search

p0p4k / vits2_pytorch

unofficial vits2-TTS implementation in pytorch
https://arxiv.org/abs/2307.16430
MIT License
477 stars 85 forks source link

Keep training on existed checkpoint #76

Closed HuuHuy227 closed 7 months ago

HuuHuy227 commented 7 months ago

Hi, your repo is awesome. I want to know how could i keep training on an existed checkpoint. I can't find any config show that way.

p0p4k commented 7 months ago

I think the same model name (-m) resumes the training.

JohnHerry commented 7 months ago

Hi, your repo is awesome. I want to know how could i keep training on an existed checkpoint. I can't find any config show that way.

Do you mean to keep the schedulers saved into checkpoint and then loaded during resume from the checkpoint?

HuuHuy227 commented 7 months ago

Hi, your repo is awesome. I want to know how could i keep training on an existed checkpoint. I can't find any config show that way.

Do you mean to keep the schedulers saved into checkpoint and then loaded during resume from the checkpoint?

Yes. I mean that I had a checkpoint I've trained before and now I want to continue train on that checkpoint.