Pruning on standard darknet model and tensorflow model so that it can be used for deepstream

Description

Hi,
I have a Yolo darknet model and tensorflow model and I want to prune these model and test it on DeepStream SDK 5.0. Can you kindly help by giving the steps or some kind of reference.

Environment

TensorRT Version: 7.0
GPU Type: dGPU
Nvidia Driver Version: 440
CUDA Version: 10.2
CUDNN Version:
Operating System + Version:
Python Version (if applicable):
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):

Relevant Files

Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)

Steps To Reproduce

Please include:

  • Exact steps/commands to build your repro
  • Exact steps/commands to run your repro
  • Full traceback of errors encountered

Hi @GalibaSashi,
Network pruning is out of DeepStream support scope.
You could try TLT Yolov3 which has been pruned by TLT, and TLT supports to prune the networks they provied based on your dataset.
Any questions about TLT can ask in TLT forum - https://forums.developer.nvidia.com/c/accelerated-computing/intelligent-video-analytics/transfer-learning-toolkit/17 .

Thanks!

Hi I actually wanted to prune my pretrained YoloV3 weights and cfg model so to do so in TLT requires us to train and create model.
Is there any other way.

Sorry! I didn’t get your point.
If you want to prune the model customized by yourself, sorry, it’s out of DeepStream support scope as DeepStream is help user quickly deploy their trained model for video analysis system.

Thanks!

Hi @mchi,
Actually I just want the pruned version of general model of yoloV2 and YoloV3.Do you have any solution?

Besides TLT, we don’t have other solutions.

Thanks!

Ok thanks