mozempk/ggml-org-llama-cpp-mirror
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-11-11 22:32:07 +01:00
Code Issues Packages Projects Releases Wiki Activity

vulkan: Allow non-pow2 n_experts in topk_moe (#17872)

Browse Source
This commit is contained in:
Jeff Bolz
2025-12-13 08:40:04 +01:00
committed by GitHub
parent 2bc94e7928
commit 07a10c1090
3 changed files with 27 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-vulkan/vulkan-shaders/topk_moe.comp
+9 -3
View File
@@ -10,6 +10,7 @@
layout (push_constant) uniform parameter
{
uint n_rows;
uint n_experts_push;
uint n_expert_used;
float clamp_min;
float clamp_max;
@@ -18,11 +19,16 @@ layout (push_constant) uniform parameter
layout(local_size_x_id = 0, local_size_y = 4, local_size_z = 1) in;
layout(constant_id = 0) const uint WARP_SIZE = 32;
layout(constant_id = 1) const uint n_experts = 512;
layout(constant_id = 1) const uint n_experts_spec = 512;
layout(constant_id = 2) const bool with_norm = true;
layout(constant_id = 3) const bool late_softmax = false;
layout(constant_id = 4) const bool nexperts_use_push = false;
const uint experts_per_thread = (n_experts > WARP_SIZE) ? n_experts / WARP_SIZE : 1;
uint n_experts = nexperts_use_push ? n_experts_push : n_experts_spec;
#define CEIL_DIV(a, b) (((a) + (b) - 1) / (b))
const uint experts_per_thread = CEIL_DIV(n_experts_spec, WARP_SIZE);
layout (binding = 0, std430) readonly buffer Logits {float logits[];};
layout (binding = 1, std430) writeonly buffer Weights {float weights[];};
@@ -94,7 +100,7 @@ void main() {
}
if (!late_softmax) {
softmax_warp_inplace(wt, n_experts, lane, false);
softmax_warp_inplace(wt, n_experts, lane, nexperts_use_push);
}
// at this point, each thread holds a portion of softmax,