为什么我在尝试通过 Lambda 调用 SageMaker 后收到错误？

Question

我尝试通过 Lambda 调用 SageMaker，但收到错误。

这是错误：

Response
{
  "errorMessage": "An error occurred (ModelError) when calling the InvokeEndpoint operation: Received client error (400) from primary and could not load the entire response body. See https://us-east-1.console.aws.amazon.com/cloudwatch/home?region=us-east-1#logEventViewer:group=/aws/sagemaker/Endpoints/jira-endpoint in account xxxx for more information.",
  "errorType": "ModelError",
  "stackTrace": [
    "  File \"/var/task/lambda_function.py\", line 19, in lambda_handler\n    response = runtime.invoke_endpoint(EndpointName=ENDPOINT_NAME,\n",
    "  File \"/var/runtime/botocore/client.py\", line 565, in _api_call\n    return self._make_api_call(operation_name, kwargs)\n",
    "  File \"/var/runtime/botocore/client.py\", line 1021, in _make_api_call\n    raise error_class(parsed_response, operation_name)\n"
  ]
}

模型端点来自预构建的模型-uri。

这是我的 Lambda 代码：

import os, io, boto3, json, csv, base64


ENDPOINT_NAME = os.environ['ENDPOINT_NAME']
runtime= boto3.client('runtime.sagemaker')

def lambda_handler(event, context):
    project = event['project']
    # project = base64.b64decode(event['project'])
    # timeline = base64.b64decode(event['timeline'])
    
    # payload = """{}.
    #     Given the project description above, sugges a project tasklist to achieve the project goals.
    #     The project timeline is {} weeks.
    # """.format(project, timeline)

        
    payload = project
    response = runtime.invoke_endpoint(EndpointName=ENDPOINT_NAME,
                                       Body=project, ContentType="text/csv")
                                     
                                     

    result = json.loads(response["Body"].read().decode())
    print(result)
    return {
        "statusCode": 200,
        "headers": { "content-type": "application/json"},
        "body":  result
    }

此外，我不确定 SageMaker 端点期望以什么格式接收 JSON 数据。

谢谢您的帮助！

Answer 1

为了调查导致“ModelError”的原因，您需要检查模型容器中的日志。来自端点的日志发送到“/aws/sagemaker/Endpoints/[EndpointName]”。请检查您是否在那里看到更具体的错误消息。

有关 SageMaker 在何处发出日志的更多信息，请参阅此文档页面： https://docs.aws.amazon.com/sagemaker/latest/dg/logging-cloudwatch.html