PredictRequest specifies which TensorFlow model to run.
signature for inference
OK
Response for PredictRequest on successful run.
Output tensors.