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
Security Question: In My Journey Of Running Clearml The "Hard Way" (Self-Hosted), One Problem I Haven'T Solved Is Security. Some Discussion Here...

When you run the docker-compose.yml on an EC2 instance, you can configure user login for the ClearML webserver. But the files API is still open to the world, right? (and same with the backend?)

We could solve this by placing the EC2 instance into a VPN.

One disadvantage to that approach is it becomes annoying to reach the model registry from outside the VPN, like if you have a deployment pipeline based in GitHub Actions. Or if you wanted to trigger a ClearML pipeline from a VPC that isn't peered with the VPN's VPC. Fixing those issues adds complexity.

Something neat about MLFlow is that the SDK supports using various types of authentication.

For example, you can have the MLFlow client SDK use JWT tokens or even AWS SigV4 auth. What that means is you could stick a load balancer or API Gateway in front of MLFlow, shielding it from the whole world. If you need to access MLFlow using the SDK in something like GitHub Actions, you just set some environment variables, whether that's a JWT token or set of AWS Access Keys.

Posted one year ago
0 Answers
one year ago
one year ago