-2-Clause. See LICENSE and NOTICE.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

freesolo_chalk-0.4.1.tar.gz (148.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

freesolo_chalk-0.4.1-py3-none-any.whl (92.3 kB view details)

Uploaded Python 3

File details

Details for the file freesolo_chalk-0.4.1.tar.gz.

File metadata

  • Download URL: freesolo_chalk-0.4.1.tar.gz
  • Upload date:
  • Size: 148.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.11.14 {"installer":{"name":"uv","version":"0.11.14","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for freesolo_chalk-0.4.1.tar.gz
Algorithm Hash digest
SHA256 17db205d183731a435ea4d381cb886a082468524868730c4aa39f28263c7c81b
MD5 7db86a51e19f64108de8faac4c74a600
BLAKE2b-256 e9f00cf11b0b89c3d3d0ad276701b465c9027ac05f5644ea0ff0ea28bb145abc

See more details on using hashes here.

File details

Details for the file freesolo_chalk-0.4.1-py3-none-any.whl.

File metadata

  • Download URL: freesolo_chalk-0.4.1-py3-none-any.whl
  • Upload date:
  • Size: 92.3 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.11.14 {"installer":{"name":"uv","version":"0.11.14","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for freesolo_chalk-0.4.1-py3-none-any.whl
Algorithm Hash digest
SHA256 c028ee0f0992603dfa156e92ce556df7ad3443e1cf686cff59696921097215a1
MD5 0f4aa29659c15a1bf03302fc7724df19
BLAKE2b-256 e773a67b6b58275fb15c43aac246a10203a213b9d4001d8b83e35aa9df65c013

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page