timmyburn/hexo-bootstrap-corpus
Hexo Human Corpus Encoding-free corpus of decisive human Hex Tac Toe games — hexagonal grid, six-in-a-row to win (player 1 opens with 1 move, then both players play 2 moves per turn; the board is theoretically infinite). Each line is one game as a raw axial move list + outcome. Nothing about any neural-network encoding is baked in — no planes, no fixed board size, no action space. Read it with the stdlib json module and build whatever representation you want. Files… See the full description on the dataset page: https://huggingface.co/datasets/timmyburn/hexo-bootstrap-corpus.
Sync corpus from hexo-pi scraper: 8300 -> 8698 human games (+398)
Sync corpus from hexo-pi scraper: 8300 -> 8698 human games (+398)
Make SCHEMA.md static (no game count; restore curated card)
Make README.md static (no game count; restore curated card)
Update README.md — 8300 games (scraper pull 2026-06-18)
Update SCHEMA.md — 8300 games (scraper pull 2026-06-18)
Update dataset_metadata.json — 8300 games (scraper pull 2026-06-18)
Update hexo_human_corpus.jsonl — 8300 games (scraper pull 2026-06-18)
Update README.md
Retire encoding-specific corpus bootstrap_corpus_v7.npz (superseded by hexo_human_corpus.jsonl)
Retire encoding-specific corpus bootstrap_corpus_v6w25.npz (superseded by hexo_human_corpus.jsonl)
Retire encoding-specific corpus bootstrap_corpus_pretrain_v6.npz (superseded by hexo_human_corpus.jsonl)
Retire encoding-specific corpus bootstrap_corpus.npz (superseded by hexo_human_corpus.jsonl)
Fix SCHEMA.md: unescaped pipes in elo row broke the table
Add encoding-free human corpus JSONL + docs (README.md)
Add encoding-free human corpus JSONL + docs (dataset_metadata.json)
Add encoding-free human corpus JSONL + docs (SCHEMA.md)
Add encoding-free human corpus JSONL + docs (hexo_human_corpus.jsonl)
Upload bootstrap_corpus.npz with huggingface_hub
Upload bootstrap_corpus_v6w25.npz with huggingface_hub
overwrite canonical with v7 (Elo-weighted, §148)
v7 corpus: Elo-weight fix (per-position weights baked into NPZ for WeightedRandomSampler)
add v7 corpus: human-only Elo-weighted (§148, 6259 games / 353091 positions, 8-plane)
bootstrap_corpus.npz: v6 human-only 8-plane (353k pos, 6259 games, f16)
add v6 human-only pretrain corpus (353k pos, 8-plane, 6259 games)
Update Corpus to V6
add dataset card
Upload bootstrap_corpus.npz with huggingface_hub
initial commit