v2.20.0rc0
📦 mlflowView on GitHub →
✨ 12 features🔧 10 symbols
Summary
MLflow 2.20.0rc0 introduces type‑hint based model signatures, Bedrock/Groq tracing, inline notebook trace rendering, uv‑accelerated model validation, a new chat panel in the Trace UI, and several enhancements such as the ChatAgent base class and improved Spark UDF handling.
✨ New Features
- Type Hint-Based Model Signature: define model signature from Python type hints in PythonModel.predict and validate input payloads.
- Bedrock / Groq Tracing Support: one-line auto-tracing for Amazon Bedrock and Groq LLMs via mlflow.bedrock.tracing or mlflow.groq.tracing.
- Inline Trace Rendering in Jupyter Notebook: render trace UI directly inside notebooks.
- Faster Model Validation with uv Package Manager: mlflow.models.predict now supports uv for faster environment validation.
- New Chat Panel in Trace UI: unified chat panel for LLM invocations showing messages, function calls, and raw payloads.
- Introduced ChatAgent base class for custom python agents.
- Supported Tool Calling in DSPy Tracing.
- Added support for invokers rights in Databricks Resources.
- Applied timeout override to within-request local scoring server for Spark UDF inference.
- Supported dictionary type for inference params.
- Made context parameter optional for calling PythonModel instances.
- Set default task for ChatModel.