llama.cpp
ggml-cuda: fallback to F32 on device without BF16 hardware acceleration
#28846
Merged
Overview
Commits
2
Changes
View On GitHub

ggml-cuda: fallback to F32 on device without BF16 hardware acceleration #28846

JohannesGaessler merged 2 commits into ggml-org:master from thelittlefireman:feature_unsupported_bf16
thelittlefireman
thelittlefireman ggml-cuda: fallback to F32 on device without BF16 hardware accelerati…
45045a75
thelittlefireman thelittlefireman requested a review 24 days ago
github-actions github-actions added ggml
github-actions github-actions added CUDA
JohannesGaessler apply logic to NVIDIA as well
bbd6d465
JohannesGaessler
JohannesGaessler
JohannesGaessler approved these changes on 2026-09-13
IMbackK
IMbackK approved these changes on 2026-09-13
JohannesGaessler JohannesGaessler merged ad6c6683 into master 24 days ago
eliah
thelittlefireman
JohannesGaessler
eliah
LostRuins
JohannesGaessler
thelittlefireman

Login to write a write a comment.

Login via GitHub

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