#include "op_plugin/AclOpsInterface.h"
#include "op_plugin/utils/OpAdapter.h"
namespace acl_op {
using npu_preparation = at_npu::native::OpPreparation;
namespace {
at::Tensor &conv3d_backward_input_nocheck(at::Tensor &grad_input, const at::Tensor &input, const at::Tensor &grad,
const at::Tensor &weight, at::IntArrayRef stride, at::IntArrayRef padding,
at::IntArrayRef dilation, int64_t groups)
{
TORCH_CHECK(stride.size() >= 3, "stride has to contain more than 3 elements, but got ", stride.size(),
OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(padding.size() >= 3, "padding has to contain more than 3 elements, but got ", padding.size(),
OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(dilation.size() >= 3, "dilation has to contain more than 3 elements, but got ", dilation.size(),
OPS_ERROR(ErrCode::PARAM));
c10::SmallVector<int64_t, N> strides_size = {1, 1, stride[0], stride[1], stride[2]};
c10::SmallVector<int64_t, N> paddings = {padding[0], padding[0], padding[1], padding[1], padding[2], padding[2]};
c10::SmallVector<int64_t, N> dilations = {1, 1, dilation[0], dilation[1], dilation[2]};
at::IntArrayRef input_size = input.sizes();
at::Tensor weight_cast = at_npu::native::custom_ops::_npu_dtype_cast(weight, grad.scalar_type());
at_npu::native::OpCommand cmd;
cmd.Name("Conv3DBackpropInput")
.Input(input_size, at::kInt)
.Input(weight_cast, "filter")
.Input(grad, "out_backprop")
.Output(grad_input, "y")
.Attr("strides", strides_size)
.Attr("pads", paddings)
.Attr("dilations", dilations)
.Attr("groups", groups)
.Attr("data_format", (string) "NCDHW")
.Run();
return grad_input;
}
at::Tensor &conv3d_backward_weight_nocheck(at::Tensor &grad_weight, const at::Tensor &input, const at::Tensor &grad,
const at::Tensor &weight, at::IntArrayRef stride, at::IntArrayRef padding,
at::IntArrayRef dilation, int64_t groups)
{
TORCH_CHECK(stride.size() >= 3, "stride has to contain more than 3 elements, but got ", stride.size(),
OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(padding.size() >= 3, "padding has to contain more than 3 elements, but got ", padding.size(),
OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(dilation.size() >= 3, "dilation has to contain more than 3 elements, but got ", dilation.size(),
OPS_ERROR(ErrCode::PARAM));
c10::SmallVector<int64_t, N> strides_size = {1, 1, stride[0], stride[1], stride[2]};
c10::SmallVector<int64_t, N> paddings = {padding[0], padding[0], padding[1], padding[1], padding[2], padding[2]};
c10::SmallVector<int64_t, N> dilations = {1, 1, dilation[0], dilation[1], dilation[2]};
at::IntArrayRef input_size = weight.sizes();
at_npu::native::OpCommand cmd;
cmd.Name("Conv3DBackpropFilter")
.Input(input, "x")
.Input(input_size, at::kInt)
.Input(grad, "out_backprop")
.Output(grad_weight, "y")
.Attr("strides", strides_size)
.Attr("pads", paddings)
.Attr("dilations", dilations)
.Attr("groups", groups)
.Attr("data_format", (string) "NCDHW")
.Run();
return grad_weight;
}
at::Tensor &conv3d_backward_bias_nocheck(at::Tensor &grad_bias, const at::Tensor &input, const at::Tensor &grad,
const at::Tensor &weight, at::IntArrayRef stride, at::IntArrayRef padding,
at::IntArrayRef dilation, int64_t groups)
{
TORCH_CHECK(input.dim() >= 3, "input has to be more than 3D, but got Tensor of dimension ", input.dim(),
OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(grad.dim() >= 3, "grad has to be more than 3D, but got Tensor of dimension ", grad.dim(),
OPS_ERROR(ErrCode::PARAM));
if (input.numel() == input.size(0) * input.size(1) * input.size(2)) {
at::Tensor grad_view = grad.contiguous().view({grad.size(0), grad.size(1), grad.size(2)});
acl_op::sum_out(grad_view, c10::SmallVector<int64_t, N>{0}, false, grad_view.scalar_type(), grad_bias);
} else {
at::Tensor grad_view = grad.contiguous().view({grad.size(0), grad.size(1), grad.size(2), -1});
acl_op::sum_out(grad_view, c10::SmallVector<int64_t, N>{0, 2, 3}, false, grad_view.scalar_type(), grad_bias);
}
return grad_bias;
}
}
std::tuple<at::Tensor, at::Tensor, at::Tensor> npu_conv3d_backward(const at::Tensor &input, const at::Tensor &grad,
const at::Tensor &weight, at::IntArrayRef stride,
at::IntArrayRef padding, at::IntArrayRef dilation,
int64_t groups, std::array<bool, 3> grad_input_mask)
{
at::Tensor grad_input;
at::Tensor grad_weight;
at::Tensor grad_bias;
if (grad_input_mask[0]) {
grad_input = npu_preparation::apply_tensor_with_format(input, ACL_FORMAT_NDC1HWC0);
conv3d_backward_input_nocheck(grad_input, input, grad, weight, stride, padding, dilation, groups);
}
if (grad_input_mask[1]) {
grad_weight = npu_preparation::apply_tensor_with_format(weight.sizes(), weight.options().dtype(at::kFloat),
npu_preparation::get_tensor_npu_format(weight));
conv3d_backward_weight_nocheck(grad_weight, input, grad, weight, stride, padding, dilation, groups);
}
if (grad_input_mask[2]) {
grad_bias = npu_preparation::apply_tensor_with_format({grad.size(1)}, grad.options(), ACL_FORMAT_NCHW);
conv3d_backward_bias_nocheck(grad_bias, input, grad, weight, stride, padding, dilation, groups);
}
return std::tie(grad_input, grad_weight, grad_bias);
}
}