Amphion icon indicating copy to clipboard operation
Amphion copied to clipboard

Published 20 hours ago • verified publisher icon open-mmlab

[Help]: Is there any loss that linearly correlate to performance of TTA autoencoder?

Open Jiang-Stan opened this issue 10 months ago • 3 comments

Problem Overview

(Briefly and clearly describe the issue you're facing and seeking help with.) I tried to reproduce the TTA recipe, but it seems that it is hard to assess performance through any loss.

Thanks in advance for your sicerely help!

Jiang-Stan avatar Apr 17 '24 09:04 Jiang-Stan

Can you share training detail of TTA autoencoderKL? I tried to reproduce the result with the origin repo with your preprocessed mel data and training configs, but I failed to get result similar to opensource ckpt. Or is there any way determining performance of ckpts of different iter?

Jiang-Stan avatar Apr 19 '24 03:04 Jiang-Stan

Hi, the detail can be found in https://github.com/open-mmlab/Amphion/blob/main/egs/tta/autoencoderkl/exp_config_latent_4_10_78.json

HeCheng0625 avatar Apr 29 '24 07:04 HeCheng0625

So the rest of the parameters are inherited from config/autoencoderkl.json, right?

fpicetti avatar May 06 '24 22:05 fpicetti