diff options
| author | Soumith Chintala <soumith@gmail.com> | 2015-03-16 00:50:44 +0300 |
|---|---|---|
| committer | Soumith Chintala <soumith@gmail.com> | 2015-03-16 00:50:44 +0300 |
| commit | ae4e4144374623cdf91e75d6c55ae4df91c71ff8 (patch) | |
| tree | 210889303bf742509da5f9961ec8d1a9e7b07465 /README.md | |
| parent | bca7790186ef354dc51e46ce6337efd581fa430e (diff) | |
Update README.md
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 2 |
1 files changed, 1 insertions, 1 deletions
@@ -17,7 +17,7 @@ Modules are API compatible their [`nn`](https://github.com/torch/nn) equivalents -- All inputs have to be 3D or 4D(batch-mode), except ReLU, Tanh and Sigmoid cudnn.SpatialConvolution(nInputPlane, nOutputPlane, kW, kH, dW, dH, padW, padH) cudnn.SpatialMaxPooling(kW, kH, dW, dH, padW, padH) -cudnn.SpatialAveragePooling(kW, kH, dW, dH) -- compared to nn, they are off by a scaling factor of (1/(kW * kH)). +cudnn.SpatialAveragePooling(kW, kH, dW, dH) -- the pointwise functions take an additional optional argument. if inplace=true then they do operations in-place without using any extra memory for themselves cudnn.ReLU(inplace[=false]) |