CTC-based Non-autoregressive Textless Speech-to-Speech Translation