Skip to content

Latest commit

 

History

History
8 lines (6 loc) · 459 Bytes

README.md

File metadata and controls

8 lines (6 loc) · 459 Bytes

Toolbox

Some handy (matlab or python) tools.

TODO

  • Add a python tool (for Caffe model) to merge statistic (i.e. mean, variance, scale and shift) of BN Layers to Conv Layers for faster inference during TESTING phase.
  • Add a python tool to prune well-trained model for faster inference with less GPU memory usage.
  • Add a python tool to convert a pretrained pytorch model to use in inference in Caffe.