TF Serving Predict API Output Interpretation

Hi,

Is TFS predict API output the same as the output of tf.keras.model.predict method?

For example, if we have a tf.keras.model compiled with BinaryAccuracy metric, will the output of the TFS predict API be a list of binary accuracy values for each one of the inputs of the predict request?

Thanks in advance!

Hello @Peter_1
Thank you for using TensorFlow
The Predict API for TF serving and model.predict method both give the percentage of prediction not directly the accuracy, If it is a binary accuracy metric, it will still return predictions (generally any one of the class with higher percentage).We have to manually calculate accuracy comparing true labels with predictions.