as-ideas / DeepPhonemizer

Grapheme to phoneme conversion with deep learning.
MIT License
352 stars 38 forks source link

Add validation loss to log #15

Closed Frank995 closed 2 years ago

Frank995 commented 2 years ago

I think it would be good to also print the validation loss after each epoch to monitor overfitting. I'm not an expert of tqdm and I wasn't able to change the description bar after the for loop, but I left the code anyway.

cschaefer26 commented 2 years ago

Hi, thanks for the PR. The validation loss is already implemented, here: https://github.com/as-ideas/DeepPhonemizer/blob/b8f170764c7648fe2acb552b787099ab4f941e58/dp/training/trainer.py#L130