What I put in the clearml.conf is the following:

agent.package_manager.pip_version = "==20.2.3" agent.package_manager.extra_index_url: [" "] agent.python_binary = python3.8

it also happens without hitting F5 after some time (~hours)

The number of documents in the old and the new env are the same though 🤔 I really don’t understand where this extra space used comes from

Ho and also use the colors of the series. That would be a killer feature. Then I simply need to match the color of the series to the name to check the tags

1.1.2

So I installed docker, added user to group allowed to run docker (not to have to run with sudo, otherwise it fails), then ran these two commands and it worked

Awesome! Thanks! 🙏

Thanks!3. I don't know, I never used Highcharts 🙂

this is the last line, same a before

There is a pinned github thread on https://github.com/allegroai/clearml/issues/81 , seems to be the right place?

hoo thats cool! I could place torch==1.3.1 there

is there a command / file for that?

Yes I agree, but I get a strange error when using dataloaders:
RuntimeError: [enforce fail at context_gpu.cu:323] error == cudaSuccess. 3 vs 0. Error at: /pytorch/caffe2/core/context_gpu.cu:323: initialization error
only when I use num_workers > 0

So it is there already, but commented out, any reason why?

Is it safe to turn off replication while a reindex operation is happening? the reindexing is rather slow and I am wondering if turning of replication will speed up the process

Update: I successfully isolated one of the reason, mem leak in matplotib itself, I opened an issue on their repo here

Hi SuccessfulKoala55 , will I be able to update all references to the old s3 bucket using this command?

the deep learning AMI from nvidia (Ubuntu 18.04)

Can I simply set agent.python_binary = path/to/conda/python3.6 ?

Maybe the agent could be adapted to have a max_batch_size parameter?

agent.cuda_version = 0 agent.cudnn_version = 0

On clearml or clearml-server?

Nevertheless there might still be some value in that, because it would allow to reduce the starting time by removing the initial setup of the agent + downloading of the data to the instance - but not as much as I described initially, if instances stopped are bound to the same capacity limitations as new instances launched

Hi, yes, you can use trains_agent.backend_api.session.client.APIClient.queues.get_all()

Isn't it overkill to run a whole ubuntu 18.04 just to run a dead simple controller task?

Sorry both of you, my problem was actually lying somewhere else (both buckets are in the same region) - thanks for you time!

I will let the team answer you on that one 🙂

It worked like a charm 😱 Awesome thanks AgitatedDove14 !

Here is the log from the instance

TimelyPenguin76 That sounds amazing! will there be a fallback mechanism as well? often p3.2xlarge are on shortage, would be nice to define one resources req as first choice (eg. p3.2xlarge) -> if not available -> use another resources req (eg. g4dn)