-
Update for latest kernel-builder (add revision)
danieldk HF Staff
-
Update for build.toml changes
danieldk HF Staff
-
Update README.md (#1)

danieldk +1: lysandre HF Staff - Update README.md (77ee5b6d9f086c00960c2289874eb3726db0fabe)
Co-authored-by: Lysandre <lysandre@users.noreply.huggingface.co>
-
Build (Torch 2.6)
danieldk HF Staff
-
Build
danieldk HF Staff -
Export `ScalarType`/`scalartypes`
danieldk HF Staff -
Update flake.lock
danieldk HF Staff
-
Build
danieldk HF Staff -
Sync with vLLM
danieldk HF Staff This fixes (among other things) a race condition in GPTQ-Marlin.
-
Build
danieldk HF Staff -
Adjust for latest kernel-builder changes
danieldk HF Staff
-
Build
danieldk HF Staff
-
Simplify `flake.nix`
danieldk HF Staff
-
Build
danieldk HF Staff -
Expose ops (handy for tests etc.)
danieldk HF Staff
-
Add a `flake.nix` for convenience
danieldk HF Staff
-
Build
danieldk HF Staff -
Add full Marlin support and tests for Marlin/CUTLASS
danieldk HF Staff
-
Import CUTLASS tests and add missing scaled mm with zp signature
danieldk HF Staff -
Add missing `scalar_type.hpp`
danieldk HF Staff -
Build
danieldk HF Staff -
Add GPTQ-Marlin
danieldk HF Staff
-
Build
danieldk HF Staff -
Add `scaled_(int|fp8)_quant` and `fp8_marlin_gemm`
danieldk HF Staff -
Build
danieldk HF Staff -
Fixup metadata
danieldk HF Staff -
Add cutlass_w8a8
danieldk HF Staff -
initial commit
danieldk HF Staff