This notebook demonstrates the process of training a 2.5 kB model using TensorFlow and converting it for use with TensorFlow Lite for Microcontrollers.

Deep learning networks learn to model patterns in underlying data. Here, we're going to train a network to model data generated by a sine function. This will result in a model that can take a value, x, and predict its sine, y.

The model created in this notebook is used in the hello_world example for TensorFlow Lite for MicroControllers.

Configure Defaults

# Define paths to model files
import os
MODELS_DIR = 'models/'
if not os.path.exists(MODELS_DIR):
    os.mkdir(MODELS_DIR)
MODEL_TF = MODELS_DIR + 'model'
MODEL_NO_QUANT_TFLITE = MODELS_DIR + 'model_no_quant.tflite'
MODEL_TFLITE = MODELS_DIR + 'model.tflite'
MODEL_TFLITE_MICRO = MODELS_DIR + 'model.cc'

Deploying TinyML: Train a Simple TensorFlow Lite for Microcontrollers model

Configure Defaults