Jan Thomas Müller

This method is unreliable. Using the `prune` method doesn't consistently prune a group; sometimes nothing gets pruned. Therefore, an alternative solution is needed. Currently, I iterate over all parameters before...

Is this the case for cpu and cuda? If it is only on cuda you may want to check out NVIDIAs optimization guide: https://docs.nvidia.com/deeplearning/performance/index.html#optimizing-performance Maybe some performance issues regarding tensor...

The pruner has a built-in safeguard to prevent layer collapse, which is probably the main issue here. If you try to prune all channels from a layer (or several layers),...

> There will be a ~5 mAP drop for DLA without DCN. Are you referring to the ctdet_coco_dlav0_1x model or another one?

Hey, really looking forward to this getting merged! I ran a quick test and found a potential issue with the current use of `with_suffix('.md')` in `_handle_batch_processing` . It replaces the...