Android z machine interpreter how to#

On Android, TensorFlow Lite inference can be performed using either Java or C++ĪPIs. Similarly, consistency with TensorFlow APIs was not anĮxplicit goal and some variance between languages is to be expected.Īcross all libraries, the TensorFlow Lite API enables you to load models, feed Should be no surprise that the APIs try to avoid unnecessary copies at theĮxpense of convenience.

TensorFlow Lite is designed for fast inference on small devices, so it In most cases, the API design reflects a preference for performance over ease of Linux, in multiple programming languages. TensorFlow inference APIs are provided for most common mobile/embedded platforms You to map the probabilities to relevant categories and present it to your Tensors in a meaningful way that's useful in your application.įor example, a model might return only a list of probabilities. When you receive results from the model inference, you must interpret the Tensors, as described in the following sections. Involves a few steps such as building the interpreter, and allocating This step involves using the TensorFlow Lite API to execute the model. For example, you might need to resize an image orĬhange the image format to be compatible with the model. Raw input data for the model generally does not match the input data formatĮxpected by the model. tflite model into memory, which contains the model's TensorFlow Lite inference typically follows the following steps:

Android z machine interpreter how to#

This page describes how to access to the TensorFlow Lite interpreter and performĪn inference using C++, Java, and Python, plus links to other resources for each The interpreter uses a static graph ordering and a custom (less-dynamic) memoryĪllocator to ensure minimal load, initialization, and execution latency. The TensorFlow Lite interpreter is designed to be lean and fast. Inference with a TensorFlow Lite model, you must run it through an On-device in order to make predictions based on input data. The term inference refers to the process of executing a TensorFlow Lite model