---
dataset_info:
  features:
  - name: video
    dtype:
      video:
        decode: false
  - name: model_output
    dtype: string
  - name: input_ids
    sequence:
      sequence: int32
  - name: attention_mask
    sequence:
      sequence: int8
  - name: pixel_values_videos
    sequence:
      sequence:
        sequence:
          sequence:
            sequence: float32
  splits:
  - name: train
    num_bytes: 2178412096.3333335
    num_examples: 50
  - name: test
    num_bytes: 566446612.6666666
    num_examples: 13
  download_size: 411680026
  dataset_size: 2744858709.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---