---
dataset_info:
  features:
  - name: img
    dtype: image
  - name: label
    dtype: int64
  - name: keywords
    dtype: string
  - name: ti_indices
    dtype: int64
  splits:
  - name: train
    num_bytes: 23138408.0
    num_examples: 10000
  - name: test
    num_bytes: 4646029.0
    num_examples: 2000
  download_size: 29212913
  dataset_size: 27784437.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---