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
Hey, I'D Like To Store My Trained Models, Results Of Transformers Training, Into Local Disk. I Tried To Set Up

Hey,
I'd like to store my trained models, results of transformers training, into local disk. I tried to set up default_output_uri: "file:///home/karol/data/" in ~/clearml.conf; however, after training the specified folder is empty. (Unfortunately default upload to fileserver constantly throws errors due to files size.) I have no idea where all files are stored. ClearML didn't raise any error and doesn't present any info in "Artifacts" tab. Do you have any ideas?

  
  
Posted 6 months ago
Votes Newest

Answers 6


What if you set the default_output_uri to false ?

  
  
Posted 6 months ago

The problem is "Artifacts" tab is empty, it wasn't with default logging (to the fileserver I guess) but then the files failed to upload.

  
  
Posted 6 months ago

Hi @<1570220844972511232:profile|ObnoxiousBluewhale25> , you can click on the model in the artifacts tab and that should take you to the model repository. What is logged in the url of the model?

  
  
Posted 6 months ago

Also, how are you saving your models?

  
  
Posted 6 months ago

I'm using a special Callback provided by Transformers: None

  
  
Posted 6 months ago

@<1570220844972511232:profile|ObnoxiousBluewhale25> it creates a new Model here
None
If you want it to log to something other than the default file server create the clearml Task before starting the training:

task = Task.init(..., outout_uri="file:///home/karol/data/")
# now training

It will use the existing Task and upload to the destination folder

  
  
Posted 6 months ago
382 Views
6 Answers
6 months ago
6 months ago
Tags