I am not very familiar with KubeFlow but as far as I know it is mainly for orchestration whereas ClearML offers a full E2E solution 🙂
I would advise using ClearML 😄
I really like ClearML
clearml-agent is for orchestration - remote execution.
clearml is the python package you need to install and add the magic lines of code:
I want to build end2end mlops platform, so i just use only clear-agent natanM?
I don't really understand about remote execution
That means I can use ClearML Agent instead of Kubernetes, right?
but you can use it with or without K8s
It depends on what you use K8s for
like docker or K8s?
It's a way to execute tasks remotely and even automate the entire process of data pre processing -> training -> output model 🙂
You can read more here:
from data versioning, model serving, model experiment
I feel ClearML is an end-to-end MLOps platform
I really don't understand about orchestration
I see two versions: clearml and clearml-agent
And orchestration 🙂
can you explain simpler?
I'm really confused when choosing a framework to build our end2end MLOps Platform.
Can you give me some advices?
It means you can run your code on a different machine very very easily using ClearML 🙂
run on dockers - i.e. execute your code inside of a docker image
I am thinking that ClearML Agent like K8s?
What's that mean? can you give me an example
It can run dockers and it can run over K8s