我正在尝试使用sagemaker automl实验中的模型进行批处理转换推断。我不确定使用SKLearnModel这种方法是否正确,但这是代码和错误:
xgb = SKLearnModel(model_data='model-16003',
image=docker_uri,
entry_point='sagemaker_serve.py',
py_version='py_3',
role=role)
xgb_transformer = xgb.transformer(instance_count,
instance_type,
strategy=strategy,
accept='text/csv',
output_path=output_path,
max_payload=max_payload,
max_concurrent_transforms=max_concurrent_transforms)
xgb_transformer.transform(input_path,
content_type='text/csv',
compression_type='Gzip',
split_type='Line')
error: No GPUs detected (normal if no gpus installed)
我要关闭还是这是错误的路径?任何建议表示赞赏!