Serving Preflight and Predict

Serving Preflight and Predict

Check model status, read the signature, then run prediction inference.

1. Confirm the ModelServer has the model loaded and reports a version in the AVAILABLE state before any inference is attempted.
2. Read the metadata of the latest available version to confirm the signature the prediction request will target, along with its expected input tensors.
3. Submit the prediction request against the model's latest available version, returning one prediction per supplied instance.

What you'll need to have ready

• modelName: The name the model was registered under in the ModelServer (e.g. "half_plus_two").
• signatureName: Optional signature to target. When omitted the ModelServer uses the model's default serving signature.
• instances: Row format input. An array of prediction inputs, where each element is a scalar, a list, or an object of named input tensors.

For your developer

Arazzo workflow tensorflow:predict-preflight, 3 operations: getModelStatus → getModelMetadata → predictModel. Specs on APIs.io.