Larq Compute Engine Converter

The LCE converter allows you to convert a Keras model built with larq to an LCE-compatible TensorFlow Lite FlatBuffer format for inference.


Before installing the LCE converter, please install:

  • Python version 3.6 or 3.7
  • Tensorflow version 1.14, 1.15, 2.0.0, or 2.1.0 (recommended):
    pip install tensorflow  # or tensorflow-gpu

You can install the LCE converter with Python's pip package manager:

pip install larq-compute-engine

Converting a Larq model

Convert your model and write the file to disk:

import larq_compute_engine as lce

model = ... # Your custom Keras model or one from larq_zoo
with open("/tmp/my_model.tflite", "wb") as f: