v4.51.3-LlamaGuard-preview
📦 transformersView on GitHub →
✨ 5 features🔧 4 symbols
Summary
This release introduces LlamaGuard 4 and Llama Prompt Guard 2, providing multimodal safety moderation for text and images. It is available as a preview tag prior to the official v4.52.0 minor release.
Migration Steps
- Install the preview version using: pip install git+https://github.com/huggingface/transformers@v4.51.3-LlamaGuard-preview
✨ New Features
- Added support for LlamaGuard 4, a multimodal 12B model for detecting inappropriate content in text and images.
- Support for Llama Prompt Guard 2 via the pipeline API or AutoModelForSequenceClassification.
- Ability to exclude specific safety categories during moderation using the 'excluded_category_keys' parameter in chat templates.
- Multilingual support for content moderation.
- Support for both input (prompt) and output (generation) moderation.