Skip to content

Commit 2a6bab5

Browse files
0cc4mggerganov
authored andcommitted
Vulkan Mixture of Experts (MoE) support (llama/7628)
* Finish Vulkan mul_mat_id implementation * Add Vulkan sum_rows and div ops * Fix MUL_MAT_ID matrix matrix shader * Fix MUL_MAT_ID matrix vector shader dispatch size * Fix MUL_MAT_ID matrix vector shader and dispatch code * Update Vulkan CPU offload for MUL_MAT_ID * Fix crash when using split mode none and setting a main GPU
1 parent 8c01c9b commit 2a6bab5

File tree

1 file changed

+448
-315
lines changed

1 file changed

+448
-315
lines changed

0 commit comments

Comments
 (0)