View a markdown version of this page

使用经过训练的模型生成新的模型构件 - Amazon Neptune

本文属于机器翻译版本。若本译文内容与英语原文存在差异,则一律以英文原文为准。

使用经过训练的模型生成新的模型构件

使用 Neptune ML 模型转换命令,您可以使用预训练的模型参数计算模型构件,例如已处理的图形数据上的节点嵌入。

用于增量推理的模型转换

增量模型推理工作流程中,处理完从 Neptune 导出的更新图形数据后,可以使用如下命令启动模型转换作业:

AWS CLI
aws neptunedata start-ml-model-transform-job \ --endpoint-url https://your-neptune-endpoint:port \ --id "(a unique model-transform job ID)" \ --data-processing-job-id "(the data-processing job-id of a completed job)" \ --ml-model-training-job-id "(the ML model training job-id)" \ --model-transform-output-s3-location "s3://(your S3 bucket)/neptune-model-transform/"

有关更多信息,请参阅《 AWS CLI 命令参考》中的 start-ml-model-transform-job

SDK
import boto3 from botocore.config import Config client = boto3.client( 'neptunedata', endpoint_url='https://your-neptune-endpoint:port', config=Config(read_timeout=None, retries={'total_max_attempts': 1}) ) response = client.start_ml_model_transform_job( id='(a unique model-transform job ID)', dataProcessingJobId='(the data-processing job-id of a completed job)', mlModelTrainingJobId='(the ML model training job-id)', modelTransformOutputS3Location='s3://(your S3 bucket)/neptune-model-transform/' ) print(response)
awscurl
awscurl https://your-neptune-endpoint:port/ml/modeltransform \ --region us-east-1 \ --service neptune-db \ -X POST \ -H 'Content-Type: application/json' \ -d '{ "id" : "(a unique model-transform job ID)", "dataProcessingJobId" : "(the data-processing job-id of a completed job)", "mlModelTrainingJobId": "(the ML model training job-id)", "modelTransformOutputS3Location" : "s3://(your S3 bucket)/neptune-model-transform/" }'
注意

此示例假设您的 AWS 证书是在您的环境中配置的。us-east-1替换为 Neptune 集群的区域。

curl
curl \ -X POST https://your-neptune-endpoint:port/ml/modeltransform \ -H 'Content-Type: application/json' \ -d '{ "id" : "(a unique model-transform job ID)", "dataProcessingJobId" : "(the data-processing job-id of a completed job)", "mlModelTrainingJobId": "(the ML model training job-id)", "modelTransformOutputS3Location" : "s3://(your S3 bucket)/neptune-model-transform/" }'

然后,您可以将此任务的 ID 传递给 create-endpoints API 调用,以创建新的端点,或者使用此任务生成的新模型构件更新现有端点。这允许新的或更新的端点为更新后的图形数据提供模型预测。

适用于任何训练任务的模型转换

您还可以提供一个trainingJobName参数来为在 Neptune ML 模型训练期间启动的任何 SageMaker AI 训练作业生成模型工件。由于 Neptune ML 模型训练作业可能会启动许多 SageMaker AI 训练作业,因此您可以灵活地基于任何 SageMaker AI 训练作业创建推理端点。

例如:

AWS CLI
aws neptunedata start-ml-model-transform-job \ --endpoint-url https://your-neptune-endpoint:port \ --id "(a unique model-transform job ID)" \ --training-job-name "(name of a completed SageMaker training job)" \ --model-transform-output-s3-location "s3://(your S3 bucket)/neptune-model-transform/"

有关更多信息,请参阅《 AWS CLI 命令参考》中的 start-ml-model-transform-job

SDK
import boto3 from botocore.config import Config client = boto3.client( 'neptunedata', endpoint_url='https://your-neptune-endpoint:port', config=Config(read_timeout=None, retries={'total_max_attempts': 1}) ) response = client.start_ml_model_transform_job( id='(a unique model-transform job ID)', trainingJobName='(name of a completed SageMaker training job)', modelTransformOutputS3Location='s3://(your S3 bucket)/neptune-model-transform/' ) print(response)
awscurl
awscurl https://your-neptune-endpoint:port/ml/modeltransform \ --region us-east-1 \ --service neptune-db \ -X POST \ -H 'Content-Type: application/json' \ -d '{ "id" : "(a unique model-transform job ID)", "trainingJobName" : "(name of a completed SageMaker training job)", "modelTransformOutputS3Location" : "s3://(your S3 bucket)/neptune-model-transform/" }'
注意

此示例假设您的 AWS 证书是在您的环境中配置的。us-east-1替换为 Neptune 集群的区域。

curl
curl \ -X POST https://your-neptune-endpoint:port/ml/modeltransform \ -H 'Content-Type: application/json' \ -d '{ "id" : "(a unique model-transform job ID)", "trainingJobName" : "(name of a completed SageMaker training job)", "modelTransformOutputS3Location" : "s3://(your S3 bucket)/neptune-model-transform/" }'

如果原始训练任务是针对用户提供的自定义模型,则在调用模型转换时必须包含一个 customModelTransformParameters 对象。有关如何实现和使用自定义模型的信息,请参阅Neptune ML 中的自定义模型

注意

modeltransform命令始终在最适合该训练的 SageMaker AI 训练作业上运行模型变换。

有关模型转换任务的更多信息,请参阅modeltransform 命令