huggingface/setfit

Health v2 · maintenance only

Adoption not part of the score

2784 stars · 264 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

SetFit is a Python library for efficient, prompt-free few-shot fine-tuning of Sentence Transformers for text classification. It achieves high accuracy with as few as 8 labeled examples per class while training and inferring much faster than large language model approaches.

Use cases

• train a text classifier with only a handful of labeled examples
• classify customer reviews as positive or negative sentiment
• do zero-shot text classification without labeled training data
• fine-tune multilingual sentence transformers for classification
• compress and speed up classification models with ONNX for production
• perform knowledge distillation and hyperparameter optimization for classifiers

When to choose

• you have very little labeled data but need an accurate text classifier
• you want fast, cheap training and inference without large LLMs or prompt engineering
• you need multilingual classification using any Sentence Transformer checkpoint
• you want a lightweight model deployable on CPU or optimized via ONNX

When to avoid

• you need generative tasks like summarization or question answering
• you have abundant labeled data and can fine-tune a full transformer for maximum accuracy
• you need state-of-the-art performance on complex reasoning tasks that require large language models

Facets

library · maturity active

machine-learning nlp llm-training machine-learning deep-learning python cross-platform few-shot-learning sentence-transformers text-classification zero-shot-classification embeddings hugging-face onnx natural-language-processing gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="huggingface/setfit")