---
dataset_info:
  features:
  - name: data_id
    dtype: string
  - name: prompt_ids
    sequence: int64
  - name: complete_training_sequence_ids
    sequence: int64
  - name: prompt_ids_len
    dtype: int64
  - name: traj_position_indices
    sequence: int64
  splits:
  - name: train
    num_bytes: 3461344541
    num_examples: 250619
  download_size: 703970632
  dataset_size: 3461344541
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---