Skip to content

Fix torch 1.8 adpater npu bug #2970

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
zhouzaida merged 1 commit into from
Oct 20, 2023
Merged

Fix torch 1.8 adpater npu bug #2970

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
1 change: 0 additions & 1 deletion mmcv/ops/csrc/common/pytorch_npu_helper.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,6 @@
#ifndef PYTORCH_NPU_HELPER_HPP_
#define PYTORCH_NPU_HELPER_HPP_

#include <torch_npu/csrc/aten/CustomFunctions.h>
#include <torch_npu/csrc/framework/utils/CalcuOpUtil.h>
#include <torch_npu/csrc/framework/utils/OpAdapter.h>

Expand Down
16 changes: 14 additions & 2 deletions mmcv/ops/csrc/pytorch/npu/focal_loss_npu.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,8 +99,20 @@ void softmax_focal_loss_forward_npu(Tensor input, Tensor target, Tensor weight,
c10::SmallVector<int64_t, 2> offsets = {0, 0};
c10::SmallVector<int64_t, 2> sizes = {n_batch, 1};
at::IntArrayRef offset = at::IntArrayRef(offsets);
at::IntArrayRef size = at::IntArrayRef(sizes);
at_npu::native::custom_ops::npu_slice_out(op_output, offset, size, output);
at::IntArrayRef size_array = at::IntArrayRef(sizes);
c10::SmallVector<int64_t, N> output_size;
for (uint64_t i = 0; i < size_array.size(); i++) {
output_size.emplace_back(size_array[i]);
}
at::Tensor result = at::empty(output_size, op_output.options());
c10::SmallVector<int64_t, N> offsetVec = array_to_small_vector(offset);
c10::SmallVector<int64_t, N> sizeVec = array_to_small_vector(size_array);
cmd.Name("Slice")
.Input(op_output)
.Input(offsetVec)
.Input(sizeVec)
.Output(output)
.Run();
}

void softmax_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight,
Expand Down