Safety and evaluation tools for Gemma models. Part of Google's responsible AI safeguards.
What is ShieldGemma?
ShieldGemma is a set of safety and evaluation tools for Gemma models, part of Google's responsible AI safeguards. See it in our AI Tools collection.
Key benefits
- Evaluate and harden Gemma-based applications for safety.
- Align with responsible AI practices when deploying or fine-tuning Gemma.
- Part of the Responsible Generative AI Toolkit.
Use cases
- Deploying or fine-tuning Gemma models in production.
- Evaluating Gemma-based applications for safety and alignment.
About ShieldGemma
ShieldGemma is a set of safety and evaluation tools for Gemma models, part of Google’s responsible AI safeguards. It helps developers evaluate and harden Gemma-based applications for safety and alignment. The documentation is part of the Responsible Generative AI Toolkit and is useful for anyone deploying or fine-tuning Gemma models in production.
Sources & further reading
- gemma
- safety
- responsible-ai
- safeguards
Similar resources in AI Tools
More ai tools to explore.
Comments