当我在AWS CloudWatch上发现关于我的模型从数据中推断的错误时，我正在从我的角度前端调用SageMaker端点(以逗号分隔的字符串形式，目标值位于第一个索引)，我正在发送：无法解析数值。我用来调用端点的字符串是："1533071820,0.05619,0.056190,0.051611,0.056111,0.006076\n">
字符串请求="1533071820,0.05619,0.056190,0.056111,0.006076\n">
ByteBufferbuf=ByteBuffer.wrapp(request.getBytes())；invokeEndpointRequest.setBody(buf)；Use the SageMaker APIAmazonSageMakerRuntimeamazonSageMaker=亚马逊SageMaker RuntimeClientBuilder.defaultClient()；Invoke the model endpoint on SageMakerInvokeEndpointResult InvokeEndpointResult=amazonSageMaker invokeEndpoint(invokeEndpointRequest)；
我期望从端点得到的结果是一个JSON对象，其"score"属性的格式为：｛"predictions"：〔｛"score"：xxxxxxx｝〕｝
我从IDE日志中得到一个"ModelError:无法评估负载"one_answers"无法在CloudWatch上解析数值">
因此，在调试这个问题时，我发现由于末尾的换行符，模型假设我的输入中有两个字符串，但只有一个。因此，出现了错误。所以我只是把字符串改为："1533071820,0.05619,0.056111,0.051611,0.006076"。如果我正在分析多个有效载荷(批)，它是："153307 1820,0.051619,0.056190,0.0561110,0.056110,0.006076\n153307820,0.056119,0.056619,0.05519,0.051610,0.05611,0006076"，其中我的两个输入由换行符分隔，更重要的是，在的末尾没有换行符