AudioLDM is a text-to-audio generation model using latent diffusion that creates speech, sound...