transformers

huggingface/transformers

Hugging Face Standard Library for Pretrained Multimodal Modelsready

Transformers is Hugging Face's core Python library that unifies the definition and usage of text, image, audio, video, and multimodal models. It is used by researchers, ML engineers, and application developers to load pretrained models and integrate them into training, inference, and serving pipelines. Releases of this package directly impact new model support, API compatibility, and integration stability with PyTorch/JAX/TF and various inference and training ecosystems.

machine-learningnlpcomputer-visionmultimodalpretrained-modelshuggingface

Choose version

5.22.21w ago

huggingface/transformers Patch release v5.22.2

Original

Key point

Hugging Face Transformers v5.22.2 patch release includes security enhancements and fixes for distributed processing errors.

Details

Security: Strengthened Jinja sandboxing to address security vulnerabilities.
Distributed processing: Fixed device_id configuration logic in NCCL environments to resolve issues related to distributed training.

transformers releases | AI Briefing