Datasets:

Modalities:
Image
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
License:
flair / README.md
fgranqvist's picture
Upload dataset
19f45d5 verified
|
raw
history blame
772 Bytes
metadata
license: cc-by-nc-4.0
dataset_info:
  features:
    - name: user_id
      dtype: string
    - name: image_id
      dtype: int64
    - name: labels
      sequence: string
    - name: partition
      dtype: string
    - name: fine_grained_labels
      sequence: string
    - name: image
      dtype: image
  splits:
    - name: train
      num_bytes: 4658458558.710612
      num_examples: 345879
    - name: test
      num_bytes: 573966691.0054843
      num_examples: 43960
    - name: val
      num_bytes: 507973541.4059026
      num_examples: 39239
  download_size: 5677054504
  dataset_size: 5740398791.122
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: val
        path: data/val-*