Are Large Language Models (LLMs) really the future of enterprise AI, or are we caught in the hype? Join us for a groundbreaking 12-hour hackathon that puts both LLMs and traditional ML approaches to the test in real-world corporate scenarios. Teams will tackle challenging tasks across genomics, time series, financial data, and NLP domains under realistic constraints – because the real world doesn't always give you unlimited GPUs and perfect datasets.
What makes this hackathon unique? You'll be part of a structured comparative study that will shape how companies approach AI solutions. Teams will be specifically assigned to either LLM or traditional ML approaches, working with diverse datasets under varying time and resource constraints. Whether you're a data scientist, ML engineer, researcher, or graduate student, join us to discover what really works in AI beyond the hype!
Requirements
Our hackathon employs a multi-metric scoring system that combines performance accuracy with real-world efficiency metrics. The score (0-100) calculated as follows:
Primary Performance Metrics (60%):
- Task-specific accuracy metrics (F1-score, RMSE, ROUGE/BLEU) based on each team's assigned dataset
- Updated every 2 hours checkpoints to track progress
Resource Efficiency Score (40%):
- Computation time (15%)
- Memory usage (15%)
- Scalability test performance (10%) - measured by model performance on larger data batches
Each team's total score will be automatically calculated and updated on the leaderboard during the six checkpoints. This scoring system incentivizes both high performance and practical efficiency, aligning with our goal of identifying truly production-ready solutions.
Prizes
Best Team
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Ravid Shwartz Ziv
NYU
Judging Criteria
-
Primary Performance Metrics
Task-specific accuracy metrics (F1-score, RMSE, ROUGE/BLEU) based on each team's assigned dataset Updated at 6h, 12h, and 24h checkpoints to track progress -
Resource Efficiency Score
Computation time Memory usage Scalability test performance - measured by model performance on larger data batches
Questions? Email the hackathon manager
Tell your friends
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.