jeong-tae / snip-filter-level-pytorch Goto Github PK
View Code? Open in Web Editor NEWEvaluate sensitivity of channel/vector connection to decrease width/depth of network. Which will accelerate inference speed and reduce storage usage without any other module or support