huggingface-hub

Hugging Face Hub SOP

Discover, load, run, and publish models and datasets on the Hugging Face Hub. Covers CLI auth, transformers pipelines, the Inference API, push_to_hub, and dataset loading.

When to Use

• User wants to find the right pre-trained model for a task
• User wants to load and run a HF model locally
• User wants to call the HF Inference API without loading weights locally
• User wants to push a fine-tuned model or dataset to the Hub
• User wants to work with HF datasets library

---

Part 1 — Auth Setup

1. Get a Token

1. Go to https://huggingface.co/settings/tokens
2. Click New token → name it → select Write role (needed for push_to_hub) or Read (for private model downloads)
3. Copy the token (starts with hf_)

2. Login via CLI

pip install huggingface_hub

huggingface-cli login
# Paste your token when prompted; it is saved to ~/.cache/huggingface/token

Or set as environment variable (CI/CD):

export HUGGINGFACE_TOKEN="hf_..."

Or in ~/.cowrangler/credentials.env:

HUGGINGFACE_TOKEN=hf_...

3. Programmatic Auth

from huggingface_hub import login
import os

login(token=os.environ.get("HUGGINGFACE_TOKEN"))

---

Part 2 — Model Discovery

Search by Task

from huggingface_hub import HfApi

api = HfApi()

# Search models by task tag
models = api.list_models(
    task="text-classification",
    sort="downloads",
    direction=-1,
    limit=10,
)

for m in models:
    print(f"{m.modelId:50s}  downloads={m.downloads:>10,}  likes={m.likes}")

Common task tags: text-classification, token-classification, question-answering, text-generation, summarization, translation, fill-mask, sentence-similarity, image-classification, object-detection, automatic-speech-recognition, text-to-speech, image-to-text

Filter by Language and Library

models = api.list_models(
    task="text-classification",
    language="tr",       # ISO 639-1 language code
    library="transformers",
    sort="downloads",
    direction=-1,
    limit=5,
)
for m in models:
    print(m.modelId, m.downloads)