search

huggingface / parler-tts

Inference and training library for high-quality TTS models.
Apache License 2.0
2.6k stars 265 forks source link

Save checkpoints as usable models #37

Open deguodedongxi opened 2 weeks ago

deguodedongxi commented 2 weeks ago

Hey everyone,

I am trying to fine-tune a model. I ran into overfitting after some training. Now I want to save a previous checkpoint as my model. As far is a can see, you are using safetensor models when using the "ParlerTTSForConditionalGeneration.from_pretrained()" method.

I cannot find an easy way how to load and save a checkpoint without starting a new training. Do you have any suggestions?

Thank you :)

ylacombe commented 1 week ago

Hey @deguodedongxi, you should simply copy the config.json and generation_config.json to the checkpoint folder you want, then you can load the model using the checkpoint folder path as your model name! let me know if that helps