So, what I am referring to is the ability of a system to allow some rigor and robustness of tracking of experiments, and also enforcing some thoughts on how things might be deployed, early on in the development process, whilst not being overly prescriptive and cumbersome
I'm cannot agree more!!
VivaciousPenguin66 We are working on trying to better understand how to solve this very delicate act of balance and offer some sort of "JIRA" for ML.
If this is okay with you, once product pe...
BattyLion34
Maybe something inside the task is different?!
Could you run these lines and send me the result:from clearml import Task print(Task.get_task(task_id='failing task id').export_task()) print(Task.get_task(task_id='working task id').export_task())
SweetGiraffe8
That might be it, could you test with the Demo server ?
GreasyPenguin14 yes there is 🙂
https://github.com/allegroai/clearml/issues/209
Set environment variable CLEARML_NO_DEFAULT_SERVER=1
MinuteGiraffe30 if you are running the following command while your current directory is where you code is, what are you getting?
$ git ls-remote --get-url origin
And you have the exact same folder structure / content, and server A/B give a different set of experiments ?
(is serverB empty, meaning no experiments at all?)
But I think this error has only appeared since I upgraded to version 1.1.4rc0
Hmm let me check something
Having the ability to pack jobs/tasks onto the same "resource" (underlying server/EC2 instance)
This is essentially a "queue". Basically a queue is a way to abstract a specific type of resource, so that you can achieve exactly what you descibed.
open up a streaming use case, wherein batch (offline) inference could be done directly inside of a ClearML pipeline in reaction to an event/trigger (like new data landing in your data lake).
Yes, that's exactly how clearml is designed, a...
If the manual execution (i.e. pycharm) was working it should have stored it on the Pipeline Task.
Any recommendation or working combinations of AMI
I would take the deeplearning AMIs from Nvidia AWS , I think they work on both CPU and GPU machines.
In terms of dockers, python dockers for CPU and nvidia runtime for GPU
[https://hub.docker.com/layers/library/python/3.11.2-bullseye/images/sha256-6128ea86d[…]d2c01646d599352f6ddd9893420eb815a06c3b90619f8?context=explore](https://hub.docker.com/layers/library/python/3.11.2-bullseye/images/sha256-6128ea86db7f6b1b286d2c01646d599352f6ddd98...
preinstalled in the environment (e.g. nvidia docker). These packages may not be available via pip, so the run will fail.
Okay that's the part that I'm missing, how come in the first run the package existed and in the cloned Task they are missing? I'm assuming agents are configured basically the same (i.e. docker mode with the same network access). What did I miss here ?
@<1595587997728772096:profile|MuddyRobin9> are you sure it was able to spin the EC2 instance ? which clearml version autoscaler are you running ?
This is very odd, can you also put here the file names? maybe an odd character is causing it?
Can you also test it with the latest clearml version (1.8.0) ?
FierceRabbit20 it seems the Pipeline Task that was created is missing the "installed requirements" section. How are you creating the actual pipeline Task? is this from code?
EnviousPanda91 notice that when passing these arguments to clearml-agent you are actually passing default args, if you want an additional argument to Always be used, set the extra_docker_arguments
here:
https://github.com/allegroai/clearml-agent/blob/9eee213683252cd0bd19aae3f9b2c65939d75ac3/docs/clearml.conf#L170
Hi UpsetTurkey67
"General/my_parameter_name" so that only this part of the configuration will be updated?
I'm assuming this is a Hyperparameter not a configuration object (i.e. task.connect not task.connect_configuration), if this is the case then Yes 🙂
What happened in the server configuration that all of a sudden you have zero ports open?
set the following:CLEARML_AGENT_DISABLE_SSH_MOUNT=1 clearml-agent daemon ...
The issue is, it will automatically mount the .ssh of the host into the container, so that if you are using SSH to clone git you have credentials, in your case, it also mounts the configuration, hence failing to login.
I will make sure we add it to the configuration file, so it is more visible
I want each remote task to execute one instance of the hydra multirun, but I suspect the remote will try to run the full multirun by itself
if config.clearml.remote and task.running_locally(): task.execute_remotely( queue_name=config.clearml.queue_name, clone=True, exit_process=False ) return
I think this ensures the local execution actually triggers the remote one, so it should be as you expect, no?
RoughTiger69
move the files locally (i.e. based on the example move folder b
into folder a
) Create a new version with two parents ('a' and 'b') then sync the local root folder ('a' in your case). Only the meta-data should change (because the referenced files are already in one of the datasets)wdyt?
Now I'm curious what's the workaround ?
regrading the actual artifact access, this is the usual Task.artifacts access: see example here:
https://github.com/allegroai/clearml/blob/master/examples/reporting/artifacts_retrieval.py
I think that what you need is to create an OutputModel , then call update weights file when you have the better model, this will also allow you to tag the model object. Would that help? Or would it make sense to use Task.models and count on the auto logging?
Okay I found it, this is due to the fact the newer versions are sending the events/images in a subprocess (it used to be a thread).
The creation of the object is done on he main process, updating file index (round robin manner), but the check itself, happens on the subprocess., which is not "aware" of the used indexes (i.e. it is always 0, hence when exceeding the history side, it skips it)
SmallBluewhale13 the final path is automatically generated, you only need to specify the bucket itself. By default it will be your "files_server"
https://github.com/allegroai/clearml/blob/c58e8a4c6a1294f8acec6ed9cba81c3b91aa2abd/docs/clearml.conf#L10
You can either change the configuration (which will make sure All uploaded artificats will always be there, including debug images etc.)
You can specify where you want the artifacts and debug images to be uploaded by setting:
https://allegro....
It does not upload, the default behavior is to log the artifact (so you know where you stored, but not enforce unnecessary uploads)
If you were to change:task = Task.init(project_name='examples', task_name='Keras with TensorBoard example')
to:task = Task.init(project_name='examples', task_name='Keras with TensorBoard example', output_uri="
")
It would also upload the model
Yes docker was not installed in the machine
Okay make sense, we should definitely check that you have docker before starting the daemon 😉
Ok, it would be nice to have a --user-folder-mounted that do the linking automatically
It might be misleading if you are running on k8s cluster, where one cannot just -v mount
volume...
What do you think?