kkitt/more_russian_homophones

--- dataset_info: features: - name: source_row dtype: int64 - name: token_index dtype: int32 - name: homophone dtype: string - name: ipa dtype: string - name: lemma dtype: string - name: pair_id dtype: int64 - name: text dtype: string - name: sample_rate dtype: int32 - name: audio dtype: audio splits: - name: train num_bytes: 23607029252.0 num_examples: 54011 download_size: 16470000090 dataset_size: 23607029252.0 configs: - config_name: default data_files: - split: train path: data/train-* ---