We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
cnn audio-superres in -> conv layers -> self-attention -> upsample -> out