
Reputation
Badges 1
8 × Eureka!Hi @<1523701070390366208:profile|CostlyOstrich36> , I deployed using the latest allegroai/clearml
image. When I execute /usr/bin/env
in the container I get:CLEARML_SERVER_VERSION=1.16.2
CLEARML_SERVER_BUILD=502
Hi @<1523701070390366208:profile|CostlyOstrich36> I have the same issue. I archived then deleted an experiment and the debug samples do not get deleted from the self-hosted fileserver. also no logs that I can find indicate an error.
I redeployed the latest versions and the button appears now
CLEARML_SERVER_VERSION=1.17.0
CLEARML_SERVER_BUILD=542
Hi @<1523701070390366208:profile|CostlyOstrich36>
For example if I want to push a task to a remote queue and have the task's name automatically be inferred from one of the configuration files. My use case is in cross validation training, the task name should include the model_name and the split_id
When you say clearml-serving you mean the CLI? Or the server also?
It was probably just a network issue on my end. Works with a larger delay than I expected.
When using clearml-agent daemon --stop
will it interrupt the running tasks or wait for the running task to finish and then stop the daemon?
Hi @<1523701070390366208:profile|CostlyOstrich36> how do I utilize the ‘presentation’:‘markdown’ solution you mentioned in conjunction with logger.report_table? Is it possible?
Hi @<1523701070390366208:profile|CostlyOstrich36>
I did something like this:
data = {"index": ["<a href=
>12345</a>"], y:[1],...}
df = pd.DataFrame(data)
clearml_logger.report_table(title, series, iteration, df)