model-publisher

What problem does it solve?

This Skill streamlines the process of packaging trained AI models, generating essential documentation like model cards, and publishing them to platforms like Hugging Face Hub, simplifying model distribution and collaboration.

Core Features & Use Cases

• Checkpoint Packaging: Converts training checkpoints to SafeTensors, ensuring integrity and compatibility.
• Model Card Generation: Automatically creates comprehensive README.md files detailing model usage, architecture, training data, and limitations.
• Hugging Face Hub Publishing: Manages repository creation, artifact uploads, and versioned updates using huggingface-cli.
• Optional GGUF Export: Generates quantized GGUF variants for efficient local inference.
• Use Case: A machine learning engineer has just finished training a new natural language processing model and needs to share it with their team. They use this Skill to package the model, generate a detailed model card, and publish it to a private Hugging Face repository.

Quick Start

Use the model-publisher skill to package and publish the latest trained model to Hugging Face Hub.