- precompute.py: Berechnet paarweise Cosine-Similarity aller Absätze,
speichert Top-10-Nachbarn pro Absatz in semantic_links Tabelle
- API: /api/similar-precomputed/{podcast}/{episode}/{idx} — liefert
vorberechnete ähnliche Stellen in <1ms
- Getestet: 728 Absätze, 7144 Links (Threshold 0.55)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
|
||
|---|---|---|
| .. | ||
| app.py | ||
| database.py | ||
| embeddings.py | ||
| precompute.py | ||
| requirements.txt | ||