| from datasets import load_dataset | |
| from tqdm import tqdm | |
| subsets = [('bg', 'lv'),('cs', 'lv'),('da', 'lv'),('de', 'lv'),('el', 'lv'),('en', 'lv'), | |
| ('es', 'lv'),('et', 'lv'),('fi', 'lv'),('fr', 'lv'),('hr', 'lv'),('hu', 'lv'), | |
| ('it', 'lv'),('lt', 'lv'),('lv', 'mt'),('lv', 'nl'),('lv', 'pl'),('lv', 'pt'), | |
| ('lv', 'ro'),('lv', 'ru'),('lv', 'sk'),('lv', 'sl'),('lv', 'sv')] | |
| with open('tilde_lv.txt', 'w') as file: | |
| for subset in tqdm(subsets): | |
| print(f'Downloading subset: {subset[0]} - {subset[1]}') | |
| dataset = load_dataset('tilde_model', lang1=subset[0], lang2=subset[1], split='train') | |
| for entry in tqdm(dataset): | |
| file.write(f'{entry["translation"]["lv"]}\n') | |
| file.close() | |
| print('Done') |