TensorFlow TensorFlow Deployment and TFLite

Question 1

What is TensorFlow Lite (TFLite) designed for?

Accepted Answer

Deploying models on mobile and embedded devices

Answer

TFLite is a lightweight runtime optimized for on-device inference on mobile phones, microcontrollers, and IoT devices.

Question 2

What is quantization in TFLite model optimization?

Accepted Answer

Reducing model weight precision (e.g., float32 to int8) to decrease size and improve speed

Answer

Quantization reduces the numerical precision of model weights and activations, significantly reducing model size and inference latency.

Question 3

Which class is used to convert a Keras model to TFLite format?

Accepted Answer

tf.lite.TFLiteConverter

Answer

tf.lite.TFLiteConverter converts TensorFlow models (SavedModel or Keras model) into the TFLite FlatBuffer format.

Question 4

What is the TensorFlow SavedModel format?

Accepted Answer

A directory containing model architecture, weights, and computation graph

Answer

SavedModel is TensorFlow's standard serialization format that saves the complete model including computation graph, weights, and metadata.

Question 5

What does TensorFlow Serving provide?

Accepted Answer

A production-grade server for serving TensorFlow models via REST and gRPC APIs

Answer

TensorFlow Serving is a flexible, high-performance serving system for machine learning models designed for production environments.

TensorFlow Practice Test