FP32 K tails in tinyBLAS on x86
#29806

Merged

ggml-cpu: support BF16/FP16/FP32 K tails in tinyBLAS on x86 #29806

ggerganov merged 3 commits into ggml-org:master from SongXiaoXi:bf16-k-tail

ggml-cpu: vectorize BF16 K tails in tinyBLAS

b5c15204

SongXiaoXi requested a review from

ggerganov 40 days ago

github-actions added ggml

tests: Skip tinyBLAS when use_ref is enabled so CPU tests compare aga…

525f8002

github-actions added testing

ggml-cpu: vectorize tinyBLAS F16/F32 tails

f03b742e

SongXiaoXi changed the title ~~ggml-cpu: support BF16 K tails in llamafile tinyBLAS~~ ggml-cpu: support BF16/FP16/FP32 K tails in llamafile tinyBLAS 38 days ago

SongXiaoXi changed the title ~~ggml-cpu: support BF16/FP16/FP32 K tails in llamafile tinyBLAS~~ ggml-cpu: support BF16/FP16/FP32 K tails in tinyBLAS on x86 38 days ago

am17an approved these changes on 2026-10-03

am17an added merge ready

ggerganov merged a7b94df2 into master 37 days ago

Reviewers

am17an

ggerganov

Assignees

No one assigned

Labels

testing ggml merge ready

Milestone

No milestone

llama.cpp ggml-cpu: support BF16/FP16/FP32 K tails in tinyBLAS on x86 #29806 Merged

ggml-cpu: support BF16/FP16/FP32 K tails in tinyBLAS on x86 #29806

llama.cpp
ggml-cpu: support BF16/FP16/FP32 K tails in tinyBLAS on x86
#29806

Merged