sent-conv-torch icon indicating copy to clipboard operation
sent-conv-torch copied to clipboard

Published 20 hours ago • verified publisher icon harvardnlp

Python3 Preprocessing and UTF-8 decoding of GoogleNews-vectors correction

Open gargrohin opened this issue 5 years ago • 0 comments

Updated preprocess.py to Python3. Corrected reading of characters from GoogleNews-vectors-negative300.bin by adding .decode("utf-8"), which was not working before, at least in Python3. Tested on custom datasets.

gargrohin avatar Jan 26 '20 19:01 gargrohin