search

daniel-kukiela / nmt-chatbot

NMT Chatbot
GNU General Public License v3.0
385 stars 213 forks source link

Outputfile in scoring.py #108

Open SofTro opened 5 years ago

SofTro commented 5 years ago

Dear Daniel, I am trying to understand where and how the file 'full_some_questions-81k.out' is created that is part of scoring.py. Could you describe where it is created and what the file producing it should contain in particular?

Thanks!

daniel-kukiela commented 5 years ago

Hi, Could you describe your problem more, please? full_some_questions-81k.out is nothing my code outputs, and none of data files is related to scoring.py as well, so I'm not sure what you mean.

SofTro commented 5 years ago

Dear Daniel, thank u for the response! Basically, Harrison explains how to set up a scoring.py script in his tutorial (I think its part 9) where he sets additional scoring parameters for the outputted data. I think it was part of sentdex lab before. And in this scoring.py he scores 'full_some_questions-81k.out' whereafter he gets a more clean output without "https://", corrections for bad responses, messed up links, unk-checker etc. And the code basically starts with:

if name == 'main':

name = 'full_some_questions-81k.out

with open(name,'r',encoding='utf8') as f:
    content = f.read().split("\n\n\n")
    for content in contents[0:-1]:
        batches = content.split(">>>")
....

And at this point I am searching the outputted data that should be used. I also tried to use the outputted data from the model folder "outputdev..."

Any hints would be very much appreciated and great job in putting these things together.