---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: CLIP_text_latent
    sequence: float32
  - name: SD_VAE_image_latent
    sequence:
      sequence:
        sequence: float32
  splits:
  - name: train
    num_bytes: 57507528731.75
    num_examples: 380530
  download_size: 60531502833
  dataset_size: 57507528731.75
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---