bigartm-users

Hi I try to run ARTM for getTheta() and 'TopTokens' using python master = mc.MasterComponent(

Hi all! Thank you very much for replies. Oleksandr, if this ability will be in the API, that will be

>The idea is to tell the model that terms w 1 1 ,…w 1 s1 always belong to class1, w 2 1 ,…w 2 s2

Thanks среда, 17 июля 2019 г., 12:42:58 UTC+3 пользователь Елена Реутт написал: Hello. Tell me,

Thank you for your response, but unfortunately not. I'll try to understand среда, 17 июля 2019 г.

Hello. I need help with understanding a meaning of coherence scores in BigArtm score tracker. Every

Hi, try to set theta_columns_naming parameter of ARTM model constructor to 'title'. --- С

Sorry, I've found the error! It works fine now. It should be: model_artm = artm.load_artm_model(

Hi, You need to pre-process collection and convert it into VW format. VW format looks as follows:

Hello! Is there any way to extract nwd matrix from ARTM model instance in Python API? I need this

Hi, I followed the installation instructions from https://bigartm.readthedocs.io/en/stable/

Alex, Murat, Thank you so much! Kind regards, Julia среда, 4 июля 2018 г., 18:13:33 UTC+3

Ah, I see, I was wrong then. Again, thank you very much!

Hi Nikolay, Thanks for you message. Actually, not sure whether we need yet another place for

Could you please clarify what do you mean when say "titles for documents are located far enough

Hi there There is BigARTM docker image here: https://github.com/bigartm/bigartm-docker I was facing

Hi BigARTM v0.9.0 is now released! For windows, download new release from https://github.com/bigartm/

No problem! Just ask if anything isn't clear enough - quit often BigARTM's documentation is

Thank you much! Know it's clear. Will try further. On Wednesday, July 19, 2017 at 3:41:05 PM UTC+

Yes, it would be a good feature. Please create an issue in our github repository, I'll fix it in

I found bug in my batch creation procedure (local batch dictionary). I processing batch protobuf

Hi! Yes, it works! Thank you! It takes time, until fit_offline ended. Thank you! вторник, 28 февраля

Hi Max, I guess it might be because you didn't specify class_id argument when creating top tokens

Hi BigARTM v0.8.3 is now released! For windows, download new release from https://github.com/bigartm/

Hi, There are some really interesting features in BigARTM that are described in theoretical papers

Hi, Yes, this is possible --- though we need to setup a better API. Currently you have to do it like

Hi BigARTM v0.8.2 is now released! For windows, download new release from https://github.com/bigartm/

Hi Svetozar, It was a long time ago since we discussed adding/removing topics from an existing model

Hi Anastasia, No, initialization of theta matrix is hardcoded in C++ code and is not available

Please keep conversations in English when communicating to bigart...@googlegroups.com. Thanks!