module TensorflowLite::EdgeTPU

Overview

An Edge TPU (Tensor Processing Unit) is a small ASIC (Application-Specific Integrated Circuit) designed for accelerating machine learning workloads at the edge, that is, on devices like smartphones, IoT devices, and embedded systems.

Google's Coral product line includes various devices that incorporate the Edge TPU, such as the Coral Dev Board, the Coral USB Accelerator, and various system-on-modules (SoMs) and PCI-E cards.

Defined in:

tensorflow_lite/edge_tpu.cr

Class Method Summary

• .devices : Array(Device)

  the list of Edge TPU devices available on the system

• .version : String

  the version of the edge tpu library in use

Class Method Detail