xtof54 @xtof54@octodon.social

@McKael @kensanata OK thu I think it's not a good policy to increase instances activity. You're a go/baduk player ?

. Didn't knew this trick but I was no Twitter guru. So your reply does not show up on the local TL... :-/ That's still not crystal clear for me why... @kensanata

This is pretty bad, because you cannot have public "replies" then ?!

@CountZero Bonjour !

I think I've just understood something about #mastodon + #api: when you *reply* to an external toot, your reply just appears in the target Home TL; further adding @mention makes it also appear in the Notifs TL. But your reply is not listed neither in source nor target local TLs !?? Got to rethink my app then... :-/

Did you consider pytorch ?

Anyone know what's the licence / copyright of our toots ?

#sluc #nancy m'a laisse croire a un miracle a la mi-temps, mais non, encore decu en fin de match... :-( Ca sent la proB ! Mais comment reconstruire une equipe ? Quels bons joueurs vont vouloir venir ???

You may find here the main language (+ linguistic activity) of most active mastodon instances at noon CET on May, 8th 2017:

http://talc1.loria.fr/users/cerisara/

@CobaltVelvet Yes, it's nice because we look at them as psychedelic images... Now did you try more recent nets like hyper-resolution networks ? They're much more realistic, which makes them also more disturbing... You know, this 'uncanny valley' stuff...

Agree ! Even little ticks ! ;-)

Eux aussi avaient honte, interieurement. Ou alors ils etaient stupides ! On juge les autres par leur apparence jusqu'a 6 ans. Ensuite, le cerveau prend le relai. Enfin, normalement... ;-)

Testing new version of app with workaround for android ssl bug

Les apps, c'est mieux je trouve...

@enkiv2 you did ? ;-) well, I guess alone against crowd is not enough. But we're so far to have good generation; would require lots of smart things: semantics, dialog model, common sense, world model... can't be done with just an lstm, whichever nb of layers are stacked...

@enkiv2 yep thats it: microsofts one :-)

@enkiv2 would be crap anyway... and you have adversarial *user*: remember facebook's bot ? can't get its name...

@enkiv2 ok :-) would not have anyway enough data per user, would need to merge user toots or train on large twitter corpus and just adapt the bot based on the few usrr toots ;-)

@enkiv2 why Markov ?

si, tres bientot j'y mettrai mousetodon :-)