[W8A8 Block Linear Refactor][2/N] Make Fp8 block linear Op use kernel abstraction.#33891

Closed

maralbahari wants to merge 15 commits intovllm-project:mainfrom

EmbeddedLLM:2n-block-scaled-rfc-pr

Commits on Feb 5, 2026

Commits on Feb 9, 2026

add missing kwargs for aiter fp8 block scaled mm func and return statement for cutlass and fix type error in dynamic deepgemm/flash-infer
maralbahari
committed
fix f-string
maralbahari
committed
improve documenetation and fix typings in init_fp8_linear_kernel
maralbahari
committed

Commits on Feb 23, 2026

Merge remote-tracking branch 'origin/main' into 2n-block-scaled-rfc-pr
maralbahari
committed

Commits on Feb 24, 2026

Merge remote-tracking branch 'origin/main' into 2n-block-scaled-rfc-pr
maralbahari
committed

Commits on Feb 25, 2026