2 ms·
This looks really impressive, do you have any writeup on how you coded all of it?
by lpellis 6y ago
This looks really impressive, do you have any writeup on how you coded all of it?
- echelon 6y agoNot yet, but that's something I can put together! There are a lot of short cuts I wish I'd known at the outset, mostly in terms of curating training data and monitoring the learning, but I also learned a few new systems-y things from the overall engineering project.
- lpellis 6y agoI'd love to read that if you ever put it together! How many hours of audio did it need to train one of the voices?
- echelon 6y agoIt varied a lot. Some speakers (that sound great) only had about 45 minutes of audio. Other speakers had up to five hours. The key to training is that all of the models were transfer learned from the Linda Johnson speech dataset (LJS).