Huggingface Hub

Access Hugging Face Hub models, datasets, and spaces via the huggingface_hub Python library. Use when you need to list, search, download, or upload HF assets.

Hugging Face Hub Skill

Interact with Hugging Face Hub — browse models/datasets/spaces, download files, upload artifacts, and more.

Prerequisites

Install the library:

pip install huggingface_hub

Save your HF token to Settings > Advanced as HF_TOKEN (get one at https://huggingface.co/settings/tokens).

Usage

python3 /home/workspace/Skills/huggingface-hub/scripts/hf_hub.py <command> [options]

Commands

Command	Description
list-models	List models. Options: --search, --sort, --direction, --limit
list-datasets	List datasets. Options: --search, --sort, --direction, --limit
list-spaces	List spaces. Options: --search, --sort, --direction, --limit
model-info	Get model info. Options: --model
dataset-info	Get dataset info. Options: --dataset
download-file	Download a file from a repo. Options: --repo-id, --filename, --revision
upload-file	Upload a file. Options: --local-path, --repo-id, --repo-type, --path-in-repo
whoami	Show authenticated user info

Examples

# Search for text-to-image models
python3 .../hf_hub.py list-models --search "text-to-image" --limit 10

# Download a model file
python3 .../hf_hub.py download-file --repo-id "stabilityai/stable-diffusion-xl-base-1.0" --filename "pytorch_model.bin"

# Upload a file
python3 .../hf_hub.py upload-file --local-path ./model.pt --repo-id "your-username/my-model" --repo-type "model"

# List top-rated datasets
python3 .../hf_hub.py list-datasets --sort "likes" --direction desc --limit 5

API Reference

This skill wraps huggingface_hub. Key functions:

• list_models, list_datasets, list_spaces — browse hub
• huggingface_hub — authentication and file ops
• InferenceClient — run inference on hosted models/spaces

Notes

• Downloads cache to ~/.cache/huggingface/
• repo-type values: model, dataset, space
• Default revision is main