PyCoral API overview | Coral (2024)

The PyCoral API (the pycoral module) is built atop the TensorFlow Lite PythonAPI to simplify your code when running aninference on the Edge TPU, and to provide advanced features for the Edge TPU such as modelpipelining across multiple Edge TPUs, and on-device transfer learning.

If you're new to this API, check out our guide to run inference on the Edge TPU withPython. For code examples, see the GitHub repo.

To install the library, see the software downloads page.

Module summary

• pycoral.utils.dataset

  Utilities to help process a dataset.

• pycoral.utils.edgetpu

  Utilities for using the TensorFlow Lite Interpreter with Edge TPU.

• pycoral.adapters.common

  Functions to work with any model.

  See Also

  What exactly is coral? [Get the Answer at BYJU’S]

• pycoral.adapters.classify

  Functions to work with a classification model.

• pycoral.adapters.detect

  Functions to work with a detection model.

• pycoral.pipeline.pipelined_model_runner

  The pipeline API allows you to run a segmented model across multiple Edge TPUs.

  For more information, see Pipeline a model with multiple EdgeTPUs.

• pycoral.learn.backprop.softmax_regression

  A softmax regression model for on-device backpropagation of the last layer.

• pycoral.learn.imprinting.engine

  A weight imprinting engine that performs low-shot transfer-learning for image classification models.

  For more information about how to use this API and how to create the type ofmodel required, seeRetrain a classification model on-device with weight imprinting.