mirror of
https://github.com/ggerganov/whisper.cpp.git
synced 2025-05-09 20:13:14 +00:00
CUDA: fix FA logic for PTX 7.0 and CC >= 7.5 (llama/12222)
This commit is contained in:
parent
b9eab73fa2
commit
e3c85e75bd
@ -310,7 +310,7 @@ void ggml_cuda_flash_attn_ext(ggml_backend_cuda_context & ctx, ggml_tensor * dst
|
||||
}
|
||||
|
||||
// The MMA implementation needs Turing or newer, use the old WMMA code for Volta:
|
||||
if (cc == GGML_CUDA_CC_VOLTA) {
|
||||
if (fp16_mma_available(cc) && !new_mma_available(cc)) {
|
||||
ggml_cuda_flash_attn_ext_wmma_f16(ctx, dst);
|
||||
return;
|
||||
}
|
||||
|
Loading…
x
Reference in New Issue
Block a user