skol101
skol101
@v-nhandt21 I simply added audiomentations (RoomSimulator, Backgroundnoise) and stored to separate directory, which I use in "input_wavs" params after step for postnet training (250k in my case, though it's 500k...
Yep, doesn't work for multi gpu training.
This does a pretty good job of removing noises from the speech https://github.com/Rikorose/DeepFilterNet
@sh-lee-prml hello, could you share the results of your E2E training?
Cheers, I'll test it out.
Yep, until all spect.pt files are generated, the training must be run on a single GPU
Could you provide link to what code exactly ?
So, you're saying this code must be UNcommented?
Same here. I'm using my dev machine behind the cloudflare tunnel. I can connect to it just fine by using domain name setup in the tunnel, but wakeonlan gives me...
Seconded, also intersted in adding custom layer/widget to OSD on the GCS.