Skip to content

Latest commit

 

History

History
79 lines (63 loc) · 3.25 KB

CONTRIBUTING.md

File metadata and controls

79 lines (63 loc) · 3.25 KB

Contribution Guidelines

Welcome to Awesome Smol Models! We’re excited to have you contribute. Please follow these steps to ensure smooth collaboration.

How to Contribute

  1. Find an area to contribute:
    • Add a new model.
    • Enhance documentation or examples.
    • Suggest new tools/resources.
  2. Fork the repository: Click the “Fork” button to create a copy of this repo in your GitHub account.
  3. Clone your fork: 
    git clone https://github.com/afondiel/awesome-smol-models  
cd awesome-smol-models
  4. Create a branch: 
    git checkout -b <branch-name>
  5. Make your changes: Add or edit models, update files, or fix issues.
  6. Run a check: Validate formatting and links using tools like markdownlint or awesome-lint.
  7. Submit a Pull Request: Push your branch and create a PR following the template below.

Pull Request Template

# Pull Request  

## Description  
Provide a summary of your changes, including the type of contribution (new model, documentation, fix, etc.).  

## Checklist  
- [ ] My code follows the repository style guide.  
- [ ] I have added appropriate references and links.  
- [ ] I have tested my changes locally.  

## Related Issue  
If applicable, link the issue number: `#<issue-number>`  

## Additional Notes  
Any other relevant information.

Add New Models

Add the model under the relevant section with the following format:

| TinyYOLO          | Object Detection      | Mobile, Edge         | [GitHub](https://github.com)     |

If you manage to run the model on any edge device/sim tool, you're welcome to add its benchmarks to highlight performance and usability.

Benchmark Table Format

Model Task Accuracy Latency (ms) Model Size (MB) Platform References
MobileNet V2 Image Classification 72.0% 25 4.3 Android, iOS, Web TensorFlow Lite

How to Add Benchmarks

  1. Download and Evaluate: Run the model on a standardized dataset (e.g., ImageNet, COCO).
  2. Measure Performance:
    • Use tools like ONNX Runtime, TensorFlow Lite Benchmark Tool, or CoreML Profiler.
    • Record latency on mobile/edge devices (e.g., Raspberry Pi, Jetson Nano).
  3. Document Results: Use the table above to document key metrics.
  4. Submit with PR: Attach the benchmark results with your PR submission.

Enhancing Resources

  • Propose new blog posts, podcasts, or tutorials.
  • Validate all links and content for accuracy.

Examples of a Good Contribution

Adding a New Model

  • Fork the repository.
  • Add the model under the relevant section with the following details: 
    | TinyYOLO          | Object Detection      | Mobile, Edge         | [GitHub](https://github.com)     |
  • Run benchmarks and attach results.
  • Submit a pull request.