Home
Projects
Resources
Alternatives
Blog
Sign In
Constant Memory Waveglow Versions
Save
PyTorch implementation of NVIDIA WaveGlow with constant memory cost.
Overview
Versions
Reviews
Resources
v1.2
3 years ago
Changelog
v1.2 - 2020-12-18
Features
Add new model 'MelGlow'.
Add mixed precision training option.
Bug Fix
Fix issue #3 .
v1.1
3 years ago
Major updates:
A modified implementation of Baidu's WaveFlow is added
Remove normalization in dataloaders
Display log likelihood score when inferencing
A memory usage record of waveglow is added, which was run on a RTX 2080 machine.
v1.0.1
3 years ago
v1.0
3 years ago
Initial release.
Home
Projects
Resources
Alternatives
Blog
Sign In
Sign In to OSA
I agree with
Terms of Service
and
Privacy Policy
Sign In with Github