nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 34 Packages Projects Releases Wiki Activity

ggml-cuda: add flash-attn support for DKQ=320/DV=256 with ncols2=32 (… (#22286)

Browse Source 
* ggml-cuda: add flash-attn support for DKQ=320/DV=256 with ncols2=32 (GQA=32)

Adds MMA-f16 and tile kernel configs, dispatch logic, template instances,
and tile .cu file for Mistral Small 4 (head sizes 320/256), restricting to
ncols2=32 to support GQA ratio 32 only.

* Adding check to return BEST_FATTN_KERNEL_NONE in case GQA!=32

* Apply suggestions from code review

Address review comments

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

* Address review comments and making kernel config default to DQK=512, DV=512 instead of DQK=256,DV=256

* Fixed bug with sinks=1, with ncols=32, there are two warp-groups created but sinks index is same(0,...,15) for both the groups hence with sinks=1, output is not matching with CPU output. Added sink_base which will be base index for each warp_group (threadIdx.y / np)

* Apply suggestions from code review

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

* Update ggml/src/ggml-cuda/template-instances/generate_cu_files.py

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

---------

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
lnigam
2026-04-28 21:37:35 +02:00
committed by GitHub
co-authored by Johannes Gäßler
parent 5d56effdee
commit 7b8443ac78
8 changed files with 86 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/fattn-tile.cuh
+28 -9
View File
@@ -68,6 +68,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_nv
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 4, 128, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 8, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2, 64, 64)
@@ -128,6 +130,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_nv
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2, 32, 128)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2, 32, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 256, 2, 32, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 4, 128, 2, 32, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 8, 256, 2, 32, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2, 32, 64)
@@ -195,6 +199,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_am
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 2, 32, 128)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 2, 32, 128)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 512, 1, 128, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 4, 128, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 8, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 2, 64, 64)
@@ -264,6 +270,8 @@ static constexpr __host__ __device__ uint32_t ggml_cuda_fattn_tile_get_config_am
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 16, 256, 5, 32, 256)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(256, 256, 32, 256, 3, 64, 128)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(320, 256, 32, 256, 2, 128, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 4, 128, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 8, 256, 2, 64, 64)
GGML_CUDA_FATTN_TILE_CONFIG_CASE(512, 512, 16, 256, 4, 64, 64)
@@ -1144,14 +1152,16 @@ static void launch_fattn_tile_switch_ncols1(ggml_backend_cuda_context & ctx, ggm
}
}
if (Q->ne[1] > 8/ncols2) {
constexpr int cols_per_block = 16;
const int nwarps = ggml_cuda_fattn_tile_get_nthreads (DKQ, DV, cols_per_block, cc) / warp_size;
const int nbatch_fa = ggml_cuda_fattn_tile_get_nbatch_fa(DKQ, DV, cols_per_block, cc);
fattn_kernel_t fattn_kernel = flash_attn_tile<DKQ, DV, cols_per_block/ncols2, ncols2, use_logit_softcap>;
launch_fattn<DV, cols_per_block/ncols2, ncols2>
(ctx, dst, fattn_kernel, nwarps, nbytes_shared, nbatch_fa, true, true, false, warp_size);
return;
if constexpr (ncols2 <= 16) {
if (Q->ne[1] > 8/ncols2) {
constexpr int cols_per_block = 16;
const int nwarps = ggml_cuda_fattn_tile_get_nthreads (DKQ, DV, cols_per_block, cc) / warp_size;
const int nbatch_fa = ggml_cuda_fattn_tile_get_nbatch_fa(DKQ, DV, cols_per_block, cc);
fattn_kernel_t fattn_kernel = flash_attn_tile<DKQ, DV, cols_per_block/ncols2, ncols2, use_logit_softcap>;
launch_fattn<DV, cols_per_block/ncols2, ncols2>
(ctx, dst, fattn_kernel, nwarps, nbytes_shared, nbatch_fa, true, true, false, warp_size);
return;
}
}
if constexpr (ncols2 <= 8) {
@@ -1210,6 +1220,14 @@ static void launch_fattn_tile_switch_ncols2(ggml_backend_cuda_context & ctx, ggm
const int gqa_limit = nvidia && gqa_ratio <= 4 && DV <= 256 ? 16 : INT_MAX;
const bool use_gqa_opt = mask && max_bias == 0.0f && Q->ne[1] <= gqa_limit && K->ne[1] % FATTN_KQ_STRIDE == 0;
if constexpr (DKQ == 320) { // Mistral Small 4
if (use_gqa_opt && gqa_ratio % 32 == 0) {
launch_fattn_tile_switch_ncols1<DKQ, DV, 32, use_logit_softcap>(ctx, dst);
return;
}
GGML_ABORT("flash-attn tile (320/256): expected GQA ratio multiple of 32");
}
if constexpr (DKQ == 576) {
if (use_gqa_opt && gqa_ratio % 16 == 0) {
launch_fattn_tile_switch_ncols1<DKQ, DV, 16, use_logit_softcap>(ctx, dst);
@@ -1221,7 +1239,7 @@ static void launch_fattn_tile_switch_ncols2(ggml_backend_cuda_context & ctx, ggm
}
}
if constexpr (DKQ <= 512) {
if constexpr (DKQ <= 512 && DKQ != 320) {
if (use_gqa_opt && gqa_ratio % 8 == 0) {
launch_fattn_tile_switch_ncols1<DKQ, DV, 8, use_logit_softcap>(ctx, dst);
return;
@@ -1275,5 +1293,6 @@ extern DECL_FATTN_TILE_CASE( 96, 96);
extern DECL_FATTN_TILE_CASE(112, 112);
extern DECL_FATTN_TILE_CASE(128, 128);
extern DECL_FATTN_TILE_CASE(256, 256);
extern DECL_FATTN_TILE_CASE(320, 256);
extern DECL_FATTN_TILE_CASE(512, 512);
extern DECL_FATTN_TILE_CASE(576, 512);