Rasa: Multiple cores when training

Created on 7 Feb 2017  路  2Comments  路  Source: RasaHQ/rasa

I know it is an inherent limitation of Python but is it possible to make the situation slightly better with the current setup? We have 32 cores machines on Digital Ocean that we are hoping to use for experimenting with Rasa NLU.

picture azzuwan

Most helpful comment

@azzuwan which backend are you using, spacy or mitie?

picture tmbo on 7 Feb 2017
❤1 😄1 👍1

All 2 comments

@azzuwan which backend are you using, spacy or mitie?

tmbo picture tmbo on 7 Feb 2017
❤1 😄1 👍1

@tmbo Brilliant! That commit is candy to me. I'm using MITIE. Going to try ASAP. Thanks!

azzuwan picture azzuwan on 7 Feb 2017
Was this page helpful?
0 / 5 - 0 ratings

Related issues

lomarceau picture lomarceau  路  3Comments
anishrav picture anishrav  路  3Comments
irfan-zoefit picture irfan-zoefit  路  3Comments
alonsopg picture alonsopg  路  3Comments
nahidalam picture nahidalam  路  3Comments