Hi, Self-hosted using docker-compose.

Yes it is! But ClearML didn't support multi node training out of the box in a way that it streamline the process. So we are trying to figure out a way to do it.

Hi AgitatedDove14 , that's what i am trying to figure out as well. The task has nothing to do with torch, and the requirements.txt doesn't have any torch packages as well.

Sorry i forgot to paste the logs.

Ok thanks, looking forward to it. Would you advise on the bug you encountered?

ok. Any idea what can go on between the setting up of clearml-agent and initialising the clearml-agent itself? Does the clearml-agent try to communicate with any internet address. From another perspective, it looks like a long time out issue. I happen to be deploying on a disconnected on-premise setup.

So these (PIP_INDEX_URL) weren't used when clearml starts running pip.

Just to put a ping for those on this side of the timezone to look at. Thanks.

Thanks CostlyOstrich36 , how do i know how is the parts indexed in the first place? Or rather, how is chunk and parts defined? Say in the context of images, videos, text documents...etc.

It would make sense on a very large resource cluster. Unfortunately we only have less than 50 GPUs to share across. A multi-tenant SAAS would cut the resources into even more smaller clusters and not help with efficiency. Or would you have a suggestion?

Hi, i will have to get back to you again. Need to check every client's repo to determine your hypothesis.

Hi. The upgrade seems to go well but i'm seeing one wierd output. When i ran a task and observe the software installed under the execution tab , i still see clearml=0.17 . Is this expected?

Hi, it make sense to automate this part just like how you automate the rest of the MLOps flow, especially when you already support Data Versioning/Lineage, Data Provenance (How it works with the experiment and as a model source) should be in too. Although i agree technically it's probably not possible to tell if the users actually used the indicated datasets after they do a datasets.get_copy() .

thanks GrumpyPenguin23 , i'll look deeper on that. This kinda fits what i am looking for but its for TRAINS and there's no technical how-to.
https://clear.ml/blog/stop-using-kubernetes-for-ml-ops/

yes its on purpose, each user would have their own AWS credentials for default_output_uri.

Yes I am.

Hi SuccessfulKoala55 , thanks. Opened issue on the CLearml-Agent GH at https://github.com/allegroai/clearml-agent/issues/67

I've been reading the documentation for a while and I'm not getting the following very well.
Given an open source codes say, huggingface. I wanted to do some training and i wanted to track my experiments using ClearML. The obvious choice would be to use Explicit Reporting in ClearML. But the part on sending my training job. and let ClearML orchestrate is vague. Would appreciate if i can be guided to the right documentation on this.

I'm also beginning to think this is related to https://clearml.slack.com/archives/CTK20V944/p1620664770492400 . Previously when i set force_repo_requirements_txt=true and system_site_packages: true , it seems to work. upgrading to v1.02 seems to change things.

Hi CostlyOstrich36 , nothing in particular. I was doing a research and noticed that ML Pipelines was mentioned not even once in the literature. So i wonder if one should be done. I'm looking at other aspects as well but i'll gradually ask on those.

Hi, the latest k8sglue-example.py was last commited about 4 months ago. Are you refering to that version?

We are deploying ClearML Server via the docker-compose.

For ClearML-Agent. We have the choice of Docker or K8S preferred (Using the Glue).
For K8S, we can't get the glue to work ( https://clearml.slack.com/archives/CTK20V944/p1614525898114200?thread_ts=1613923591.002100&cid=CTK20V944 ) so we can't make an assessment of whether it actually works for us.

This is the log i extracted.

Ok that worked. So every time i have changes in codes, i will have to rerun the experiment on my own machine that doesn't have any GPUs?

Kinda defeat the purpose of using ClearML Agent.

yup. in this case it wasn't root. Removing that USER and -u in pip solves the problem. However, in our production images, we are required to remove root access.
` FROM nvidia/cuda:10.1-cudnn7-devel

ENV DEBIAN_FRONTEND noninteractive
RUN apt-get update && apt-get install -y
python3-opencv ca-certificates python3-dev git wget sudo ninja-build
RUN ln -sv /usr/bin/python3 /usr/bin/python

create a non-root user

ARG USER_ID=1000
RUN useradd -m --no-log-init --system --uid ${USER_ID} a...

docker exec clearml-elastic curl zsh: no matches found:

Thanks SuccessfulKoala55 , how might I do this clean up? Does this increase with more use of ClearML? And to add, we save all artifacts onto a remote S3 server.

No issues. I know its hard to track open threads with Slack. I wish there's a plugin for this too. 🙂

Yes of cos, its a long one.

Hi, scenario as follows.

client.py runs task.execute_remotely(queue='myqueue', exit_process=True) The API section of clearml.conf at client side is read in. client side calls clearml server and insert task into queue. K8S glue retrieves task from queue. Spawn a K8S pod. K8S pod performs git clone Error. ssh keys not found.
Each individual has their own key in the gitlab profile and gitlab is configured to only work via ssh.
We can't place the key in the image as this is as good as ...