Open cao-nv opened 2 years ago
Did you ever figure out a solution? I'm getting a different error UnicodeDecodeError in repositories_dataset.py", line 67, but really just looking for an updated version of the model using TF 2, so I can convert it to CoreML
Impressed by the simplicity and accuracy of GuessLang model, but the model is built on Tensorflow Estimator which is out-dated and cumbersome to use in other language, I want to update the GuessLang model to TF 2.0 with better architecture for both Python and other language for deployment, and hopefully a higher accuracy. I got this error when running GuessLangTools to download and prepare the dataset. How can I get around this problem? For example: manually add TOML repos, skip this language....
Thanks