llama.cpp
ggml : group all experts in a single ggml_mul_mat_id
#6505
Merged

Loading