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
I Have A Question. If I Do A Training On A Cloud Instance For Example The Models Are Saved There First. This Is The Path That Is Also Logged In The Clearml As Model Url. However, After The Training The Resulting Model Is Copied To A Central Fileserver And

I have a question. If I do a training on a cloud instance for example the models are saved there first. This is the path that is also logged in the clearml as model url. However, after the training the resulting model is copied to a central fileserver and the cloud instance is shutdown. How to make sure all models in the "models" overview in the clearml dashboard have a current url?

Posted 2 years ago
Votes Newest


GreasyPenguin14 Hi!

If I understand you correctly, you would have to change the url's of the models yourself so they would point to the now downed instances.

You can also use the following setting:
sdk.development.default_output_uri: "SOME_URL"
in your ~/clearml.conf to set it to send the models anywhere you want them to go from the get go 🙂

Is that helpful?

Posted 2 years ago