![Profile picture](https://clearml-web-assets.s3.amazonaws.com/scoold/avatars/TrickySheep9.png)
Reputation
Badges 1
383 × Eureka!Would adding support for some sort of post task script help? Is something already there?
Do we support GPUs in a) docker mode b) k8s glue?
Thanks AgitatedDove14 - i get overall what you are saying. Have to get glue setup, which I couldn’t understand fully, so that’s a different topic 🙂
The job itself doesn’t have any other param
how do you see things being used as the most normal way?
The Optimizer task is taking a lot of time to complete. Is it doing something here:
You mean the job with the exact same arguments ?
Yes
No, all of them completed!
Beyond this have the UI running, have to start playing with it. Any suggestions for agents with k8s?
Thanks for the fast responses as usual AgitatedDove14 🙂
AgitatedDove14 - does having this template work for updating hte base image:
` spec:
containers:
- image: nvidia/cuda:11.4.1-cudnn8-runtime-ubuntu20.04 `
I am seeing that it still picks up nvidia/cuda:10.1-cudnn7-runtime-ubuntu18.04
AgitatedDove14 aws autoscaler is not k8s native right? That's sort of the loose point I am coming at.
Having a pipeline controller and running actually seems to work as long as i have them as separate notebooks
PipelineController with 1 task. That 1 task passed but the pipeline says running
Basic question - i am running clearml agent in a ubuntu ec2 machine. Does it use docker by default? I thought it uses docker only if I add the --docker
flag?
I then install this wrapper SDK in my containers, notebook instances etc
What happens if I do blah/dataset_url
?
` if project_name is None and Task.current_task() is not None:
project_name = Task.current_task().get_project_name()
if project_name is None and not Task.running_locally():
task = Task.init()
project_name = task.get_project_name() `
Does a pipeline step behave differently?
Maybe two thing here:
If Task.init() is called in an already running task, don’t reset auto_connect_frameworks? (if i am understanding the behaviour right) Option to disable these in the clearml.conf
Any reason to not have those as two datasets?