Skip to content

[CANN]MUL_MAT optimization #12382

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 2 commits into from
Mar 15, 2025
Merged

[CANN]MUL_MAT optimization #12382

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
531b3a8
[CANN]MUL_MAT optimization
noemotiovon Mar 14, 2025
3d5b3a7
code adjustment
noemotiovon Mar 14, 2025
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
9 changes: 7 additions & 2 deletions ggml/src/ggml-cann/aclnn_ops.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2790,10 +2790,15 @@ static void ggml_cann_mul_mat_quant(ggml_backend_cann_context& ctx,
(char*)output_buffer + batch1 * output_stride, ACL_FLOAT16,
output_elem_size, output_ne, output_nb, 2, ACL_FORMAT_ND,
output_ne_offset);

int64_t antiquantGroupSize = 0;
if (src0->ne[0] > QK8_0) {
antiquantGroupSize = QK8_0;
}

ACL_CHECK(aclnnWeightQuantBatchMatmulV2GetWorkspaceSize(
acl_input_tensor, acl_weight_tensor, acl_scale_tensor, nullptr,
nullptr, nullptr, nullptr, QK8_0, acl_output_tensor,
nullptr, nullptr, nullptr, antiquantGroupSize, acl_output_tensor,
&workspaceSize, &executor));
if (workspaceAddr == nullptr) {
workspaceAddr = workspace_allocator.alloc(workspaceSize);
Expand Down Expand Up @@ -2833,7 +2838,7 @@ static void ggml_cann_mul_mat_quant(ggml_backend_cann_context& ctx,

ACL_CHECK(aclnnWeightQuantBatchMatmulV2GetWorkspaceSize(
acl_input_tensor, acl_weight_tensor, acl_scale_tensor,
nullptr, nullptr, nullptr, nullptr, QK8_0,
nullptr, nullptr, nullptr, nullptr, antiquantGroupSize,
acl_output_tensor, &workspaceSize, &executor));
ACL_CHECK(aclnnWeightQuantBatchMatmulV2(
workspaceAddr, workspaceSize, executor, ctx.stream()));
Expand Down
5 changes: 0 additions & 5 deletions ggml/src/ggml-cann/ggml-cann.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1689,11 +1689,6 @@ static bool ggml_backend_cann_supports_op(ggml_backend_dev_t dev,
case GGML_OP_MUL_MAT: {
switch (op->src[0]->type) {
case GGML_TYPE_Q8_0:
// Current groupsize should not be greater than k-1 in
// aclnnWeightQuantBatchMatmulV2GetWorkspaceSize
if (op->src[0]->ne[0] <= QK8_0) {
return false;
}
case GGML_TYPE_F16:
case GGML_TYPE_F32:
case GGML_TYPE_Q4_0:
Expand Down
Loading