baartmar

/

DeepNSM-8B

like 1

This PR significantly enhances the model card for baartmar/DeepNSM-8B by:

• Adding pipeline_tag: text-generation to the YAML metadata, ensuring the model is discoverable under the correct task filter on the Hugging Face Hub (e.g., at https://huggingface.co/models?pipeline_tag=text-generation).
• Correcting library_name to transformers in the YAML metadata, as transformers is the primary library used for loading and performing inference with this model, rather than peft which was used for fine-tuning.
• Adding relevant tags (nlp, paraphrase, semantic-analysis, llama) to improve categorization and searchability.
• Populating and expanding various sections with detailed information extracted from the paper abstract and the GitHub repository:
  • Model Description: Provides a comprehensive overview of the model's purpose and the Natural Semantic Metalanguage (NSM) framework.
  • Developed by: Includes the full list of authors.
  • Model Sources: Adds a link to the Colab Demo, serving as a direct project page/demo.
  • Uses: Clarifies direct and downstream applications, and outlines out-of-scope uses.
  • Bias, Risks, and Limitations: Discusses potential issues such as language specificity, generative model characteristics, and resource requirements.
  • How to Get Started with the Model: Includes a ready-to-use Python code snippet for inference using the transformers library.
  • Training Details: Provides specific training hyperparameters and links to the training dataset (baartmar/nsm_dataset).
  • Evaluation: Summarizes the evaluation methodology and key results.
  • Citation: Adds the BibTeX entry for the associated paper.
• Removing all [More Information Needed] placeholders, making the model card complete and highly informative.

These updates aim to provide users with a much richer understanding and easier access to the model.