README.md

metadata

language:
  - en
license: mit
task_categories:
  - image-classification
  - image-segmentation
pretty_name: BirdSnap Lite
tags:
  - biology
dataset_info:
  features:
    - name: image
      sequence:
        sequence:
          sequence: float32
    - name: common
      dtype: string
    - name: scientific
      dtype: string
    - name: common_idx
      dtype: int64
    - name: scientific_idx
      dtype: int64
  splits:
    - name: train
      num_bytes: 1904373787
      num_examples: 38096
    - name: test
      num_bytes: 92505076
      num_examples: 1851
    - name: val
      num_bytes: 349937
      num_examples: 7
  download_size: 1989604716
  dataset_size: 1997228800
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: val
        path: data/val-*

This is a version of BirdSnap that will be easier on your free Google Colab quota.