sinaalemohammad

/

Neon

like 0

This Pull Request updates the model card for the "Neon: Negative Extrapolation From Self-Training Improves Image Generation" repository.

Key changes include:

• Metadata: Added pipeline_tag: unconditional-image-generation for better discoverability. The library_name has been omitted as the original repository does not provide direct integration with Hugging Face libraries (e.g., diffusers) for automated usage snippets.
• Content:
  • The model card now features the full paper title as the main heading, along with a direct link to the paper (Neon: Negative Extrapolation From Self-Training Improves Image Generation).
  • A link to the official GitHub repository (https://github.com/SinaAlemohammad/Neon) has been included.
  • An introductory section summarizing the paper's abstract and contributions has been added.
  • The "Benchmark Performance" table from the GitHub README is now included, showcasing model results across various architectures and datasets.
  • The comprehensive "Quickstart" section, detailing environment setup, model download, and evaluation procedures (with original bash commands), has been incorporated to guide users on how to run the models.
  • Additional sections like "Toy Experiment", "Repository Map", "Citation", "Contact", and "Acknowledgments" have been added to provide complete context from the original GitHub README.

These changes aim to make the model card more informative, discoverable, and aligned with best practices for documenting AI artifacts on the Hugging Face Hub.