Description
TensorFlow Lite models can run from Python on lightweight inference workflows. Developers can load compact models, feed input tensors, and integrate local inference into applications or edge-device tools. It processes caller-supplied model inputs, which may include sensitive data.