---
dataset_info:
  features:
  - name: top_acts
    sequence:
      sequence:
        sequence: float32
  - name: top_indices
    sequence:
      sequence:
        sequence: int32
  - name: word_splits
    sequence: string
  - name: code
    dtype: string
  - name: input_ids
    sequence: int32
  - name: lang
    dtype: string
  splits:
  - name: train
    num_bytes: 317814290374
    num_examples: 50000
  download_size: 105945545037
  dataset_size: 317814290374
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---