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
If I Ran A Hyperparemeter Sweep And I Wanted To Create A Graph Where The X-Axis Was One Of The Hyperparameters, Let'S Say The Momentum Term Of The Optimizer, And I Wanted To Plot That Vs The Min-Loss Over All Epochs, Is There A Good Way To Do This With Cl

If I ran a hyperparemeter sweep and I wanted to create a graph where the x-axis was one of the hyperparameters, let's say the momentum term of the optimizer, and I wanted to plot that vs the min-loss over all epochs, is there a good way to do this with ClearML? So each x-y pair is from a different experiment. I'm not sure how I'd do this with compare.

  
  
Posted 10 months ago
Votes Newest

Answers 5


Maybe you want to use some other functions then the ones I quoted, so feel free to read the docs, you should be able to do this

  
  
Posted 10 months ago

@<1545216070686609408:profile|EnthusiasticCow4> yes, that's true. I would aggregate the tasks by tags (the steps will be tagged with opt: ID ), None then get the metrics to get the losses None , and look into the tasks config to get the term you wanted to optimize None , then plot everything

  
  
Posted 10 months ago

Hi Again Eugen,

If I use the hyperparameter tool in ClearML, won't that create a different experiment for every step of the hyperparameter-optimizer? So this will be run across experiments. I could do something with pipelines but since the metrics are already available in the ClearML hyperparameter/metric tables I thought it would make sense to be able to plot against those values.

  
  
Posted 10 months ago

I see, that could work.

  
  
Posted 10 months ago

Hi @<1545216070686609408:profile|EnthusiasticCow4> ! Can't you just get the values of the hyperparameters and the losses, then plot them with something like mathplotlib then just report the plot to ClearML?

  
  
Posted 10 months ago
568 Views
5 Answers
10 months ago
10 months ago
Tags