Sq Format Sparse Quantized

Unify sparse and quantized representations in a single hardware-friendly format for efficient LLM inference. Exploit complementary acceleration properties—high precision for sparse operations, low precision for dense—when W4A8 bottlenecks GPU throughput.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 5b0739e340

Frequently asked questions

npx get-skill add adu2021/sq-format-sparse-quantized