Reputation
Badges 1
533 × Eureka!AgitatedDove14 worked like a charm, thanks a lot!
bottom line I want to edit the cleanup service code to only delete tasks under a specific project - how do I do that?
So if I'm collecting from the middle ones, shouldn't the callback be attached to them?
AgitatedDove14 ⬆ please help 🙏
Loading part from task B:
` def get_models_from_task(task: clearml.Task, model_artifact_substring: str = 'iter_') -> dict:
"""
Extract all models saved as artifacts with the specified substring
:param task: Task to fetch from
:param model_artifact_substring: Substring for recognizing models among artifacts
:return: Mapping between iter number and model instance
"""
# Extract models from task (models are named iter-XXX where XXX is the iteration number)
model_...
later today or tomorrow, I'll update
AgitatedDove14 permanent. I want to start with a CLI interface that allows me add users to the trains server
Config == conf_obj
no?
whatttt? I looked at config_obj
didn't find any set
method
Manual model registration?
the link to manual model registry doesn't work
Very nice thanks, I'm going to try the SA server + agents setup this week, let's see how it goes ✌
Do i need to copy this aws scaler task to any project I want to have auto scaling on? what does it mean to enqueue hte aws scaler?
essentially editing apiserver.conf
section auth.fixed_users.users
Martin: In your trains.conf, change the valuefiles_server: '
s3://ip :port/bucket'
Isn't this a client configuration ( trains-init
)? Shouldn't be any change to the server configuration ( /opt/trains/config...
)?
I know I can configure the file server on trains-init
- but that only touches the client side, what about the container on the trains server?
AgitatedDove14 clearml version on the Cleanup Service is 0.17.0
I'm trying it now
Sorry I meant this link
https://azuremarketplace.microsoft.com/en-us/marketplace/apps/apps-4-rent.clearml-on-centos8
google store package could be the cause, because indeed we have the env var set, but we don't use the google storage package
Makes sense
So I assume, trains assumes I have nvidia-docker installed on the agent machine?
Moreover, since I'm going to use Task.execute_remotely
(and not through the UI) is there any code way to specify the docker image to be used?
it seems that only the packages that are on the script are getting installed
Mmm maybe, lets see if I get this straight
A static artifact is a one-upload object, a dynamic artifact is an object I can change during the experiment -> this results at the end of an experiment in an object to be saved under a given name regardless if it was dynamic or not?
So dynamic or static are basically the same thing, just in dynamic, I can edit the artifact while running the expriment?
Second, why would it be overwritten if I run a different run of the same experiment? As I saw, each object is stored under a directory with the task ID which is unique per run, so I assume I won't be overriding artifacts which are saved under the same name in different runs (regardless of static or dynamic)