# ADI MAX78000/MAX78002 Model Training and Synthesis

A model-training and code-synthesis toolchain for Analog Devices' MAX78000 and MAX78002 ultra-low-power machine-learning accelerator chips, converting a trained PyTorch model into C code for the device.

This repository is the synthesis half of a five-repository project for Analog Devices' MAX78000 and MAX78002 edge AI chips: a companion training repository handles model development and training in PyTorch, while this repository's "izer" tool converts a trained model into C code that runs on the device. A software development kit repository supplies drivers and example programs for the evaluation kits, and a reference-design repository provides host and sample applications such as a camera reference design.

The project documents installation requirements, including a specific PyTorch version, and file-system layout expectations for combining the training and synthesis repositories with the software development kit. Support for TensorFlow and Keras is deprecated in favor of PyTorch.

It is written in Python and released under the Apache-2.0 license.

Website: https://github.com/analogdevicesinc/ai8x-synthesis