Remove XGBoost ML + spam scorer; add image-aware AI curation #93
Loading…
Reference in a new issue
No description provided.
Delete branch "feat/issue-80-collect"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Collecting branch for the AI-curation rework: replaces Fenliu's XGBoost ML model and rule-based spam scorer with a two-stage AI classifier that checks post text for promotion and the image for a real cat.
Sub-PRs merged into this branch:
reviewer_notestoauto_reject_reasonreview.htmltemplateCloses #80
Closes #82
Closes #83
Closes #84
Closes #85
Closes #87
Closes #91
Closes #94
Closes #96
This PR is too large to review automatically (14451 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (14451 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (14660 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (14660 lines; limit is 3000). Please consider splitting it into smaller PRs.
Extract the scheduler's inline classification into a shared classify_posts() in services/ai_pipeline.py and call it (plus reject_ai_flagged_posts) from the manual POST /hashtags/{id}/fetch endpoint and the review page's _fetch_all_active_streams, so every path that saves new posts also classifies them.This PR is too large to review automatically (14965 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (14965 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (15058 lines; limit is 3000). Please consider splitting it into smaller PRs.
This PR is too large to review automatically (15058 lines; limit is 3000). Please consider splitting it into smaller PRs.