This dataset is a compact, ML-oriented chess game dataset generated from
angeluriot/chess_games
plus additional 2025 games extracted from
Lumbras Giga Base.
It keeps only the move sequence in UCI notation, a derived minimum Elo field,
and a source label. This makes the dataset smaller and simpler to consume than
the base dataset when the downstream task only needs legal move sequences and a
coarse player-strength signal.
Games: 15… See the full description on the dataset page:
https://huggingface.co/datasets/mrmlengineer/chess_games.
