[Python] Training call to sagemaker from local code hangs/freezes on Windows OS

#Using the PyTorch estimator tells SageMaker to use an AWS provided PyTorch container
estimator = PyTorch(
entry_point = "training_application.py", # training script
framework_version = "1.12", #PyTorch Framework version, keep it same as used in default example
py_version = "py38", # Compatible Python version to use
instance_count = 1, #number of EC2 instances needed for training
instance_type = "ml.p3.2xlarge", #Type of EC2 instance/s needed for training
disable_profiler = True, #Disable profiler, as not needed
role = role #Execution role used by training job
)

#Start the training
estimator.fit()


When I run this program or any other similar ones like huggingface_estimator the call to sagemaker just hangs and has to be interrupted at keyboard on my windows machine. AWS configure is all good. The job is never visible or reaches to be in sagemaker-->training --> list on aws console in my region. It is same even if i run by notebook or even accelerate package of aws. Is it even compatible with Windows.

DEBUG:botocore.loaders:Loading JSON file: C:\Users\VSCodeProjects\myenv\Lib\site-packages\botocore\data\endpoints.json
DEBUG:botocore.regions:Creating a regex based endpoint for ecr, us-east-1
DEBUG:sagemaker.estimator:Train args after processing defaults: {'input_config': None, 'role': 'arn:aws:iam::xxxx5411284:role/service-role/AmazonSageMaker-ExecutionRole-202409xxxxxxx', 'output_config': {'S3OutputPath': 's3://sagemaker-us-east-1-xxxxx5411284/'}, 'resource_config': {'VolumeSizeInGB': 30, 'InstanceCount': 1, 'InstanceType': 'ml.p3.2xlarge'}, 'stop_condition': {'MaxRuntimeInSeconds': 86400}, 'vpc_config': None, 'input_mode': 'File', 'job_name': 'pytorch-training-2024-09-17-01-30-47-493', 'hyperparameters': {'sagemaker_submit_directory': '"s3://sagemaker-us-east-1-xxxxx5411284/pytorch-training-2024-09-17-01-30-47-493/source/sourcedir.tar.gz"', 'sagemaker_program': '"training_application.py"', 'sagemaker_container_log_level': '20', 'sagemaker_job_name': '"pytorch-training-2024-09-17-01-30-47-493"', 'sagemaker_region': '"us-east-1"'}, 'tags': None, 'metric_definitions': None, 'experiment_config': None, 'environment': None, 'enable_network_isolation': False, 'retry_strategy': None, 'image_uri': '763104351884.dkr.ecr.us-east-1.amazonaws.com/pytorch-training:1.12-gpu-py38', 'debugger_hook_config': {'S3OutputPath': 's3://sagemaker-us-east-1-xxxxx5411284/', 'CollectionConfigurations': []}, 'enable_sagemaker_metrics': True, 'profiler_config': {'DisableProfiler': True}}

Continue reading...