[18:36:55] tizianop: re T151969 have you enabled Kerveros authentication for Hadoop? [18:36:56] T151969: Tiziano Piccardi shell request + analytics-privatedata-users - https://phabricator.wikimedia.org/T151969 [18:37:47] tizianop: check https://lists.wikimedia.org/pipermail/wiki-research-l/2019-December/007038.html [18:38:03] leila: hi! :) yes, it worked fine with kinit before Christmas [18:39:40] tizianop: ok. then I'm of no help, sadly. :/ [18:40:40] tizianop: while you're here, btw: congratulations on the acceptance of the paper. You've put a /lot/ of work in that research and it's great to see it receive strong reviews and acceptance. I'm happy for you. [18:40:44] thank you! Luca is looking into it and there is something wrong with the exchange of the key [18:40:57] tizianop: ok. [18:41:08] thank you very much! :D [22:57:41] o/ isaacj [22:57:53] I'm looking for a good way to output vectors from a supervised fasttext run. [22:58:09] Do you know how to get the vectors in sorted order by word frequency? [23:00:14] if model is your trained model, then model.words is a list of words sorted by frequency [23:00:28] can't guarantee it but that certainly seems to be what it is going on [23:00:37] Awesome. So you just call get_word_vector() for each one then to get the vector? [23:11:50] yeah -- i was about to share some code that will do this on the phab task so will update that in ~30 minutes [23:14:50] Cool thank you :) [23:14:54] Also, don't work too late :P [23:35:09] Oooh. It trains so fast! [23:40:10] yeah, it'll spoil you with how fast it is and little fine-tuning it needs