feat(weave): add wandb.agent_user_feedback feedback type#7184
Open
nikumar1206 wants to merge 7 commits into
Open
feat(weave): add wandb.agent_user_feedback feedback type#7184nikumar1206 wants to merge 7 commits into
nikumar1206 wants to merge 7 commits into
Conversation
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
|
Preview this PR with FeatureBee: https://beta.wandb.ai/?betaVersion=dd6aa4b2342035463ae56382ab5b9922ec4c78fb |
nikumar1206
force-pushed
the
nikhil/agent-reaction-feedback-type
branch
from
d16f1e6 to
3f564b6
Compare
nikumar1206
changed the title
nikumar1206
force-pushed
the
nikhil/agent-reaction-feedback-type
branch
from
3f564b6 to
432b9f6
Compare
nikumar1206
changed the title
Adds a human-applied feedback type for agents (vs scorer-applied wandb.agent_monitor). The value is stored as a tag in the existing scorer_tags column (e.g. an emoji glyph like ["👍"], or any manual label), with no scorer refs. Validation allows scorer_tags on this type and requires it non-empty; the type is multi-value so multiple tags per ref aggregate correctly.
nikumar1206
force-pushed
the
nikhil/agent-reaction-feedback-type
branch
from
432b9f6 to
8fcd562
Compare
AgentChatMessage now carries agent_version + status_code (from the message's span), and AgentTraceChatRes carries the root span's agent_name/agent_version/status_code. This lets the agents UI attach span_agent_name/version/status_code to human reaction feedback (wandb.agent_user_feedback) so reactions are filterable/groupable in the agent feedback aggregate endpoint.
Pins that the chat view surfaces agent_version + status_code (per-message and trace root) for reactions, and that an agent_user_feedback row round-trips the span_agent_name/version/status_code the UI attaches.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Backend support for human reactions on agents.
wandb.agent_user_feedbackfeedback type for human-applied tags (vs scorer-appliedwandb.agent_monitor). The value is stored as a tag inscorer_tags.Frontend: wandb/core#45619.
Testing