Skip to content

Expose mxfp8 grouped gemm in torch ops #4820

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

Closed
wants to merge 1 commit into from
Closed

Expose mxfp8 grouped gemm in torch ops #4820

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
11 changes: 11 additions & 0 deletions fbgemm_gpu/experimental/gen_ai/src/quantize/include/fbgemm_gpu/torch_ops.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,17 @@ at::Tensor f8f8bf16_rowwise_grouped_mm(
std::optional<at::Tensor> offsets,
at::Tensor& output);

#else

// Torch compliant MXFP8 grouped GEMM only on CUDA for now.
at::Tensor mx8mx8bf16_grouped_mm(
at::Tensor XQ,
at::Tensor WQ,
at::Tensor x_scale,
at::Tensor w_scale,
at::Tensor offsets,
std::optional<at::Tensor> output = std::nullopt);

#endif

} // namespace fbgemm_gpu
7 changes: 0 additions & 7 deletions fbgemm_gpu/experimental/gen_ai/src/quantize/quantize.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,13 +64,6 @@ at::Tensor f4f4bf16_grouped_stacked(
std::optional<at::Tensor> global_scale = std::nullopt,
std::optional<at::Tensor> starting_row_after_padding = std::nullopt,
bool use_mx = true);
at::Tensor mx8mx8bf16_grouped_mm(
at::Tensor XQ,
at::Tensor WQ,
at::Tensor x_scale,
at::Tensor w_scale,
at::Tensor offsets,
std::optional<at::Tensor> output = std::nullopt);
at::Tensor f8f8bf16(
at::Tensor XQ,
at::Tensor WQ,
Expand Down
Loading