bohdanchap

If you use a python 3 and the latest PyTorch you can read the classifier by specifying additional encoding parameter that escapes Unicode bytes: `classifier_ckpt = torch.load('model/classifier.ckpt', encoding= 'unicode_escape')` More...