example of deploying a model to SageMaker using the Boto3 SDK for Python

boto.py

#Prepare and upload your model artifacts to an Amazon S3 bucket. The bucket must be in the same region as the SageMaker endpoint you will create.

#Create an IAM role with the necessary permissions for SageMaker to access the S3 bucket and other AWS services on your behalf.

#Use the AWS SDK for Python (Boto3) to create a SageMaker model resource. The model resource describes the location of the model artifacts and the Docker image containing the inference code.

#Create an endpoint configuration, which defines the resources that SageMaker should use when hosting the endpoint. This includes the number and type of instances to use, and the configurations for the model and the data processing.

#Use the endpoint configuration to create an endpoint. This deploys the model and makes it available for inference.

#Use the SDK to send requests to the endpoint and receive inferences.

import boto3

# Specify your AWS Region
aws_region = '<aws_region>'

# Create a low-level SageMaker service client
sagemaker = boto3.client('sagemaker', region_name=aws_region)

# Create a SageMaker model
model_name = '<model_name>'
model_s3_path = 's3://<bucket_name>/<path_to_model_artifacts>'
image = '<docker_image>'
sagemaker.create_model(ModelName=model_name,
                      Containers=[{'Image': image,
                                   'ModelDataUrl': model_s3_path}])

# Create an endpoint configuration
config_name = '<config_name>'
instance_type = '<instance_type>'
sagemaker.create_endpoint_config(EndpointConfigName=config_name,
                                 ProductionVariants=[{'InstanceType': instance_type,
                                                      'ModelName': model_name}])

# Create an endpoint
endpoint_name = '<endpoint_name>'
sagemaker.create_endpoint(EndpointName=endpoint_name,
                         EndpointConfigName=config_name)