tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-09-07 07:29:39 -04:00
Code Issues Packages Projects Releases Wiki Activity

CUDA: add FP32 FlashAttention vector kernel (#7188)

Browse Source 
* CUDA: add FP32 FlashAttention vector kernel

* fixup! CUDA: add FP32 FlashAttention vector kernel

* fixup! fixup! CUDA: add FP32 FlashAttention vector kernel

* fixup! fixup! fixup! CUDA: add FP32 FlashAttention vector kernel
This commit is contained in:
Johannes Gäßler
2024-05-12 19:40:45 +02:00
committed by GitHub
parent 6f1b63606f
commit dc685be466
9 changed files with 899 additions and 458 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
ggml-cuda/fattn-vec-f16.cuh Normal file
View File

@@ -0,0 +1,5 @@
#include "common.cuh"
void ggml_cuda_flash_attn_ext_vec_f16(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
void ggml_cuda_flash_attn_ext_vec_f16_no_mma(ggml_backend_cuda_context & ctx, ggml_tensor * dst);