Mini-hackathon on Multimodal Apps and Training LLMs

Come join us for a fun and innovative hackathon for building multimodal applications and training LLMs! Let's build exciting projects together 🎉

1. General Information
2. Themes
3. Resources
4. Schedule
5. Team Formation
6. Submission Process
7. Communication (Discord)
8. Evaluation Criteria
9. Acknowledgements

1. General Information

Date & Time:

• Date: Friday, May 16 to Saturday, May 17
• Time: 6:00 PM - 10:00 PM (May 16), 10:00 AM - 6:00 PM (May 17)

Location:

• EPFL AI Center Lounge (ELE 117)

2. Themes

LauzHack and the EPFL AI Team are hosting a mini-hackathon with two exciting tracks:

1. Multimodal Applications:

• The use of GenAI APIs and models (together.ai, OpenAI, HuggingFace, etc). together.ai is providing a lot of credits so let's use them!
• The Telegram bot builder API can be optionally used for a quick-and-dirty user interface, or a web app (Lovable, Gradio, Python Web Dev).
• Build innovative, multimodal products by combining APIs for new applications!

2. LLM Training:

• Fine-tune a small LLM to a specific domain
• Running a small LLM on-device, e.g. Android or Raspberry Pi
• Improving modeling efficiency, e.g. with mixture of experts or parameter efficient training

3. Resources

1. GenAI Applications:

• Tutorial on making a Telegram bot with GenAI APIs. This will be presented on Friday!
• Message Eric (organizer) on the LauzHack Discord for a together.ai key!

2. LLM Training (latest version will be presented on Friday!):

• Tutorial for setup.
• LLM baselines which will be presented on Friday.
• Cloud compute setup with EPFL resources: link (request access)

Hugging Face's LLM course is a great collection of resources for both tracks (LLM scientist and LLM engineer).

4. Schedule

Note you must have an "Approved" status on Luma to attend the event.

Friday, May 16:

• 6:00 PM: Tutorials/Workshops (AI Center Lounge)
  • Pizza
  • Intro to the event
  • Tutorial on making a Telegram bot with GenAI APIs
  • Pizza
  • (Around 8) Tutorial on training LLMs
  • More pizza and spontaneous tutorials based on people's interests.
• 10:00 PM: End of day. You CANNOT stay overnight, but you can continue working remotely.

Saturday, May 17:

• 10:00 AM: Breakfast (AI Center Lounge)
• (Deadline) Till 3:30 PM: Hack, hack, hack!
• 4:00 PM: Demos then prizes 🏆
• 6:00 PM: Apéro

😋 We will provide dinner on Friday and (breakfast, lunch, apero) on Saturday.

5. Team Formation

Up to 4 members per team. Declare your team here (need to request access).

6. Submission Process

⚠️ All projects should be submitted on DevPost. And make sure your team is up-to-date in the team formation document.

In order to be considered for a prize, all projects should demo/present (3 min) on Saturday afternoon.

When submitting your project, in the Additional info step, please select the track that you are competing for (Multimodal track or Training track) as in the attached image.

7. Communication (Discord)

Real-time information about the event, food service details, and questions related to the challenges will be posted in our personal Discord server (#2025-llm-multimodal-hackathon channel).

Please use the link sent to you via Luma (as this will add you to the private channel for the event). If you don't see the channel, please write to Eric (Organizer) or Petr (Organizer) on Discord.

8. Evaluation Criteria

Only work done during hackathon will be considered (and should be made explicit) for the project evaluation.

The following criteria will be used to guide the evaluation of the projects.

For multimodal track, we also consider how different modalities are incorporated into the application.

Keep in mind 3 min limit for presenting your project!

Criteria	Allocated points
1. Technical Impressiveness	6
1.1. How impressive is the project from a technical perspective?	3
1.2. How reasonable the technical and programming solutions are, given the limited timeframe of a hackathon?	3
2. Idea	6
2.1. How innovative, original and unexpected the project is?	3
2.2. How usable the idea is for the real-world target population to which the project is aimed? hackathon?	3
3. Prototype	8
3.1. Does the prototype work as advertised by the team, and as expected from a one-day work?	4
3.2. Does the prototype provide a good user experience and usability?	4
4. Presentation	4
5. Integrity Check
TOTAL	24

9. Acknowledgements

A BIG thanks to the MLO lab and Research Compute Platform for technical support, and to together.ai for providing credits 🙏