I think an important gap in the documentation is how to train the models. It shouldn't be necessary to download an 8GB tar file to make this project go!
The README would be improved by a very short "# Training" section, even perhaps just pointing to the correct notebook or script, with advice on whether any preparation is needed to run it.
I think an important gap in the documentation is how to train the models. It shouldn't be necessary to download an 8GB tar file to make this project go!
The README would be improved by a very short "# Training" section, even perhaps just pointing to the correct notebook or script, with advice on whether any preparation is needed to run it.