llama.cpp
ggml-cpu: add RVV repack GEMM and GEMV for quantization types
#19121

Merged

ggml-cpu: add RVV repack GEMM and GEMV for quantization types #19121

ggerganov merged 7 commits into ggml-org:master from riseproject-dev:10x/riscv-quant-repack

taimur-10x requested a review from

ggerganov 45 days ago

github-actions added ggml

taimur-10x force pushed from eba08396 to 1f89bd78 44 days ago

taimur-10x force pushed from 1f89bd78 to d4197d9e 38 days ago

taimur-10x force pushed from d4197d9e to ccce8408 33 days ago

taimur-10x force pushed from ccce8408 to 35bcbdcc 26 days ago

ixgbe approved these changes on 2026-02-27

ixgbe commented on 2026-02-27

ggml-cpu: add rvv ggml_quantize_mat_4x8 for q8_0

ff49890d

ggml-cpu: add rvv repacking for iq4_nl

9bd6c9e3

ggml-cpu: add generic impl for iq4_nl gemm/gemv

dbd16674

ggml-cpu: add rvv repacking for q8_0

d9fbb69f

ggml-cpu: refactor; add rvv repacking for q4_0, q4_K

6502d03b

ggml-cpu: refactor; add rvv repacking for q2_K

981bbc58

taimur-10x force pushed from f41b1298 to 264a9712 8 days ago

taimur-10x force pushed from 264a9712 to fb95e742 8 days ago

ggml-cpu: refactor rvv repack

a037b851

taimur-10x force pushed from fb95e742 to a037b851 8 days ago

ggerganov merged af237f30 into master 2 days ago

Reviewers

ixgbe

ggerganov

Assignees

No one assigned

Labels

ggml

Milestone

No milestone