[Proposal]: machine-learning-llm-python

[tompollard]

1. Lesson Topic

Introduction to large language models in Python

2. Lesson Language

English

3. Draft materials

The materials are already being developed as part of a short workshop being run by our research group.

4. Requirements for existing materials

• [ ] do your lesson materials conform to our Code of Conduct?
• [ ] are your materials already on GitHub and do they use The Carpentries lesson template or The Carpentries Workbench? (you can visit the Workbench tutorial to learn more about how to use our lesson infrastructure).

5. New repository creation

• [X] I need a new repository created

6. Transfer existing repository

• [ ] my lesson meets both criteria in part 4 and I would like to transfer the repository to The Carpentries Incubator

7. Collaborators

@Gallifantjack

8. Info/Questions

The course would be based around our work on large language models (e.g. https://arxiv.org/abs/2406.12066), with the goal of working through a simple task of exploring and evaluating large language models around a clinical narrative. The materials are already being developed as part of a short workshop being run by our research group.

[qualiaMachine]

Excited to see this, @tompollard! I wanted to put this related lesson on your radar in case you haven't seen it: https://carpentries-incubator.github.io/python-text-analysis/

That one focuses more on NLP fundamentals (i.e., creating text embeddings) but it does have a short section on LLMs. We've been considering extending that lesson with a handful of additional LLM examples since demand is high, but a full workshop on the topic honestly may be more appropriate. Feel free to borrow from our existing materials if anything looks useful. I might also be interested in joining your dev team if you're looking for help! Just give me a few months to get the extended Trustworthy AI/ML lesson in better shape first haha.

[tobyhodges]

Thanks for this lesson proposal @tompollard.

I have created the repository for you at https://github.com/carpentries-incubator/intro-LLMs-python/.

You and @Gallifantjack should have received an invitation to join the maintainer team for this lesson. I have also granted you Admin access to the repository and the ability to add and remove collaborators on the maintainer team.

The README of the repository includes a brief list of recommended next steps. You might find the Workbench documentation a helpful resource while working with The Carpentries lesson infrastructure. The Carpentries Curriculum Development Handbook and our Collaborative Lesson Development Training curriculum provide an overview of the approach we recommend to lesson design and development. But please feel free to post back here or ping me if you get stuck or if you have any further questions. I am always happy to support lesson development!

To keep in touch with The Carpentries curriculum community you may also find it useful to join the incubator-developers list on The Carpentries TopicBox and/or the lesson-dev channel on The Carpentries Slack.