Closed valentinitnelav closed 1 year ago
I tried pip install pytz
in the environment, but this didn't solve the issue. I get the same error message when I try the demo.
# Update the git repo
cd ~/PAI/detectorscd ~/PAI/detectors/
git pull
module purge
module load PyTorch/1.10.0-foss-2021a-CUDA-11.3.1
module load TensorFlow/2.6.0-foss-2021a-CUDA-11.3.1
module load torchvision/0.11.1-foss-2021a-CUDA-11.3.1
source ~/venv/detectron2/bin/activate
pip install detectron2 -f \
https://dl.fbaipublicfiles.com/detectron2/wheels/cu113/torch1.10/index.html
# Didn't get any error message; only a series of `Requirement already satisfied` messages for each dependency.
pip install pytz
Related issues:
In this initial stage of the PAI project, we decided to not pursue further the employment of detectron2. detectron2 looks to be a bit heavier. Because we would like to deploy the trained weights on an edge device, then we need to focus on light models.
FYI: I will also get the help of the cluster team from Uni Leipzig because like ScaledYOLOv4, there are some dependency issues and certain modules need to be loaded from the software tree.
The environment looks like this for now: