Skip to content

v2.1.0

Choose a tag to compare

@galileo-automation galileo-automation released this 09 Apr 19:38
· 67 commits to main since this release
v2.1.0
eedf91f

v2.1.0 (2026-04-09)

This release is published under the Apache-2.0 License.

Bug Fixes

  • Cannot pass prompt_settings as dict to run_experiment (#527, a32ea66)

  • Missing tasks in hierarchical CrewAI crew traces (#532, 6afaa04)

  • NewSDK: get_version* are [very] confusing (#530, cf606c4)

  • Regression issues on experiments (#511, b7cc9bd)

  • Resolve OpenAI test failures when Azure env vars are set (#540, 5322d63)

  • datasets: Adjust default model alias name (#535, 95a4e45)

  • docs: Update span docstrings with accurate input formats (#541, 63197c1)

Chores

  • Remove CODEOWNERS (#533, 354491b)

  • Update dev environment to recommend Python 3.13 (#518, 4b197ab)

  • Update ruff target-version and fix mypy pre-commit hook (#536, 4b6bffb)

Features

  • Implement google a2a instrumentor (#529, ca24de6)

  • experiments: Add on_error callback to flush and experiment run (#537, e08596c)

  • langchain: Add multimodal content support to LangChain handler (#515, fb6c1a1)

  • logger: Validate input types in start_trace (#539, e07ab69)

Refactoring

  • Add @deprecated decorators to legacy service functions in datasets, experiments, prompts, log_streams (#524, 546e00f)

Detailed Changes: v2.0.0...v2.1.0