mirror of
https://github.com/discourse/discourse.git
synced 2026-08-05 17:25:34 +08:00
Previously, sentiment and emotion classification only used configured classification model endpoints, which blocked sites that could not run those models. This change lets admins choose agent-backed LLM classifiers for sentiment and emotion while storing results under stable model keys so LLM changes do not force historic reclassification. |
||
|---|---|---|
| .. | ||
| ai_moderation | ||
| sentiment | ||