Skip to content

Latest commit

 

History

History
26 lines (21 loc) · 1.82 KB

Submission_Guide.md

File metadata and controls

26 lines (21 loc) · 1.82 KB
Raw

Submission Guide

1. Create a Github repo with your documented source code and a README detailing your submission and documenting your journey, including general and category specific items below:

  • Describe your submission, how it works, what are the goals.
  • Reasoning behind choosing framework, model, dataset, technique.
  • Describe findings, difficulties and limitations, and future perspectives.

Build

Submissions to this category can be described as open-source applications employing AI for good.

  • Project dependencies, environments, APIs, or additional considerations required for your application to run
  • User / use case analysis
  • Feasibility, scaling, operating cost

Train

Submissions to this category are efforts, experiments, and demonstrations of training AI models. Source code, notebook, or something resembling https://huggingface.co/models

  • Dataset selection methodology, cleaning, synthesizing
  • Framework, model/technique selection, expectations, outcome
  • Progress, Benchmarking, examples

Analyze

Submissions to this category are in depth analyses into AI topics, alongside research and experimentation. Can include visualization, classification, predictive methods, and data storytelling, resembling a Jupyter notebook, or Kaggle.

  • Topic, dataset, objective, approach selection and reasoning
  • Findings, predictions, prescriptions, future endevors.

2. Create a short video demonstrating your submission running, outlining key relevant details and highlighting innovative approaches and findings.

3. Create an Issue at https://github.com/numfocus/numhack-2024/issues, select what category you are submitting to and link to your Github repository and video. Once Issue has been created, also select a Label on the Issue page, indicating the technical category using the Label options; a-build, a-train, a-analyze.