Skip to content

N
ngraph
Commit d38aba91 authored by Fenglei's avatar Fenglei Committed by Robert Kimball
Browse files
Options

add find algorithm for convolution without extra padding (#1710)

parent 2aa7899c
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 6 additions and 3 deletions
src/ngraph/runtime/gpu/cudnn_emitter.cpp
View file @ d38aba91
This diff is collapsed.
src/ngraph/runtime/gpu/cudnn_emitter.hpp
View file @ d38aba91
......@@ -76,7 +76,8 @@ namespace ngraph
const Shape& output_tensor_shape,
const Strides& window_movement_strides,
const Strides& window_dilation_strides,
const Shape& padding_below);
const Shape& padding_below,
const bool find_algo = false);
size_t build_convolution_backward_data(const std::string& dtype,
const Shape& input_filter_shape,
......@@ -84,7 +85,8 @@ namespace ngraph
const Shape& output_tensor_shape,
const Strides& window_movement_strides,
const Strides& window_dilation_strides,
const Shape& padding_below);
const Shape& padding_below,
const bool find_algo = false);
size_t build_convolution_backward_filter(const std::string& dtype,
const Shape& input_tensor_shape_0,
......@@ -92,7 +94,8 @@ namespace ngraph
const Shape& output_filter_shape,
const Strides& window_movement_strides,
const Strides& window_dilation_strides,
const Shape& padding_below);
const Shape& padding_below,
const bool find_algo = false);
size_t build_reduce_forward(const cudnnReduceTensorOp_t& reduce_op,
const std::string& dtype,
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment