Llama Guard 3 8B is Meta's safety classifier designed to detect violations across 14 different harm categories in LLM conversations. It can classify both user and assistant messages according to a detailed safety policy covering violence, sexual content, self-harm, and other risks. The model uses a specific prompt format and chat template for consistent safety evaluations.
Llama Guard 3 8B sits in PulseGate's Foundation models & chat category. It focuses on identifying and categorizing potentially unsafe or harmful content in user-assistant conversations according to a detailed safety policy. It is built as an open-source project for developers. Llama Guard 3 8B is open source under the Open Source license. It runs on the web, the command line, and API.
It is developed by Meta, and it first shipped in 2024. The project is developed in the open on GitHub with 7.7k stars. Among its 3 catalogued features are Content Safety, Harm Classification, and Policy Enforcement. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do