1. 首先确认您已经正确地配置了IAM角色和权限。您可以使用以下代码片段来检查您的角色是否具有必要的权限：

import boto3

sm = boto3.client('sagemaker')
role_arn = 'arn:aws:iam:::role/'

try:
    sm.list_notebook_instances()
except sm.exceptions.ClientError as e:
    if "AccessDenied" in str(e) and "is not authorized to perform" in str(e):
        print("Role does not have required permissions.\n")
        print("Missing %s permission on the role: %s" % (e.response['Error']['Code'], role_arn))
    else:
        print(e)
else:
    print("Role has required permissions.")

2. 如果您确认角色和权限配置正确，则可能是因为您的AWS账户没有在Sagemaker中启用假定角色。您可以使用以下代码片段来启用假定角色：

import boto3

sagemaker = boto3.client('sagemaker')
sagemaker_response = sagemaker.create_notebook_instance(
    NotebookInstanceName='my-notebook-instance-####',
    # ... Other arguments ...
    ExecutionRoleArn='arn:aws:iam:::role/my-execution-role',
    # ... Other arguments ...
)

response_status = sagemaker.describe_notebook_instance(NotebookInstanceName=sagemaker_response['NotebookInstanceName'])
print("Notebook Instance status: ", response_status['NotebookInstanceStatus'])

在上面的代码中，注意提供ExecutionRoleArn参数以指定假定角色的ARN。

如果这些解决方法均不起作用，则建议您检查Sagemaker服务的更新，并联系AWS支持团队寻求帮助。