TensorFlow · Schema

TensorFlow Serving Prediction Response

Schema for a prediction inference response from TensorFlow Serving. Returns predictions in row format or outputs in column format.

Artificial IntelligenceDeep LearningJavaScriptMachine-LearningModel ServingNeural NetworksOpen-SourcePython

Properties

Name Type Description
predictions object Row format predictions. One prediction per input instance. Can be a scalar, list, or object.
outputs object Column format outputs. Map of output tensor names to arrays of values.
error string Error message if the prediction failed
View JSON Schema on GitHub

JSON Schema

tensorflow-serving-prediction-response-schema.json Raw ↑
{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "$id": "https://github.com/api-evangelist/tensorflow/blob/main/json-schema/tensorflow-serving-prediction-response-schema.json",
  "title": "TensorFlow Serving Prediction Response",
  "description": "Schema for a prediction inference response from TensorFlow Serving. Returns predictions in row format or outputs in column format.",
  "type": "object",
  "properties": {
    "predictions": {
      "description": "Row format predictions. One prediction per input instance. Can be a scalar, list, or object.",
      "oneOf": [
        {
          "type": "array",
          "description": "Array of predictions, one per input"
        },
        { "type": "number" },
        { "type": "string" }
      ]
    },
    "outputs": {
      "type": "object",
      "description": "Column format outputs. Map of output tensor names to arrays of values.",
      "additionalProperties": {
        "type": "array"
      }
    },
    "error": {
      "type": "string",
      "description": "Error message if the prediction failed"
    }
  },
  "examples": [
    {
      "predictions": [1.0, 2.0, 3.0]
    },
    {
      "predictions": [
        {"output_label": "cat", "score": 0.92},
        {"output_label": "dog", "score": 0.08}
      ]
    },
    {
      "outputs": {
        "output_label": ["cat", "dog"],
        "score": [0.92, 0.87]
      }
    }
  ]
}