Julia Kreutzer

Hi @tarekzaki115, yes you can resume on any machine, as long as you copy over checkpoint and vocabulary files and adjust the config to read these files in and to...

search tests in #72

This is awesome! I think Chris Emezue has already done that for the Masakhane JoeyNMT models.

https://huggingface.co/spaces/chrisjay/masakhane-benchmarks

Awesome, thank you @chrisemezue! Can you send a PR?

Thanks! I requested some changes, once they are resolved we can merge.

Thanks @KashyapKishore! `build_vocab` has changed a lot with v2.0. As @may- described, you now pass it the config directly as argument, and it will read the data paths and tokenizer...

Yes, that's what it does. Since we have it for the other scripts as well, I think doesn't hurt to add it.

Yes, a "resume" option in addition to overwriting would be helpful. It would require the following: 1. Continue logging into the same log file. Or rename the existing one. 2....

That looks good. In the 3x TRUE case, we could throw a warning and set overwrite=False, as I assume the case you described would be the best reason for having...