CostlyOstrich36 , I am not sure what you mean, but if you refer to the name of the category in the configuration, it doesn't matter.. the names are arbitrary for this example..

I have some custom configuration that describes the experiment I am running..
Some dictionary.. nothing to do with frameworks

the above represent my case

CostlyOstrich36 , thanks!

is there a chance you help me with the specific POST call for debug? I was trying to implement it using requests package but I got errors.. didn't work for me.. I believe it something trivial

nice, saved some electrons from running 🙂

yes

https://demoapp.demo.clear.ml/projects/3de9c825c8c94a6eb1578a0e00cc4ec4/experiments/4bc7f2a64ae440cf85c5972d3d169dcd/output/hyper-params/hyper-param/General

I am getting 405

if I remove the import and use the path I get this:
raise MissingConfigException(hydra.errors.MissingConfigException: Cannot find primary config 'config'. Check that it's in your config search path.

mostly the second case

OS:Fedora
browser: brave but also on chrome
reproduce like in the example I gave above.. drag the right corner across more then a single column

it happens on task init

if I can't "pull", execute, report tasks from the same persistent python script it doesn't solve the problem of avoiding rerunning some heavy setup for a lightweight trial

AgitatedDove14 Thanks, I am aware of the auto_connect_frameworks and it makes sense for torch.save (and probably other stuff) however if it is called about 10 times during model initialization it seem excessive and I would be happy to avoid this behavior during model initialization if possible. do you know what is actually triggering the uploads?

😕 well I have py38 and debian.. strange I'll give it another look

(value, "")->(value, "\")

AgitatedDove14 , this was my starting point, however this is only a list of workers ids and names for example:
{'id': 'ip-777-77-7-77.us-east-7.compute.internal:0', 'name': <function Entity.__getattr__.<locals>.new_func at 0x782477c68ca0>}

AgitatedDove14 , the hydra people say its not on "their side".. so it is still a mystery to me

AgitatedDove14 ,yes my own server
is there an available reference to such a post request? I was trying some variations and it didn't work for me

this is what I get with curl

response:
{"meta":{"id":"33e9e80e94ee4384b45962acafcd2af5","trx":"33e9e80e94ee4384b45962acafcd2af5","endpoint":{"name":"","requested_version":1.0,"actual_version":null},"result_code":400,"result_subcode":0,"result_msg":"Invalid request path /v2.14/debug/ping","error_stack":null,"error_data":{}},"data":{}}

How can we increase the number of API server processes?

very strange that you don't see the same in the community server

AgitatedDove14 thanks, at peak usage we have 6-8 gb of free RAM

thanks AgitatedDove14 , Not that I know I didn't do any special setup other than the straightforward pip install hydra-core --pre

AlertBlackbird30 I saw you asked about wanted features to add to the roadmap.. this my top one 🙂

https://github.com/allegroai/clearml/issues/488

I am familiar with the above..
I am talking about an overview.. I can count the workers manually but when I have a 100 workers its too much

Bonus 2: having a way to push experiments up in the queue