Encoding bug with word_language_model


I’m new to PyTorch and in order to get started I decided to download this word language model directly from the github of the PyTorch team --> https://github.com/pytorch/examples/tree/master/word_language_model

I have a problem : the output I get (the language model) is a bunch of weird symbols. This is clearly a sign of an encoding problem.

A colleague of mine told me that the problem could be one of the following (or maybe more than one of the following combined) :

  • incompatibility of Python versions
  • incompatibility of PyTorch versions
  • incompatibility of Pickle versions

So I checked the requirements file but it only states that you need torch and nothing more (no specifications on the versions I have to use).

Could you please help me ?

Thanks a lot !