w2s_dataset

--- dataset_info: features: - name: write format dtype: string - name: speak format dtype: string splits: - name: train num_bytes: 13223140117 num_examples: 16043579 downlo