Conversation
|
GitHub CI seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account. You have signed the CLA already but the status is still pending? Let us recheck it. |
web/ee/src/components/pages/evaluations/customEvaluation/CustomEvaluation.tsx
Dismissed
Show dismissed
Hide dismissed
There was a problem hiding this comment.
Pull Request Overview
This PR bumps the version from 0.60.2 to 0.61.0 and introduces significant enhancements to support custom evaluations, improve metric handling, and refactor workflow utilities. Key changes include adding evaluator revision support, enhancing metric resolution with better slug tracking, implementing CSV export for custom evaluations, and refactoring SDK workflow utilities.
Key Changes
- Added evaluator revision support with data merging and fallback mechanisms
- Enhanced metric column factory with nested object support, improved slug resolution, and dynamic type inference
- Implemented custom evaluation type ("custom") throughout the evaluation flow
- Added CSV export functionality for custom evaluations with robust column flattening
- Refactored SDK workflow utilities (renamed
PARAMETERS_REGISTRYtoCONFIGURATION_REGISTRY, madeslugoptional in builtin workflows)
Reviewed Changes
Copilot reviewed 113 out of 118 changed files in this pull request and generated 55 comments.
Show a summary per file
| File | Description |
|---|---|
| web/package.json, web/oss/package.json, web/ee/package.json, sdk/pyproject.toml | Version bump to 0.61.0 |
| web/oss/src/state/evaluators/atoms.ts | Added evaluator revision fetching and merging logic |
| web/oss/src/state/app/hooks.ts, atoms/fetcher.ts | Enhanced app filtering and validation logic |
| web/ee/.../evaluation pages | Added "custom" eval type support in routing |
| web/ee/src/lib/metricColumnFactory.tsx | Major refactoring with nested metrics and improved resolution |
| web/ee/.../CustomEvaluation.tsx | Added CSV export with column flattening |
| web/ee/.../utils.tsx | Enhanced evaluator metric resolution with run indexes |
| sdk/agenta/sdk/workflows/utils.py | Renamed registry and refactored URI handling |
| sdk/agenta/sdk/workflows/builtin.py | Made slug parameter optional |
| sdk/agenta/sdk/models/* | Added new model files for git, evaluations, testsets, blobs |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
release/v0.61.0