This repository is for active development of the Azure SDK for Python. For consumers of the SDK we recommend visiting our public developer docs at https://learn.microsoft.com/python/azure/ or our versioned developer docs at https://azure.github.io/azure-sdk-for-python.
Hello !
I'll use the bug template but it's maybe more a lack of documentation / information.
azure-ai-ml:
1.5.0:
Ubuntu 20.04.4 LTS:
Python 3.8:
Describe the bug
I'm working with azureml-sdk and i try to switch to azure-ai-ml.
In the past, i was naming my tensorboard logdir programatically in order to have an information about the date and to do the link between my azureml job and my output dir.
I was doing this way (from inside a job) :
class AMLTensorboardLogger(IDistributed):
def __init__(self,rootLogDir='tb_results',LogDir_suffix='',logbatch:bool=False,log_freq:int=1):
self.logname = datetime.now().strftime("%y%m%d_%H%M")+f"_{LogDir_suffix}"
logdir=os.path.join(rootLogDir,self.logname)
os.makedirs(logdir,exist_ok=True)
self.logdir=logdir
self.logbatch = logbatch
self.logwriter=SummaryWriter(logdir)
###################################################################################
### We add this part in order to be able to run in AzureML Dashboard natively
###################################################################################
if os.getenv('AZUREML_LOGGING'):
print('Logging enable in AzureML Pipelines...')
from azureml.core import Run
current_run=Run.get_context()
current_run:Run
if 'Offline' not in current_run._run_id:
self._azrun = current_run
print(f'Run ID : {self._azrun._run_id}')
current_run.display_name = f"AZRun - {self.logname}"
print(f'Run Name : {current_run.display_name}')
To Reproduce
NA
Expected behavior
I would like a way to get the current run with the new SDK.
Screenshots
In the bottom, you can see a job that have been renamed from inside it.
I'd like to do this with sdk v2.0 but i don't know how.
Hello ! I'll use the bug template but it's maybe more a lack of documentation / information.
Describe the bug I'm working with azureml-sdk and i try to switch to azure-ai-ml.
In the past, i was naming my tensorboard logdir programatically in order to have an information about the date and to do the link between my azureml job and my output dir.
I was doing this way (from inside a job) :
To Reproduce NA
Expected behavior I would like a way to get the current run with the new SDK.
Screenshots In the bottom, you can see a job that have been renamed from inside it. I'd like to do this with sdk v2.0 but i don't know how.