CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-03 09:22:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

CUDA: remove incorrect precision check (#7454)

Browse Source
This commit is contained in:
Johannes Gäßler
2024-05-22 10:24:29 +02:00
committed by GitHub
parent 3e5faa8503
commit 95fb0aefab
1 changed files with 0 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
ggml-cuda/fattn-tile-f32.cu
View File

@@ -286,9 +286,6 @@ void ggml_cuda_flash_attn_ext_tile_f32(ggml_backend_cuda_context & ctx, ggml_ten
const ggml_tensor * KQV = dst; const ggml_tensor * KQV = dst;
const ggml_tensor * Q = dst->src[0]; const ggml_tensor * Q = dst->src[0];
const int32_t precision = KQV->op_params[2];
GGML_ASSERT(precision == GGML_PREC_DEFAULT);
if (Q->ne[1] <= 16) { if (Q->ne[1] <= 16) {
constexpr int cols_per_block = 16; constexpr int cols_per_block = 16;
constexpr int parallel_blocks = 4; constexpr int parallel_blocks = 4;