#include "op_plugin/AclOpsInterface.h"
#include "op_plugin/OpApiInterface.h"
#include "op_plugin/utils/op_api_common.h"
const int DIM_TWO = 2;
const int DIM_THREE = 3;
namespace op_api {
using npu_preparation = at_npu::native::OpPreparation;
using tensor_list = std::tuple<at::Tensor, at::Tensor, at::Tensor>;
tensor_list npu_moe_gating_top_k_softmax_v2 (const at::Tensor &x, int64_t k, const c10::optional<at::Tensor> &finished_opt,
const c10::optional<int64_t> renorm_opt, const c10::optional<bool> softmax_flag_opt)
{
TORCH_CHECK(x.dim() == DIM_TWO or x.dim() == DIM_THREE, "The x's shape should be 2D or 3D", OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(x.scalar_type() == at::kHalf || x.scalar_type() == at::kFloat || x.scalar_type() == at::kBFloat16,
"float16, float32 or bfloat16 tensor expected but got a tensor with dtype: ",
x.scalar_type(), OPS_ERROR(ErrCode::PARAM));
auto x_size = x.sizes();
TORCH_CHECK(k >= 0 and k <= x_size[x.dim() - 1], "The k's shape should be in [0, ", x_size[x.dim() - 1], "]", OPS_ERROR(ErrCode::PARAM));
int64_t renorm = c10::value_or_else(renorm_opt, [] {return 0;});
TORCH_CHECK(renorm == 0 || renorm == 1, "renorm must be 0 or 1, but got: ", renorm, OPS_ERROR(ErrCode::PARAM));
bool softmax_result_flag = c10::value_or_else(softmax_flag_opt, [] {return false; });
const at::Tensor &finished = c10::value_or_else(finished_opt, [] {return at::Tensor(); });
if (finished.defined()) {
TORCH_CHECK(finished.scalar_type() == at::kBool, "bool tensor expected but got a tensor with dtype: ", finished.scalar_type(), OPS_ERROR(ErrCode::PARAM));
auto finished_size = finished.sizes();
TORCH_CHECK((x.dim() - 1) == finished.dim(), "x.dim() should be 1 more than finished.dim().", OPS_ERROR(ErrCode::PARAM));
TORCH_CHECK(x_size[0] == finished_size[0], "Input rows should be same as finished rows.", OPS_ERROR(ErrCode::PARAM));
if (x.dim() == DIM_THREE) {
TORCH_CHECK(x_size[1] == finished_size[1], "Input rows should be same as finished rows.", OPS_ERROR(ErrCode::PARAM));
}
}
at::Tensor y;
at::Tensor expert_idx;
at::Tensor softmax_result;
if (x.dim() == DIM_THREE) {
y = npu_preparation::apply_tensor_without_format({x_size[0], x_size[1], k}, x.options());
expert_idx = npu_preparation::apply_tensor_without_format(
{x_size[0], x_size[1], k}, x.options().dtype(at::kInt));
} else {
y = npu_preparation::apply_tensor_without_format({x_size[0], k}, x.options());
expert_idx = npu_preparation::apply_tensor_without_format(
{x_size[0], k}, x.options().dtype(at::kInt));
}
bool softmaxFlag = (renorm == 0) && softmax_result_flag;
if (softmaxFlag) {
if (x.dim() == DIM_THREE) {
softmax_result = npu_preparation::apply_tensor_without_format(
{x_size[0], x_size[1], x_size[2]}, x.options().dtype(at::kFloat));
} else {
softmax_result = npu_preparation::apply_tensor_without_format(
{x_size[0], x_size[1]}, x.options().dtype(at::kFloat));
}
} else {
softmax_result = npu_preparation::apply_tensor_without_format({0}, x.options().dtype(at::kFloat));
}
if (k == 0) {
return std::tie(y, expert_idx, softmax_result);
}
for (int32_t i = 0; i < x.dim(); i++) {
if (x_size[i] == 0) {
return std::tie(y, expert_idx, softmax_result);
}
}
EXEC_NPU_CMD(aclnnMoeGatingTopKSoftmaxV2, x, finished, k, renorm, softmaxFlag, y, expert_idx, softmax_result);
return std::tie(y, expert_idx, softmax_result);
}
}