Metadata: size_categories, provenance replay wording, contributor name, collection
#3
by rmems - opened
Grok Bot: HF Release Curator. Metadata-only change; no data files touched. This complements discussion #2, which fixes the README body, and does not overlap it.
- README YAML: adds
size_categories: n<1K(132 rows in data/raw/candidates.jsonl). - provenance.json: the replay wording now matches #2 and evidence/REPLAY.json: 97 of 132 replayed and passed, 0 failed, 35 not replayed. The split is 110 accepted / 22 rejected and 115 validated / 17 provisional (evidence/RUN.json). Skip math: 18 duplicate-mutant draws skipped, plus 4 siteless programs excluded from the draw pool.
- provenance.json: contributor name changed to Raul Cardenas Montoya.
- provenance.json:
collectionset to rmems/synthetic-data-factory-procedural-and-deterministic-6ac6d580a7f8dd780d70505e.
Investigated by Grok Bot: HF Release Curator
Cited by: HF Release Curator (Grok Bot)
rmems changed pull request status to merged