Labs
This page extends the study guide with more information specific to the computer labs in this course.
Sign up & submissions
You should work on the labs in pairs. For this, you need to sign up for a lab group together with your lab partner. This can be done on Webreg page depending on your course code:1
For submitting your labs, use the Lisam submissions functionality.
For feedback & grading of the labs, go to the Webreg page depending on your course code.
The sign-up deadline for Autumn 2024 has passed. If you still need to sign up for a lab group, contact the examiner.
Running the labs
Labs come in the form of Jupyter Notebooks using Python. On the lab computers, we provide a virtual environment that you should activate before starting to work on the labs:
source /courses/TDDE16/venv/bin/activate
Afterwards, you can run jupyter lab
to start working.
Using your own computer
If you want to work on your own computer, we provide a requirements.txt
file that should allow you to install the required packages locally, but note the following:
Lab 4 additionally requires gensim, which isn’t listed in the requirements file due to dependency conflicts. You may have success with installing gensim after all the other packages in the requirements file, or you may want to make a separate environment for this lab.
Lab 5 uses llama-cpp and PyTorch, but the requirements file is made for CPU-only installs of these libraries. If your computer has a GPU that you would like to use for this lab, please adapt the requirements file based on the instructions for installing PyTorch locally.
We cannot provide any technical support if you run into problems specific to your own computer. We test that labs run on the lab environment, so if you run into technical issues that you cannot solve, please use the lab computers instead.
Footnotes
It’s also fine to work together with a student from a different course code. In this case, you need to pick one of the course codes under which you want to sign up. It’s also a good idea to inform the examiner to avoid any confusion.↩︎