1. Eligibility

   - Open to data scientists, ML engineers, researchers, and graduate students

   - Must be 18 years or older

   - Teams of 4-5 members

   - Total of 60 participants will be selected based on pre-hackathon survey

   - Both students and professionals welcome

2. Project Requirements

   - Solutions must be developed during the hackathon period

   - Teams will be assigned either LLM or traditional ML approach - no switching allowed

   - Must use the provided dataset for your assigned category

   - All code must be original or properly attributed open-source

   - Solutions must be reproducible on standard computing resources

3. Technical Requirements

   - Must use the provided computing resources - no external GPU clusters

   - All dependencies must be documented in requirements.txt or environment.yml

   - Code must include proper documentation and setup instructions

   - Must log and report resource utilization metrics

   - Must implement checkpoints at 2h, 4h... marks

4. Submission Requirements

   - Complete source code on GitHub

   - Project write-up (max 500 words)

5. Judging Criteria

   - Performance Metrics (40%)

   - Resource Efficiency (25%)

   - Scalability & Adaptability (20%)

   - Code Quality & Documentation (15%)

6. Intellectual Property

   - Teams retain ownership of their code

   - Solutions must be open-source (MIT License)

   - Aggregate results may be published in research papers

7. Code of Conduct

   - Professional and respectful behavior required

   - No discrimination or harassment

   - Share knowledge and collaborate within teams

   - Follow ethical AI development practices

Organizers reserve the right to modify rules for clarity. All decisions by judges are final.