Skip to content

whisper: validate get_rows support for cpu extra buffer #3323

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
merged 1 commit into from
Jul 14, 2025
Merged

whisper: validate get_rows support for cpu extra buffer #3323

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
15 changes: 11 additions & 4 deletions src/whisper.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1438,7 +1438,8 @@ static bool weight_buft_supported(const whisper_hparams & hparams, ggml_tensor *
op_supported = true;
} else {
switch (op) {
// The current extra_buffer_type implementations only support GGML_OP_MUL_MAT
// The current extra_buffer_type implementations only support GGML_OP_MUL_MAT and GGML_OP_GET_ROWS
case GGML_OP_GET_ROWS:
case GGML_OP_MUL_MAT: {
ggml_init_params params = {
/*.mem_size =*/ 2 * ggml_tensor_overhead(),
Expand All @@ -1454,9 +1455,15 @@ static bool weight_buft_supported(const whisper_hparams & hparams, ggml_tensor *

ggml_tensor * op_tensor = nullptr;

int64_t n_ctx = hparams.n_audio_ctx;
ggml_tensor * b = ggml_new_tensor_4d(ctx, GGML_TYPE_F32, w->ne[0], n_ctx, w->ne[2], w->ne[3]);
op_tensor = ggml_mul_mat(ctx, w, b);
if (op == GGML_OP_MUL_MAT) {
int64_t n_ctx = hparams.n_audio_ctx;
ggml_tensor * b = ggml_new_tensor_4d(ctx, GGML_TYPE_F32, w->ne[0], n_ctx, w->ne[2], w->ne[3]);
op_tensor = ggml_mul_mat(ctx, w, b);
} else if (op == GGML_OP_GET_ROWS) {
int64_t num_indices = 8;
ggml_tensor * indices = ggml_new_tensor_1d(ctx, GGML_TYPE_I32, num_indices);
op_tensor = ggml_get_rows(ctx, w, indices);
}

// create a temporary dummy buffer for the weight so that supports_op can check the buffer type
GGML_ASSERT(w->buffer == nullptr);
Expand Down
Loading