Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
Is There Any Way To: Within The Ui, Select And Compare The Scalars For More Than 10 Experiments? I'D Like To Do Something Like:

Is there any way to: within the UI, select and compare the scalars for more than 10 experiments? I'd like to do something like:
select these 10 run in such and such a way and color them green Select these other 10 and color them blue

  
  
Posted 2 years ago
Votes Newest

Answers 18


Oh look, the blue setting is best!

  
  
Posted 2 years ago

SmallDeer34 , as a workaround for the issue of the graph maximize you can simply rename some of the experiments and this should fix the coloring issue for now until it is fixed in the next version 🙂

  
  
Posted 2 years ago

SmallDeer34 , great, thanks for the info 🙂

  
  
Posted 2 years ago

They don't, no

  
  
Posted 2 years ago

generally I include the random seed in the name

  
  
Posted 2 years ago

As an alternate solution, if I could group runs and get stats across the group, that would be cool

  
  
Posted 2 years ago

image

  
  
Posted 2 years ago

maximized:

  
  
Posted 2 years ago

And the reason is, because I have a bunch of "runs" with the same settings, and I want to compare broadly across several settings. So if I select "a bunch" with setting A I can see a general pattern when compared with setting B.

I've got 7-10 runs per setting, and about 7 or 8 settings

  
  
Posted 2 years ago

Or at least not conveniently

  
  
Posted 2 years ago

SmallDeer34 Hi! 🙂

I'm afraid that currently we support only up to 10 experiments in comparison at once. However you can add/remove experiments mid comparison. Is there a specific reason why you'd need to compare specifically 20?

Regarding the coloring, if you do 5/5 it would be supported.

  
  
Posted 2 years ago

SmallDeer34 , just out of curiosity, how many in a single compare would be needed for your use case?

  
  
Posted 2 years ago

something like this is what I'm looking for

  
  
Posted 2 years ago

no maximized:

  
  
Posted 2 years ago

Aggregating the sort of range of all the runs, maybe like a hurricane track?

  
  
Posted 2 years ago

SmallDeer34 , and they still have the same colors when you maximize the graph?

  
  
Posted 2 years ago

Well they do all have different names

  
  
Posted 2 years ago

No, not specifically 20,in fact more than 20

  
  
Posted 2 years ago