search

NVIDIA / waveglow

A Flow-based Generative Network for Speech Synthesis
BSD 3-Clause "New" or "Revised" License
2.29k stars 530 forks source link

Continue Training from a checkpoint saved in checkpoints folder #267

Open pgovil1998 opened 2 years ago

pgovil1998 commented 2 years ago

How can we continue the training of a model from a checkpoint saved in the folder. Please write back as soon as possible. It will be of great help.

msalhab96 commented 2 years ago

Just add the path of the checkpoint to this line https://github.com/NVIDIA/waveglow/blob/8afb643df59265016af6bd255c7516309d675168/config.json#L11