Narration: stop the 1B model copying the example title verbatim

Narration: kill label-parroting at the source + reduce gate false positives

Narration: stop category-label parroting + force flowing prose

Narration: stop the gate rejecting the LLM's own title + 'Just' (traces-driven)

Trace: create the dataset repo before pushing (upload_file does not auto-create)

Trace: log push failures + boot self-check for HF_TOKEN write access

Narration: route titles, varied phrasing, flowing LLM prose, looser gate

fixed the narration+traces

Round-4 fixes: drop overclaiming match-tag, dedup stops, hidden excludes famous

Round-2 route-fit: top-N category cap + weak-match honesty

fixed ui

v2 with model changed and using gradio server func

Design system, live freshness, autocomplete, and clay-pin markers

Wanderlust: taste-aware Paris discovery routing