Skip to content

optimize and simplify ggml_ext_chunk #1084

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

Merged
leejet merged 1 commit into from
Dec 13, 2025
Merged

optimize and simplify ggml_ext_chunk #1084

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
4 changes: 3 additions & 1 deletion common.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -194,10 +194,12 @@ class GEGLU : public UnaryBlock {
auto proj = std::dynamic_pointer_cast<Linear>(blocks["proj"]);

x = proj->forward(ctx, x); // [ne3, ne2, ne1, dim_out*2]
auto x_vec = ggml_ext_chunk(ctx->ggml_ctx, x, 2, 0);
auto x_vec = ggml_ext_chunk(ctx->ggml_ctx, x, 2, 0, false);
x = x_vec[0]; // [ne3, ne2, ne1, dim_out]
auto gate = x_vec[1]; // [ne3, ne2, ne1, dim_out]

gate = ggml_cont(ctx->ggml_ctx, gate);

gate = ggml_gelu_inplace(ctx->ggml_ctx, gate);

x = ggml_mul(ctx->ggml_ctx, x, gate); // [ne3, ne2, ne1, dim_out]
Expand Down
34 changes: 11 additions & 23 deletions ggml_extend.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -732,34 +732,22 @@ __STATIC_INLINE__ struct ggml_tensor* ggml_ext_slice(struct ggml_context* ctx,
__STATIC_INLINE__ std::vector<struct ggml_tensor*> ggml_ext_chunk(struct ggml_context* ctx,
struct ggml_tensor* x,
int num,
int64_t dim) {
int64_t dim,
bool cont = true) {
GGML_ASSERT(dim >= 0 && dim < 4);
GGML_ASSERT(x->ne[dim] % num == 0);

int perm[4] = {0, 1, 2, 3};
for (int i = dim; i < 3; ++i)
perm[i] = perm[i + 1];
perm[3] = dim;

int inv_perm[4];
for (int i = 0; i < 4; ++i)
inv_perm[perm[i]] = i;

if (dim != 3) {
x = ggml_ext_torch_permute(ctx, x, perm[0], perm[1], perm[2], perm[3]);
x = ggml_cont(ctx, x);
}

std::vector<struct ggml_tensor*> chunks;
int64_t chunk_size = x->ne[3] / num;
int64_t chunk_size = x->ne[dim] / num;
int64_t stride = chunk_size * x->nb[dim];
int64_t chunk_ne[4] = {x->ne[0], x->ne[1], x->ne[2], x->ne[3]};
chunk_ne[dim] = chunk_size;
for (int i = 0; i < num; i++) {
auto chunk = ggml_view_4d(
ctx, x,
x->ne[0], x->ne[1], x->ne[2], chunk_size,
x->nb[1], x->nb[2], x->nb[3], x->nb[3] * i * chunk_size);

if (dim != 3) {
chunk = ggml_ext_torch_permute(ctx, chunk, inv_perm[0], inv_perm[1], inv_perm[2], inv_perm[3]);
chunk_ne[0], chunk_ne[1], chunk_ne[2], chunk_ne[3],
x->nb[1], x->nb[2], x->nb[3], stride * i);
if (cont) {
chunk = ggml_cont(ctx, chunk);
}
chunks.push_back(chunk);
Expand All @@ -772,7 +760,7 @@ __STATIC_INLINE__ ggml_tensor* ggml_ext_silu_act(ggml_context* ctx, ggml_tensor*
// x: [ne3, ne2, ne1, ne0]
// return: [ne3, ne2, ne1, ne0/2]

auto x_vec = ggml_ext_chunk(ctx, x, 2, 0);
auto x_vec = ggml_ext_chunk(ctx, x, 2, 0, false);
ggml_tensor* gate;
if (gate_first) {
gate = x_vec[0];
Expand All @@ -781,7 +769,7 @@ __STATIC_INLINE__ ggml_tensor* ggml_ext_silu_act(ggml_context* ctx, ggml_tensor*
x = x_vec[0];
gate = x_vec[1];
}

gate = ggml_cont(ctx, gate);
gate = ggml_silu_inplace(ctx, gate);

x = ggml_mul(ctx, x, gate); // [ne3, ne2, ne1, ne0/2]
Expand Down
Loading