Here you will find all the info you need to know to participate in the 2024 EAGE Annual Hackathon in Oslo!
This year, we seek for ways of solving geoscience problems in a computationally and energy-efficient way. Use AI to mimic the first-principle physics, or train a model to suggest hyperparameters for an algorithm and save multiple rounds of search iterations, or bring your code and make it work faster!
This event will be facilitated in English and will be an in-person only event. To register, select the hackathon option during the registration for the annual event. Once registered do not hesitate to post in the following threads:
💡 Brainstorm Ideas here
🛠️ Explore resources here
📣 Find your team here
👉 Register here
- Join the online meetup in May to start thinking about ideas and forming a team if you don't already have one.
- We will use a Github discussion board – if you are looking for a team to join or have a topic that you would like some people to help you work on – please post.
- There will be some time on Sunday morning to finalize the teams, but the expectation is that you have tried to find a team upfront.
- Get your team ready!
- You will need to bring your laptop and power cable.
- Ideate around the technical challenge your team will work on. You can use our Github discussion board for this. Remember the theme!
- What data do you need? The hackathon facilitators will put together some data for you to use.
- Are you going to use an existing model or build one from scratch?
- What technology will your team use (python packages, dashboards, interpreters..etc).
The facilitators will provide some ideas and provide you with some technology to get you thinking - you can find this thread in the discussion board above.
- What technology will your team use (python packages, dashboards, interpreters..etc).
- Familiarize yourself with the agenda.
- This is a fairly short hackathon. Your team will be expected to present your final outcome on Monday afternoon at 14:00. That means code freeze will happen at 12:00.
- With your team, try to come up with a plan of attack keeping in mind the short time.
- Yes, there will be a final presentation from each team with two teams up for award from a judging panel of experts and from the hackathon participants.
- The facilitators will reveal the criteria for the presentation Monday morning (this is so you can focus on your work and not on creating a presentation 😊)
- Remember to take screenshots, record methods, and document your thought process along the way – this will help you to faster put together a final presentation.
- You are expected to upload all code to a repository in this hackathon Github organization.
- Award winners will be asked to present their results at the Energy Transition Theatre within the main exhibition floor during the conference - schedule to be announced soon!.
Sunday, June 9:
9:00 - 9:30 CEST - Welcome and coffee
9:30 - 10:30 CEST - Introduction & hackathon modalities
10:30 – 11:00 CEST – Project and team discovery time
11:00 - 18:00 CEST - Hacking all day
Monday, June 10:
9:00 - 9:15 CEST - Welcome and coffee
9:15 - 12:00 CEST - Hacking and preparation of presentations
12:00 CEST: CODE FREEZE – Final Presentation Instructions
12:15-14:00 CEST – Time for working and practicing presentations
14:00 - 15:15 CEST - Presentations to the Judging Panel
15:15 - 15:45 CEST - Announcement of the winners and prizes *
🍔 Food: Lunch will be provided - including a vegetarian option.
📍 Location: The hackathon will be at the same conference venue as the rest of EAGE Annual: NOVA Spektrum - Messeveien 8, 2004 Lillestrøm, Norway. It is a short walk from the train station at Lillestrøm - with connections to both airport and hotels and to the center of Oslo. We will be in ROOM 8.
To equip our hackathon participants with crucial insights into LLMs and physics-based ML, we're excited to offer two comprehensive workshops. These sessions are designed not only to introduce the foundational technology behind our quickstart materials but also to serve as a prime networking opportunity, potentially helping you find your ideal teammate before the competition begins.
- NeMo Framework for LLM Training and Fine-tuning: A dive into the essentials of Language Model training and fine-tuning. This workshop aims to walk you through examples and methods needed to enable the adaptation of an LLM to a new knowledge domain. By Oleg Ovcharenko
- NVIDIA Modulus for Reservoir Simulation and History Matching: Explore the end-to-end reservoir simulation and history matching with physics-ML. Participants will understand the math behind the problem as well the details of its implementation. By Clement Etienam
Registered participants will receive an invitation email for workshops a few weeks before the EAGE Annual event. It's a perfect chance to ask questions, clarify doubts, and connect with like-minded innovators.
Follow this thread for datasets, packages, and tips and tricks: We will have access to sufficient GPU resources.
You will be supported during the hack by:
- Oleg Ovcharenko, NVIDIA
- Ashley Russel, Equinor
- Lukas Mosser, AkerBP
- Clement Etienam, NVIDIA
- Issam Said, NVIDIA