Question regarding the pre-trained checkpoint

by soujanyaporia - opened Mar 31, 2023

Mar 31, 2023

Hi,

Excellent work! Thanks. Can I ask how are the checkpoints shared in this repo obtained? Did you follow the complete process as explained in the paper?

Also, could you let me know if there is a trainer code?

Thanks.

-Soujanya

sanchit-gandhi

Centre for Vision, Speech and Signal Processing - University of Surrey org Mar 31, 2023

Hey @soujanyaporia - this checkpoint is obtained by converting the weights from the official checkpoint (under ckpt). The code is made to match the results of the official implementation one-to-one: https://github.com/haoheliu/AudioLDM. Thus, the model is entirely equivalent to the one released with the AudioLDM paper.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment