feat(usage): add model dimension dropdown for analysis and realtime o… - #427
Open
Adkimsm wants to merge 2 commits into
Open
feat(usage): add model dimension dropdown for analysis and realtime o…#427Adkimsm wants to merge 2 commits into
Adkimsm wants to merge 2 commits into
Conversation
…verview Add global model_dimension query parameter to analysis and realtime overview APIs, supporting 'model' (default) and 'alias' modes. In alias mode, model statistics aggregate by model_alias with fallback to the normalized model name when alias is empty. Pricing continues to use the real model name and alias for rule matching. Frontend adds a global dropdown to AnalysisPanel and OverviewRealtimePanel controlling model ranking, efficiency, composition, heatmap, and realtime current usage. Analysis responses are cached per dimension to avoid refetching on toggle. Realtime query keys include model_dimension to prevent stale cache reuse.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 8f87918077
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
…back 1. resolveDisplayRealtime: remove scope fallback so switching model dimension no longer reuses the previous dimension's realtime data. 2. KeyOverviewPage loadRealtime: clear stale realtime response on dimension/window change to prevent mismatched data display. 3. UsagePage loadAnalysis: skip core fetch on cache hit; add force option for manual refresh to bypass the per-dimension cache. 4. resolveUsageModelDimensionKey and resolveUsageOverviewRealtimeModelKey: test trimmed alias for emptiness instead of comparing against the 'unknown' sentinel, so a legitimate alias of 'unknown' is no longer treated as missing.
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.
Summary
Add global model_dimension query parameter to analysis and realtime overview APIs, supporting 'model' (default) and 'alias' modes. In alias mode, model statistics aggregate by model_alias with fallback to the normalized model name when alias is empty. Pricing continues to use the real model name and alias for rule matching.
Frontend adds a global dropdown to AnalysisPanel and OverviewRealtimePanel controlling model ranking, efficiency, composition, heatmap, and realtime current usage. Analysis responses are cached per dimension to avoid refetching on toggle. Realtime query keys include model_dimension to prevent stale cache reuse.