I’m often deploying AWS endpoints with a script mode (adding my custom python to the model) and sometimes there might be stupid spelling errors in the Python. I see it immediately in the CloudWatch and I want to interrupt the deployment, but seems there’s no such an option.
Otherwise the deployment will re-try the deployment really for a long time.
Any thoughts on how is it possible to kill the deployment running?
Or at least make it fail after 2 min or so, not wait half-hour. I find such a behaviour quite inconvenient.
Thanks!
P.S. local model not always work good with custom code and locally trained models (seems due to inconsistencies with prebuilt containers) and it requires time to understand why – i prefer to deploy.

So, I expect either to be able to kill the deployment process or to set the fail timeout to 2 min.

xgb_inference_model = XGBoostModel(
        model_data=model_url,
        role=role,
        entry_point="inference1.py",
        source_dir="./code",
        framework_version="1.7-1",
        sagemaker_session=sagemaker_session
    )
predictor = xgb_inference_model.deploy(
        initial_instance_count=1,
        endpoint_name='opportunity-win-score',
        instance_type="ml.t2.medium"
    )
```   
`