Datasets:

samaritan-ai
/

hebrew_synth_lines

dataset_info:
  features:
    - name: image
      dtype: image
    - name: text
      dtype: string
    - name: format
      dtype: string
  splits:
    - name: train
      num_bytes: 13099604057.927
      num_examples: 133819
    - name: validation
      num_bytes: 1561427305.097
      num_examples: 15743
    - name: test
      num_bytes: 775148754.804
      num_examples: 7871
  download_size: 15576770771
  dataset_size: 15436180117.828001
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
license: mit
language:
  - he
  - arc
task_categories:
  - text-generation