PR #13101 Force FP32 compute in GLM4 FFN Down

Force FP32 compute in GLM4 FFN Down #13101

JohannesGaessler merged 4 commits into ggml-org:master from city96:glm4_cublas_fix

Force FP32 compute in cuBLAS GEMM

6efd8727

github-actions added Nvidia GPU

github-actions added ggml

JohannesGaessler requested changes on 2025-04-25

Revert "Force FP32 compute in cuBLAS GEMM"

db52579a

Force F32 compute in GLM4 ffn down

70975676

JohannesGaessler approved these changes on 2025-04-25

city96 changed the title ~~CUDA: Force FP32 compute in cuBLAS GEMM~~ Force FP32 compute in GLM4 FFN Down 232 days ago

Edit comment to clarify issue

06113f00

JohannesGaessler merged 558a7647 into master 232 days ago

city96 deleted the glm4_cublas_fix branch 232 days ago

Reviewers

JohannesGaessler

Assignees

No one assigned

Labels

Nvidia GPU ggml

Milestone

No milestone