Repository for the Special Course in Deep Learning in Computer Vision @DTU
This repo will include 3 of the projects developed:
- P1_1: Image Classification: HotDog/NoHotDog
- P1_2: Object Detection: RCNN Waste Classification using the TACO dataset.
- P_2: Generative Adversarial Networks
- Open your terminal and write:
- Login with your credentials
- Start up an interactive node using
voltash
- Create a directory in the HPC where you want your venvs to be stored
mkdir venvs - Check the available modules to load
module avail - Load the python module:
module load python3/3.10.7 - Create the venv:
python3 -m venv NAME_VENV - You can now activate your environment wherever using:
source NAME_VENV/bin/activate
Make sure you have installed it in your env. If not do
pip install notebook
- Start it inside your interactive node
jupyter notebook --no-browser --port=40000 --ip=$HOSTNAME
- Copy the values in your url that look like this:
n-62-20-1:40001
- Open another terminal and write. Please note substitute USER by your username and make sure the above line is included after L8080:
ssh [email protected] -g -L8080:n-62-20-1:40001 -N
- Enter your credentials
- It will look as if nothing happened, but open your browser and write
http://localhost:8080/tree?
Once you have this text file (the job script, let us call it submit.sh), you must submit it by typing in a terminal the command (Source: https://www.hpc.dtu.dk/?page_id=1416):
bsub < submit.sh
Job Statistics:
bstat