lucidrains / voicebox-pytorch

Implementation of Voicebox, new SOTA Text-to-speech network from MetaAI, in Pytorch
MIT License
589 stars 49 forks source link

Fixed a bug where the scheduler would get None #30

Closed wassimseif closed 11 months ago

wassimseif commented 11 months ago

Fixed a bug where the scheduler would get None for T_max if user is passing num_epochs instead of num_train_steps.

Without this fix the user would get an exception

lucidrains commented 11 months ago

thanks!