Downloaded the pre-trained English-German Transformer from http://opennmt.net/Models-py/, and when I unzipped it got a .pt and a .model file. Then, I fed the path of the .pt file into the command given http://opennmt.net/OpenNMT-py/FAQ.html#how-do-i-use-the-transformer-model (with the --train_from flag) and got the error in the title. From some searching, this seems to be a Windows problem, but I'm not sure how to resolve it. I would like to load the pre-trained model using the state_dict given in .pt, but I'm not sure how to make sure that it fits the same parameters of the original model without using the train.py script.
Please use the forum for this type of general question. Also read more previous post, and give exact command lines you used + trace of your errors.
However not sure about the windows problem.