v0.2.0-alpha.23

0.2.0-alpha.23 (2024-08-26)

Full Changelog: v0.2.0-alpha.22...v0.2.0-alpha.23

Features

• improvement: updates to custom metric runner
• improvement: skip metrics if already computed, surface errors for each metric
• feat: add --dataset flag so custom metrics can be forced to run on only specific datasets

0.2.0-alpha.22 (2024-08-21)

Full Changelog: v0.2.0-alpha.21...v0.2.0-alpha.22

Bug Fixes

• add missing dependency for Anthropic notebook example (eddc160)

Chores

• ci: also run pydantic v1 tests (#319) (6959e23)
• client: fix parsing union responses when non-json is returned (#318) (1b18e64)