https://github.com/Microsoft/CNTK
Raw File
Tip revision: 2287f5ab5f2aec19d87ecb3d76649610c878019e authored by Bowen Bao on 25 July 2018, 20:42:31 UTC
change default to not pad for channel in convolution.
Tip revision: 2287f5a
current_iteration.md
# CNTK Current Iteration

## Efficient group convolution
The implementation of group convolution in CNTK has been updated. The updated implementation moves away from creating a sub-graph for group convolution (using slicing and splicing), and instead uses cuDNN7 and MKL2017 APIs directly. This improves the experience both in terms of performance and model size. 

As an example, for a single group convolution op with the following attributes:

- Input tensor (C, H, W) = (32, 128, 128)
- Number of output channels = 32 (channel multiplier is 1)
- Groups = 32 (depth wise convolution)
- Kernel size = (5, 5)

The comparison numbers for this single node are as follows:

| First Header  | GPU exec. time (in millisec., 1000 run avg.) | CPU exec. time (in millisec., 1000 run avg.) | Model Size (in KB, CNTK format)
| ------------- | ------------- | ------------- | ------------- |
| Old implementation  | 9.349  | 41.921  | 38  |
| New implementation  | 6.581  | 9.963  | 5  |
| Speedup/savings	Approx.  | 30%	Approx.  | 65-75%	Approx.  | 87% |

## Operators
### depth_to_space and space_to_depth
There is a breaking change in the **depth_to_space** and **space_to_depth** operators. These have been updated to match ONNX specification, specifically
the permutation for how the depth dimension is placed as blocks in the spatial dimensions, and vice-versa, has been changed. Please refer to the updated doc
examples for these two ops to see the change.


## Bug fixes


## ONNX
### Updates
- Updated CNTK's ONNX BatchNormalization op export/import to latest spec.

### Bug or minor fixes:


## Misc

back to top