TensorFlow Lite is an open source deep learning framework for on-device inference.
Pick a new model or retrain an existing one.
Convert a TensorFlow model into a compressed flat buffer with the TensorFlow Lite Converter.
Take the compressed .tflite file and load it into a mobile or embedded device.
Quantize by converting 32-bit floats to more efficient 8-bit integers or run on GPU.
Explore optimized models to help with common mobile and edge use cases.
See more ways to participate in the TensorFlow community.
Post-training quantization produces integer quantized models that are fast on CPU, and can be deployed in accelerators
Run inference on GPU can improve inference up to ~4x on Pixel 3.
In this three-part blog series, we walk through how we designed a simple system that demonstrates how computers can see, understand, and react to the world around us.
The new optimization toolkit in TensorFlow provides a suite of techniques that developers—both novice and advanced—can use to optimize machine learning models.