metadata
dataset_info:
features:
- name: source
dtype: string
- name: target
dtype: string
- name: dialect
dtype: string
splits:
- name: train
num_bytes: 51504403
num_examples: 125608
- name: validation
num_bytes: 5639442
num_examples: 13755
- name: test
num_bytes: 76972
num_examples: 200
download_size: 29932609
dataset_size: 57220817
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
Mirror created 2026-10-11 by the JoTalbot/ukraine pipeline. Original source: https://huggingface.co/datasets/DenysKovalML/ukrainian-dialect-normalization — please credit and respect the original dataset's license.