centernet_prune
centernet_prune copied to clipboard
How does the video memory(cuda memory ) change?
Thank you for your outstanding work!
original model: 204M ap50: 0.49
first prune model: 139M ap50: 0.48
second prune model: 96M ap50: 0.47
How does the video memory(cuda memory ) change?
Hello, this work has been done for a long time. I forgot some parameters, but obviously, cuda memory will be reduced a lot, especially in centernet_model train and inference. After you finish pruning, you can use a larger batch_size which is almost multiplied for training. Of course, this is only the number of parameters.
Thank you for your reply!
this prune version only support res101_no_dcn backbone?
This version only support resnet101, dcn and no_dcn is ok, dcn modue added at backbone last in centernet, dcn does not affect pruning
This version only support resnet101, dcn and no_dcn is ok, dcn modue added at backbone last in centernet, dcn does not affect pruning
Hi,author. This project is very helpful to me, thank you very much. In addition, I would like to ask that slimming pruning has been updated in the pruning method?looking forward your answer!