llama.cpp
CUDA/HIP: fix race in flash_attn_ext_f16_process_tile when nbatch_combine != DKQ/2
#30103
Merged
Overview
Commits
1
Changes
View On GitHub

CUDA/HIP: fix race in flash_attn_ext_f16_process_tile when nbatch_combine != DKQ/2 #30103

ggerganov merged 1 commit into ggml-org:master from IMbackK:fattn_fix
IMbackK
IMbackK CUDA/HIP: fix race in flash_attn_ext_f16_process_tile when nbatch_com…
1cc47333
IMbackK IMbackK requested a review 37 days ago
github-actions github-actions added ggml
github-actions github-actions added CUDA
JohannesGaessler
JohannesGaessler approved these changes on 2026-10-07
am17an
IMbackK
JohannesGaessler
am17an
IMbackK IMbackK added merge ready
IMbackK
ggerganov ggerganov merged 1bb2b9fc into master 34 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
ggml merge ready CUDA
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum