Labs
There will be several labs on selected Thursdays from 14:15–16:00 in 1019.
The labs will give you practical experience with different types of data, training and fine-tuning Transformer-based models, and using Hugging Face tools and resources. Labs are not graded and are intended for practice and self-study.
Most labs are provided as Jupyter (.ipynb) notebooks, so some familiarity with Jupyter Notebook or Google Colab will be helpful. While labs are designed to be completed via Colab Notebook resources, you may also work with your own resources, use a Jupyter notebook connected to HPC, or other HPC resources. If you are not familiar with using HPC, please check this tutorial or take the short course offered by the HPC support team.
You can complete the labs wherever and whenever it is most convenient for you. The scheduled classroom sessions are there to give you dedicated time to work on the lab and to get help if you need it. Please note that we release the lab materials several days before the scheduled classroom session.
An instructor will be available during the session to answer questions and provide guidance. If no students arrive within the first 15 minutes, the instructor may leave.
The lab schedule below is still being finalized and will be updated as more labs are added. Please check this page and Slack for updates.
| Lab | Date | Title | Materials |
| 1 | September 17 | TBA | TBA |
| 2 | September 24 | TBA | TBA |
| ... | ... | ... | ... |