# Transformers

Hugging Face's Python library of model definitions for text, vision, audio and multimodal machine learning, used for both inference and training.

Transformers is a model-definition framework for machine learning models across text, computer vision, audio, video and multimodal tasks. A model defined in Transformers works with most training frameworks, such as DeepSpeed and PyTorch Lightning, inference engines such as vLLM and TGI, and libraries such as llama.cpp and mlx.

Its Pipeline API is a high-level class for inference that handles preprocessing and output for tasks such as text generation, speech recognition, image classification and visual question answering. Over a million model checkpoints on the Hugging Face Hub can be used with it.

It requires Python 3.10+ and PyTorch 2.5+, and is Apache-2.0 licensed.

Website: https://huggingface.co/transformers