Jetson inference Hello AI world detectnet train_ssd.py error

File “train_ssd.py”, line 309, in
net.init_from_pretrained_ssd(args.pretrained_ssd)
File “/jetson-inference/python/training/detection/ssd/vision/ssd/ssd.py”, line 119, in init_from_pretrained_ssd
state_dict = torch.load(model, map_location=lambda storage, loc: storage)
File “/usr/local/lib/python3.6/dist-packages/torch/serialization.py”, line 608, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File “/usr/local/lib/python3.6/dist-packages/torch/serialization.py”, line 777, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
EOFError: Ran out of input

Hi,

It looks like there are some issues when opening the pickle file.
Do you follow the instructions below to run the retraining?

Could you share more details about your steps with us?

Thanks.

It would seem that the base model is corrupted - can you try downloading it again?

cd jetson-inference/python/training/detection/ssd
wget https://nvidia.box.com/shared/static/djf5w54rjvpqocsiztzaandq1m3avr7c.pth -O models/mobilenet-v1-ssd-mp-0_675.pth

Hi Dusty,

Thank you for your help. Yes, indeed the model wasn’t downloaded properly. I found another source to download since the box link isn’t working in China.

Simon

发自我的iPad

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.