CODEWITHSUNDEEP
tensorflowtensorflow-servingtfx
Max Power
Max Power

Reputation: 992

Generate instances or inputs for TensorFlow Serving REST API

I'm ready to try out my TensorFlow Serving REST API based on a saved model, and was wondering if there was an easy way to generate the JSON instances (row-based) or inputs (columnar) I need to send with my request.

I have several thousand features in my model and I would hate to manually type in a JSON. Is there a way I can use existing data to come up with serialized data I can throw at the predict API?

I'm using TFX for the entire pipeline (incl. tf.Transform), so I'm not sure if there is a neat way built into TFX I can use.

The output from saved_model_cli is this:

The given SavedModel SignatureDef contains the following input(s):
  inputs['examples'] tensor_info:
      dtype: DT_STRING
      shape: (-1)
      name: input_example_tensor:0

Which does not tell me much.

Upvotes: 4

Views: 2585

Answers (2)

RakTheGeek
RakTheGeek

Reputation: 425

You can try the below code:

examples = []
for _, row in Inputs.iterrows():
  example = tf.train.Example()
  for col, value in row.iteritems():
    example.features.feature[col].float_list.value.append(value)
  examples.append(example)
print(examples)

It's output will be a json as shown below:

[features {
  feature {
    key: "PetalLength"
    value {
      float_list {
        value: 5.900000095367432
      }
    }
  }
  feature {
    key: "PetalWidth"
    value {
      float_list {
        value: 2.0999999046325684
      }
    }
  }
  feature {
    key: "SepalLength"
    value {
      float_list {
        value: 7.099999904632568
      }
    }
  }
  feature {
    key: "SepalWidth"
    value {
      float_list {
        value: 3.0
      }
    }
  }
}
]

Then you can perform inference using the below command:

curl -d '{"inputs":examples}' \
  -X POST http://localhost:8501/v1/models/1554294699:predict

Upvotes: 0

Happy Gene
Happy Gene

Reputation: 492

You can use a Python REST client to make the call programatically, instead of manually composing the request. This is a sample code in tensorflow_serving github:

https://github.com/tensorflow/serving/blob/master/tensorflow_serving/example/resnet_client.py

Upvotes: 1

Related Questions