I'm not there yet but will keep it in mind. I would probably run same version with multiple params, is there an easy way to keep track of which params led to which CSV?
-
-
Replying to @DrVeronikaCH @vcheplygina
You can store them in the CSV and/or in the same of the CSV, really depends. Give me more details, e.g., how many params?
2 replies 0 retweets 0 likes -
Replying to @o_guest
I will explain once I can type on a real keyboard again instead of phone :P
1 reply 0 retweets 1 like -
-
Replying to @o_guest
I guess 5-10 or so, so far I have been using the name, but that becomes complicated when you add a new param, and it means all previous results had a default value... but the filename structure changes. Does that make sense?
1 reply 0 retweets 1 like -
Replying to @DrVeronikaCH @o_guest
I think there are some packages for managing this, came across this recently but haven't tried it (so many things to try out first!): http://sacred.readthedocs.io/en/latest/index.html …
2 replies 2 retweets 3 likes -
Replying to @DrVeronikaCH @vcheplygina
Oh, wow! Tell me more about it. Looks really interesting.
1 reply 0 retweets 0 likes -
Replying to @o_guest
I know just as much as you at this point. But it feels like it's something I would need. I want to have an easy lookup between params and results, also see easily which param combinations haven't run yet.
2 replies 0 retweets 1 like -
Replying to @DrVeronikaCH @o_guest
And if an experiment produces any figures etc, that the figures know which params/code version they were produced by
1 reply 0 retweets 1 like -
Oh, I see the documentation is pretty good! I'll read up! 
-
-
Replying to @o_guest @vcheplygina
I remember coming across Sumatra a few years ago, but this Reddit thread has many others I keep meaning to try!https://www.reddit.com/r/MachineLearning/comments/5gyzqj/d_how_do_you_keep_track_of_your_experiments/ …
0 replies 0 retweets 1 likeThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

