unless of course an urgent dumpling mission comes about. :)
2019-04-03 09324, 2019
Mr_Monkey
Data dumpl.
2019-04-03 09345, 2019
iliekcomputers
Not right now, gonna get dinner and catch up on some other stuff.
2019-04-03 09320, 2019
ruaok
I'll just dump what I got in my brain. when you have the bandwidth take a look, ok?
2019-04-03 09327, 2019
iliekcomputers
Yup, sounds good.
2019-04-03 09339, 2019
iliekcomputers
From what I remember off the AB migration
2019-04-03 09347, 2019
iliekcomputers
* create consul template
2019-04-03 09303, 2019
iliekcomputers
* make sure data dumps are sane
2019-04-03 09305, 2019
ruaok
shit, I've just been ordered on dumpling missing.
2019-04-03 09307, 2019
ruaok
misson.
2019-04-03 09328, 2019
iliekcomputers
* run a new.bb on hetzner
2019-04-03 09318, 2019
iliekcomputers
There will need to be a separate docker image for production, I guess.
2019-04-03 09312, 2019
ruaok
meh, I'll do the mission tomorrow. a bit too far for tonight.
2019-04-03 09313, 2019
iliekcomputers
Anyways, I'll get dinner and come back to this later tonight.
2019-04-03 09316, 2019
ruaok focuses on the doc
2019-04-03 09341, 2019
Mr_Monkey
Bon appétit iliekcomputers
2019-04-03 09307, 2019
ruaok
ok, I think we've captured what needs doing for BB.
2019-04-03 09347, 2019
ruaok has a pointy haired boss moment.
2019-04-03 09357, 2019
ruaok
I added "make google doc" to a google doc.
2019-04-03 09335, 2019
Mr_Monkey
ruaok: Cool ! Frist step :)
2019-04-03 09341, 2019
Mr_Monkey
First, even.
2019-04-03 09305, 2019
djwhitey_ joined the channel
2019-04-03 09326, 2019
djwhitey has quit
2019-04-03 09353, 2019
aliamo joined the channel
2019-04-03 09353, 2019
ruaok
yeah, knowing iliekcomputers he'll be done with it before sundown tomorrow.
2019-04-03 09304, 2019
ruaok
or 4am. whichever comes first in india.
2019-04-03 09316, 2019
ruaok is a wee confused on timezones atm.
2019-04-03 09332, 2019
code_master5
Mr_Monkey: A question about BB database 🙋♂️
2019-04-03 09351, 2019
code_master5
Why does BB database in docker image doesn't show tables when we hit `\d` command ? Why do we need to access tables using syntax like `bookbrainz.<table_name>` ?
2019-04-03 09308, 2019
code_master5
* docker container
2019-04-03 09350, 2019
code_master5
*two questions 😅
2019-04-03 09319, 2019
aliamo has quit
2019-04-03 09339, 2019
Mr_Monkey
code_master5: I don't know what to tell you. My SQL-fu is not strong
2019-04-03 09305, 2019
alastairp
code_master5: bookbrainz. is a schema, postgres probably needs to be configured to also know about that schema
once that does, we should have space for BB. well, we probably have space now anyway.
2019-04-03 09337, 2019
pristine__
ruaok: I can try and replace RDDs with Dataframe because that is what we are using (there is a lot of usage of rdd in the code that troubles me) or do you have anything else for this week?
"Recommendations generated for Vansika Pareek in 28.22102117538452" -> "Recommendations generated for Vansika Pareek in 28.2s"
2019-04-03 09358, 2019
ruaok
can you please run one for user "rob"? I'm curious.
2019-04-03 09344, 2019
ruaok
loads of info there, just hard to consume.
2019-04-03 09310, 2019
ruaok
I think we ought to split the model training data and save it to a file that clearly identifies the model.
2019-04-03 09322, 2019
ruaok
and then reference that model clearly in the generated playlist.
2019-04-03 09357, 2019
pristine__
These results are for a small dataset
2019-04-03 09324, 2019
amCap1712
hi all
2019-04-03 09346, 2019
pristine__
I haven't tested on the whole dataset, but it would be more then 2 min for every user.
2019-04-03 09352, 2019
pristine__
amCap1712: hey:)
2019-04-03 09357, 2019
amCap1712
i was working on the proposal and have a doubt.
2019-04-03 09305, 2019
aliamo joined the channel
2019-04-03 09310, 2019
amCap1712
hi pristine__ . how's everything going on?
2019-04-03 09303, 2019
pristine__
Okayish. Machines are having hard time learning with me.
2019-04-03 09306, 2019
pristine__
Lol
2019-04-03 09323, 2019
code_master5
pristine__: 🤣
2019-04-03 09324, 2019
amCap1712
😂 That was a good one
2019-04-03 09346, 2019
pristine__
😆
2019-04-03 09325, 2019
amCap1712
yvanzo, ruaok: My doubt is should I embed my video mockups directly in the pdf which will require flash to play or should i upload the video to youtube and just provide the links in the pdf
2019-04-03 09304, 2019
yvanzo
Which video?
2019-04-03 09315, 2019
yvanzo
(link?)
2019-04-03 09330, 2019
pristine__
ruaok: training data is a df? We want to save the df, but why?
2019-04-03 09301, 2019
ruaok
I have no idea.
2019-04-03 09301, 2019
amCap1712
yvanzo: the links to youtube videos of the mockups i provide in the proposal on community.meb.o
2019-04-03 09340, 2019
pristine__
we are saving the model anyway (as of now). I am not sure if saving the df would help in any way.
2019-04-03 09344, 2019
amCap1712
on the forums, it is automatically embedded. so no probs there but what about the final proposl
2019-04-03 09327, 2019
ruaok
possible faster than re-collecting the data each time you want to train the model?
2019-04-03 09331, 2019
pristine__
For user 'rob' , I will run the script on leader so that we can have an idea of the actual time
2019-04-03 09334, 2019
pristine__
Tomorrow?
2019-04-03 09351, 2019
pristine__
We will train the model only when we have new listens (data) and for that we will have to recollect data (eliminating some data and adding some data)
2019-04-03 09330, 2019
pristine__
In any case saving the training data isn't much of use.
2019-04-03 09308, 2019
ruaok
well, when we're in production, saving the DF may not make sense.
