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
Hi! When Logging Datasets To A Task (By Using

Hi! When logging Datasets to a Task (by using Dataset.get() ), they appear under the Hyperparameter -> Datasets tab. This is all well.

But if I want to change my input dataset for a cloned Task, currently I'm doing that through an input string name or dataset ID. But this is a crude approach. It would be better if there was a possibility to click on Edit and get a list of available datasets , similar to how it works for Artifacts -> Input Models.

Does this feature exist today?

  
  
Posted 21 days ago
Votes Newest

Answers 4


No problem, thanks for replying!

  
  
Posted 13 days ago

Hi @<1864479785686667264:profile|GrittyAnt2> ! For OS datasets, this is currently not supported unfortunately

  
  
Posted 20 days ago

My apologies, Mikael - I didn't realize this wasn't supported on the OS version. :face_palm:

  
  
Posted 20 days ago

@<1864479785686667264:profile|GrittyAnt2> - Yes, it does. From within your cloned task you can go to the Dataviews tab and click the Edit button. From there you can:

  • Click the pencil icon to select a different version of the current dataset;
  • Click the trashcan icon to remove the current dataset; or
  • Click the plus icon, which will open a dialog where you can select a different dataset.
    Hope this helps!
    image
    image
  
  
Posted 20 days ago
118 Views
4 Answers
21 days ago
12 days ago
Tags