Seldon · Schema

Seldon Inference Request (Open Inference Protocol V2)

Schema for an inference request payload sent to a Seldon model endpoint following the Open Inference Protocol V2.

MLOpsMachine LearningModel ServingInferenceKubernetesAI OperationsDrift DetectionExplainabilityCanary DeploymentA/B TestingLLMOps

Properties

Name Type Description
id string Optional unique identifier for the inference request.
inputs array List of input tensors for the inference request.
outputs array Optional list of requested output tensors.
parameters object Optional request-level parameters.
View JSON Schema on GitHub

JSON Schema

seldon-inference-request.json Raw ↑ 
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://raw.githubusercontent.com/api-evangelist/seldon/main/json-schema/seldon-inference-request.json",
  "title": "Seldon Inference Request (Open Inference Protocol V2)",
  "description": "Schema for an inference request payload sent to a Seldon model endpoint following the Open Inference Protocol V2.",
  "type": "object",
  "properties": {
    "id": {
      "type": "string",
      "description": "Optional unique identifier for the inference request."
    },
    "inputs": {
      "type": "array",
      "description": "List of input tensors for the inference request.",
      "items": {
        "$ref": "#/$defs/InferenceTensor"
      },
      "minItems": 1
    },
    "outputs": {
      "type": "array",
      "description": "Optional list of requested output tensors.",
      "items": {
        "$ref": "#/$defs/InferenceTensor"
      }
    },
    "parameters": {
      "type": "object",
      "description": "Optional request-level parameters.",
      "additionalProperties": true
    }
  },
  "required": ["inputs"],
  "$defs": {
    "InferenceTensor": {
      "type": "object",
      "title": "InferenceTensor",
      "description": "A named tensor with shape, datatype, and data.",
      "properties": {
        "name": {
          "type": "string",
          "description": "The name of the tensor."
        },
        "shape": {
          "type": "array",
          "description": "The shape of the tensor as an array of dimensions.",
          "items": {
            "type": "integer"
          }
        },
        "datatype": {
          "type": "string",
          "description": "The datatype of the tensor.",
          "enum": ["BOOL", "UINT8", "UINT16", "UINT32", "UINT64", "INT8", "INT16", "INT32", "INT64", "FP16", "FP32", "FP64", "BYTES", "STRING"]
        },
        "data": {
          "description": "The tensor data as a flattened array or nested array matching the shape.",
          "oneOf": [
            { "type": "array" },
            { "type": "string" }
          ]
        },
        "parameters": {
          "type": "object",
          "description": "Optional per-tensor parameters.",
          "additionalProperties": true
        }
      },
      "required": ["name", "shape", "datatype", "data"]
    }
  },
  "examples": [
    {
      "id": "req-001",
      "inputs": [
        {
          "name": "input",
          "shape": [1, 4],
          "datatype": "FP32",
          "data": [5.1, 3.5, 1.4, 0.2]
        }
      ]
    }
  ]
}