Reputation
Badges 1
282 × Eureka!Thanks 👍 . Should i create an issue on Github?
[root@2c7498711bef elasticsearch]# curl
`
{
"index" : "events-training_stats_scalar-d1bd92a3b039400cbafc60a7a5b1e52b",
"shard" : 0,
"primary" : false,
"current_state" : "unassigned",
"unassigned_info" : {
"reason" : "CLUSTER_RECOVERED",
"at" : "2021-05-22T11:33:38.932Z",
"last_allocation_status" : "no_attempt"
},
"can_allocate" : "no",
"allocate_explanation" : "cannot allocate because allocation is not permitted to any of the nodes",
"node_allocation_decisi...
I also see this on my logs, noting that the config is read in but its still printing the supposedly hidden keys on the logs and UI.agent.hide_docker_command_env_vars.enabled = true agent.hide_docker_command_env_vars.extra_keys.0='TRAINS_AGENT_GIT_USER' ..... docker_cmd=harbor.ai/public/detectron2:v3 --env TRAINS_AGENT_GIT_USER=gituser
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?
Ok thanks. that explains alot. We have been doing this wrongly the whole time, thinking that the clearml.conf on the client side would be acknowledged by the remote agent execution. In reality, only the API section is utilised.
Any idea where i can find the relevant API calls for this?
Hi, i have the same question. Why would this be ignored if called remotely?
https://clear.ml/docs/latest/docs/references/sdk/task/#set_base_docker
Hi AgitatedDove14 , i've got the same error. It would appear that the code references clearml_agent/helper/base.py
which i believe is part of clearml-agent v0.17.1. Could that be the issue?
Hi, i changed it, but it still point to https://files.pythonhosted.org/packages .
Sorry, dev end I was referring to my developers.
I didn't think Horovod needs to be as complicated as you described. It can also work by running on multiple known nodes. How would i add a glue for multinode?
Horovod does also work with other similar products such as yours (E.g. Polyaxon).
Hi SuccessfulKoala55 , just wondering how i can follow up on this.
and out of curiosity, what did you think we were talking about? cos i didn't see anywhere else that might print the secrets.
yeah, someone should call them out.
Hi SuccessfulKoala55 , would they need the fileserver to route to minio then? E.g.
This will ensure that any actions by clearml-data and models are saved into the S3 object store.
api {
files_server: s3://ecs.ai:80/clearml-data/default
}
aws {
s3 {
credentials {
host: http://ecs.ai:80
## Insert the iam credentials provided by your SAs here.
}
}
}
But if user forgot to do above, they will be saved on ClearML server. If I switch off f...
Hi,
basically i run this block first and ended the script.task = Task.init(project_name="afro-nmt", task_name=args.taskname, continue_last_task=args.taskid) Logger.current_logger().report_scalar(title="BLEU",series="JW300",value=args.jwbleu, iteration=args.lastiter)
Then i run another script, with series different.
` task = Task.init(project_name="afro-nmt", task_name=args.taskname, continue_last_task=args.taskid)
Logger.current_logger().report_scalar(title="BLEU",series="SS900",value=arg...
Hi FriendlySquid61 , AgitatedDove14 , the issue and possible fix is in this issue raise. https://github.com/allegroai/clearml-agent/issues/51
Yeah that sounds good. But from user perspective, especially the untrained, they wouldn't know what to point to. Example, some may think it's an exe, some think it's a zip bundle, and others think it's any github repo with the word vscode.
For example, it would useful to integrate https://github.com/whylabs/whylogs#features into ClearML as part of data and model monitoring. WhyLogs would have their own static page that would preferably be displayed as a new custom tab (besides logs, scalars and plots.).
f you can directly access the machine running the agent, yes you could. If not reverse proxy is in the working
Hi AgitatedDove14 , i might have misunderstood your previous comment above. Do you mean that clearml-session can only work regardless of whether xforwarding is configured, if we have direct access to the Kubernetes worker when we run K8S glue?
We did some testing today and clearml-session tried to tunnel with a k8s cluster ip, and thus failed.
If we setup a ingress with Me...
Hi AgitatedDove14 , thanks.
In this case i am running k8s glue (machine glue), which will then spawn off pods in kubernetes worker (machine worker). So when you say direct access, are you refering to the Glue machine or K8S Worker machine?
They don't have the same version. I do seem to notice that if the client is using version 3.8, during remote execution will try to use that same version despite the docker image not installed with that version.
Unfortunately due to security, clients can't have direct access to the nodes. Is there any possible workarounds at the moment?
Thanks. That's easy to miss as its not quite apparent in the main docs. How should i pass in env variables with Task?
Thanks TimelyPenguin76 , is there an env var for the S3 connection as well?
Ok. Problem was resolved with latest version of clearml-agent and clearml.
Hi TimelyPenguin76 ,
If you notice in the last screenshot, it state the bucket name to be http://ecs.ai . It then it tries to open http://s3.amazonaws.com/ecs.ai/clearml-models/artifact/uploading_file?X-Amz-Algorithm= ....
Hi, i will have to get back to you again. Need to check every client's repo to determine your hypothesis.