ModuleNotFoundError: No module named 'run_vot'

[Luc Stiemer]

My machine:
Ubuntu 20.04
RTX 3090 GPU
CUDA 11
pytorch 1.7.1

I installed your tracker using the install.sh just changing the pytorch line to the following, since my gpu does only support cuda 11:

conda install pytorch torchvision torchaudio cudatoolkit=11.0 -c pytorch

There were no errors during the installation. Now, running the vot test RLT_DiMP command, im facing the error:

Checking for new version
Scanning registry /home/linx123-rtx/vot-workspace/trackers.ini
Found 12 trackers
Generating dummy sequence
Obtaining runtime for tracker RLT_DiMP
Initializing tracker (1/3)
Running process: /home/linx123-rtx/anaconda3/envs/RLT_DiMP/bin/python -c "import sys;sys.path.insert(0, '/home/linx123-rtx/vot-workspace/RLT_DiMP/atom/pytracking # you should change directory'); import run_vot; run_vot.run_vot2020_LT('RLT_dimp', 'new', None, 0) # 0 means gpu_id"
Traceback (most recent call last):
File "", line 1, in
ModuleNotFoundError: No module named 'run_vot'
Error during tracker execution: Unable to connect to tracker

I tried python 3.6,3.7 and 3.8. I also tried other gcc versions (e.g. the 7.5 which was suggested by you) but nothing worked. Could you please help me with this?

Kind regards

[Luc Stiemer]

issue closed, I had to remove "# you should change directory" from the tracker.ini.

Further, I had to add
pad = tuple(map(int,pad))
in functional.py and add in line 3488 of torch/nn since I was facing the error:

replication_pad2d(): argument 'padding' must be tuple of ints, but found element of type float at pos 1