llama.cpp
CUDA: remove -sm row, refactor cuBLAS
#24216
Open

CUDA: remove -sm row, refactor cuBLAS #24216

JohannesGaessler
JohannesGaessler JohannesGaessler requested a review from CISC CISC 15 days ago
JohannesGaessler JohannesGaessler requested a review 15 days ago
github-actions github-actions added documentation
github-actions github-actions added Nvidia GPU
github-actions github-actions added ggml
JohannesGaessler
JohannesGaessler
CISC
CISC approved these changes on 2026-06-06
JohannesGaessler CUDA: remove -sm row, refactor cuBLAS
3e9d26a1
JohannesGaessler fix CDNA + BF16 logic
874bd7ef
JohannesGaessler fix bad return
fdf64b88
JohannesGaessler JohannesGaessler force pushed from 6bcdfe5a to fdf64b88 15 days ago
gaugarg-nv
gaugarg-nv commented on 2026-06-06
JohannesGaessler fix src0 strides, contiguous requirements
d739f87e
JohannesGaessler JohannesGaessler requested a review from ggerganov ggerganov 14 days ago
gaugarg-nv
gaugarg-nv commented on 2026-06-07
JohannesGaessler fix GGML_CUDA_FORCE_CUBLAS
31cd8b42
gaugarg-nv
JohannesGaessler fix casts to BF16
e9e46b02

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone