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
Unanswered
Trying To Create A Data Pipeline On My Own. Wanted To Ask, For Each Batch Of Data, Do I Have To Create A New Dataset Object Or Do I Just Create One Dataset Object And Add Batches To It. If Its The Latter, Then How.


I'm kind of new to developing end to end applications so I'm also learning how the predefined pipelines work as well. I'll take a look at the clear ml custom pipelines

  
  
Posted 3 years ago
180 Views
0 Answers
3 years ago
one year ago