llamia-chess-data / README.md
ssingh22's picture
Upload dataset
5e6aea8 verified
|
raw
history blame
1.55 kB
metadata
dataset_info:
  - config_name: LichessValidMultiEvaluationLines
    features:
      - name: fen
        dtype: string
      - name: multi_evals
        dtype: bool
      - name: meta
        struct:
          - name: depth
            dtype: int64
          - name: knodes
            dtype: int64
      - name: pvs
        list:
          - name: cp
            dtype: int64
          - name: line
            dtype: string
          - name: mate
            dtype: int64
      - name: num_lines
        dtype: int64
    splits:
      - name: train
        num_bytes: 16902444248
        num_examples: 82976204
    download_size: 7945561664
    dataset_size: 16902444248
  - config_name: projector_align
    features:
      - name: fen
        dtype: string
      - name: meta
        struct:
          - name: ASK_LEGAL_MOVES
            dtype: bool
          - name: SAN_LINES
            dtype: bool
          - name: TABULAR_DESC
            dtype: bool
          - name: VERBALIZE
            dtype: bool
          - name: depth
            dtype: int64
          - name: knodes
            dtype: int64
      - name: id
        dtype: string
      - name: conversations
        list:
          - name: from
            dtype: string
          - name: value
            dtype: string
    splits:
      - name: train
        num_bytes: 15979867893.746607
        num_examples: 19971226
      - name: test
        num_bytes: 16002891.253392864
        num_examples: 20000
    download_size: 3751712090
    dataset_size: 15995870785
configs:
  - config_name: LichessValidMultiEvaluationLines
    data_files:
      - split: train
        path: LichessValidMultiEvaluationLines/train-*
  - config_name: projector_align
    data_files:
      - split: train
        path: projector_align/train-*
      - split: test
        path: projector_align/test-*