search

CoCoPIE-Pruning / CoCoPIE-ModelZoo

MIT License
124 stars 46 forks source link

How fast to depoly in Nvidia Jetson TX2 #3

Open zhucheng725 opened 4 years ago

zhucheng725 commented 4 years ago

Can this repo be depolyed in jetson tx2? How fast in it Many thanks

MuhammadAsadJaved commented 4 years ago

@zhucheng725 I have the same question. But seems like they did not provide complete project yet. They only provide test code, no train and demo etc codes yet. Please let me know if you are planning to do by yourself. I have tried to use the weights and cfg for demo using (https://github.com/ultralytics/yolov3) but it does not work.

zhucheng725 commented 4 years ago

@MuhammadAsadJaved Hey bro, Although I have not tested this repo, it looks like that it can be used yolov4 not v3. So maybe you can try to download weights and cfgs from here:https://drive.google.com/drive/folders/1FcWdXcWc3vScV-guIrxWsWGhjQwPOEQW

MuhammadAsadJaved commented 4 years ago

@zhucheng725 I know your meaning. I am using the same weights. My meaning is this project is build based on the PyTorch Yolov3 (https://github.com/ultralytics/yolov3). So I have tried to test these weights and .cfg with the PyTorch Yolov3 project but It does not work. Now I have checked the code and it's very different from the original project, so we can not use the demo, training, and test codes of that project for this project. So we can only wait for the author/developer to update this project for model training and demo codes.

MuhammadAsadJaved commented 3 years ago

@zhucheng725 Have you tested this code for demo or not? Is there any progress?