vuron/adept: Платформа для обучения и вывода нейронных сетей, с API приближенным к PyTorch.

src/dispatch/poolNd.hpp

50 строк2 KB

kolkir

MaxPool2d implementation

09 фев 2025, 16:36

09 фев 2025, 16:36b44f8a7

#pragma once

#include <adept/dispatch/dispatcher.hpp>
#include <adept/param_array.hpp>
#include <adept/tensor.hpp>

namespace adept {

inline Tensor avg_pool2d_fwd(const Tensor& input,
                             const ParamArray<2>& kernel,
                             const ParamArray<2>& stride,
                             const ParamArray<2>& padding) {
  static auto& func = Dispatcher::instance().find("avg_pool2d_fwd");
  return Dispatcher::instance()
      .call<Tensor, const Tensor&, const ParamArray<2>&, const ParamArray<2>&,
            const ParamArray<2>&>(func, input, kernel, stride, padding);
}

inline Tensor avg_pool2d_bwd(const Tensor& out_grad,
                             const Shape& input_shape,
                             const ParamArray<2>& kernel,
                             const ParamArray<2>& stride,
                             const ParamArray<2>& padding) {
  static auto& func = Dispatcher::instance().find("avg_pool2d_bwd");
  return Dispatcher::instance()
      .call<Tensor, const Tensor&, const Shape&, const ParamArray<2>&, const ParamArray<2>&,
            const ParamArray<2>&>(func, out_grad, input_shape, kernel, stride, padding);
}

inline Tensor max_pool2d_fwd(const Tensor& input,
                             Tensor& indices,
                             const ParamArray<2>& kernel,
                             const ParamArray<2>& stride,
                             const ParamArray<2>& padding,
                             const ParamArray<2>& dilation) {
  static auto& func = Dispatcher::instance().find("max_pool2d_fwd");
  return Dispatcher::instance()
      .call<Tensor, const Tensor&, Tensor&, const ParamArray<2>&, const ParamArray<2>&,
            const ParamArray<2>&, const ParamArray<2>&>(func, input, indices, kernel, stride,
                                                        padding, dilation);
}

inline Tensor max_pool2d_bwd(const Tensor& out_grad,
                             const Tensor& indices,
                             const Shape& input_shape) {
  static auto& func = Dispatcher::instance().find("max_pool2d_bwd");
  return Dispatcher::instance().call<Tensor, const Tensor&, const Tensor&, const Shape&>(
      func, out_grad, indices, input_shape);
}
}  // namespace adept