Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Faster AVX2 matrix multiplications for MoE models #428

Merged
merged 3 commits into from
May 21, 2024
Merged

Faster AVX2 matrix multiplications for MoE models #428

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
473b96d
Adding MOE support: initial commit
Kawrakow May 20, 2024
dbb04fd
Adding MOE support: seems to be working
Kawrakow May 20, 2024
d292ca0
Address PR comment
Kawrakow May 21, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions llama.cpp/ggml.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10642,6 +10642,20 @@ static void ggml_compute_forward_mul_mat_id(
const int64_t nr0 = ne01; // src0 rows
const int64_t nr1 = cne1; // src1 rows

if ((vec_dot_type == GGML_TYPE_Q8_K || vec_dot_type == GGML_TYPE_Q8_0 ||
vec_dot_type == GGML_TYPE_Q8_1) && dst->type == GGML_TYPE_F32) {
if (ne13 == 1) {
if (!llamafile_mixmul_iqk(nr0, nr1, ne00, ne11, src0->type,
(const char *)src0_cur,
(const char *)wdata,
(float *)dst->data, nb1, nb2,
matrix_rows + cur_a*ne12,
ith, nth)) goto IQK_MulMat_Not_Available;
continue;
}
}
IQK_MulMat_Not_Available:;

// distribute the thread work across the inner or outer loop based on which one is larger

const int64_t nth0 = nr0 > nr1 ? nth : 1; // parallelize by src0 rows
Expand Down
Loading