llama.cpp
ggml : add BF16 unary, GLU, binary and scale ops (CPU, CUDA)
#29675

Merged

ggml : add BF16 unary, GLU, binary and scale ops (CPU, CUDA) #29675

am17an merged 4 commits into master from aman/cuda-bf16-elementwise

ggml : add BF16 unary, GLU, binary and scale ops (CPU, CUDA)

248fecc0

am17an requested a review from

ggerganov 22 days ago

am17an requested a review 22 days ago

ggerganov commented on 2026-09-29

github-actions added testing

github-actions added ggml

github-actions added CUDA

JohannesGaessler approved these changes on 2026-09-29

ggml-cpu : use per-op _bf16 functions for BF16 unary and GLU ops

30a6e87b

CUDA: use ggml_cuda_cast in binbcast and unary kernels to fix the HIP…

f575038e

ggml-openvino : reject BF16 SCALE and mixed-type BF16 ADD/MUL/SUB

71d72e14

am17an requested a review from

wine99 22 days ago

github-actions added OpenVINO

ggerganov approved these changes on 2026-09-30

am17an merged 2090f60f into master 22 days ago

Reviewers

ggerganov

JohannesGaessler

wine99

Assignees

No one assigned

Labels

testing ggml CUDA OpenVINO

Milestone

No milestone