seq2seq-chatbot icon indicating copy to clipboard operation
seq2seq-chatbot copied to clipboard

how to train my own data

Open DhruvMevada opened this issue 6 years ago • 4 comments

How any I chane the dataset . Can I take the dataset of mine rather then twitter data & corpus movie data if yes then how?

DhruvMevada avatar Feb 22 '18 06:02 DhruvMevada

hi, if you like to train your own data, you should make sure the data import to python script have the same format with the twitter data we are using.

Alternatively, you can prepare the file with the same format with twitter data, and import it to the python program in the same way.

zsdonghao avatar Mar 10 '18 15:03 zsdonghao

Either programming or data preparation.

wagamamaz avatar Mar 12 '18 14:03 wagamamaz

DhruvMevada you can train your chatbot on your data also,for that purpose you have to do data preprocessing before training your chatbot. you have to make the use of field separator in your data for separating the text from ID's . your data should be in csv format.And please do data cleaning on your data to get the best result for example you can see my respository also to see how data preprocessing is done while using data in chatbot

vishal2612200 avatar Aug 15 '18 07:08 vishal2612200

@zsdonghao What format would the data be in. I have comments in one file and replies in the other file.

jessedoka avatar Apr 16 '21 14:04 jessedoka