Joseph Turian

Actually I am an nnAudio user :)

@ABC0408 Just curious if you post a colab, to see what the performance differences are there

@ashleve Can you explain why `ddp_spawn` is better than `ddp`? (I skimmed the issues but couldn't grok it.) I understand from the lightning docs that spawn is slower than ddp

@ashleve Just curious because I am using hydra + DDP in a current project. How would I be able to detect if this issue is occurring for me? What evidence...

@ashleve woof that's gross. If you have a good fix, we might considering seeing if we can push it upstream to lightning.

@ashleve Lightning team appears to be working on this issue? https://github.com/Lightning-AI/lightning/pull/11617#issuecomment-1245842064 I've been lightly commenting in that PR

@ashleve This is cool but what if your artifacts are stored in wandb?

Hi @btrude I have a few questions about how to generate VQ-VAE codes without the priors. You can email me at username at gmail.com

@lostanlen `inv_frame` is good because it is similar to `frame`.

Tensorflow uses overlap_and_add https://www.tensorflow.org/api_docs/python/tf/signal/overlap_and_add