Create and upload a Hugging Face model card
Retrain models for new data at edge devices
Evaluate open LLMs in the languages of LATAM and Spain.
Measure over-refusal in LLMs using OR-Bench
Display model benchmark results
Display and submit LLM benchmarks
Quantize a model for faster inference
Demo of the new, massively multilingual leaderboard
View and compare language model evaluations
Create and manage ML pipelines with ZenML Dashboard
Display benchmark results
Upload ML model to Hugging Face Hub
Download a TriplaneGaussian model checkpoint
Modelcard Creator is a tool designed to simplify the process of creating and managing model cards for machine learning models. Specifically, it helps users generate and upload model cards to the Hugging Face platform, making it easier to document and share model performance, dataset details, and other relevant information. This tool is particularly useful for model benchmarking and transparency in AI development.
pip install modelcard-creator to install the package.modelcard init to create a basic template.modelcard create to generate the final model card.modelcard push command to share your model card on the Hugging Face hub.What is a model card?
A model card is a standardized document that provides detailed information about a machine learning model, including its architecture, training data, performance metrics, and ethical considerations.
Can I customize the model card template?
Yes, Modelcard Creator allows you to customize the template to include specific details, visuals, and formatting tailored to your model's needs.
Where can I find more information about Hugging Face model cards?
You can find additional details and examples of model cards on the Hugging Face Model Cards documentation page.