• Fenglei's avatar
    cuda optimize softmax (#1310) · 154dc47a
    Fenglei authored
    * Updated softmax.
    
    * Formatting.
    
    * Updated convolution.
    
    * Use build_primitive overloading. Add helper to emit type_string given a node.
    
    * Formatting.
    
    * Update ConvolutionBackpropData.
    
    * convolution backprop & max pool memory primitive cacheing (#1303)
    
    * Updated ConvolutionBackpropFilters.
    * Update MaxPool.
    
    * Update Max and Min. (#1307)
    
    * softmax optimization
    
    * fix bug
    
    * fix bugs
    
    * clang format
    
    * remove comments
    
    * add softmax divide
    
    * fix bugs
    
    * fix bug
    
    * fix bug
    
    * clang format
    
    * remove unused header
    
    * register
    
    * using single parameters instead of array
    
    * using build_elementwise instead of build_elementwise_collective
    
    * remove workspace as csullivan suggested
    154dc47a
Name
Last commit
Last update
..
cpu Loading commit data...
gpu Loading commit data...
intelgpu Loading commit data...
interpreter Loading commit data...
reference Loading commit data...
CMakeLists.txt Loading commit data...
aligned_buffer.cpp Loading commit data...
aligned_buffer.hpp Loading commit data...
backend.cpp Loading commit data...
backend.hpp Loading commit data...
host_tensor_view.cpp Loading commit data...
host_tensor_view.hpp Loading commit data...
performance_counter.hpp Loading commit data...
tensor_view.cpp Loading commit data...
tensor_view.hpp Loading commit data...