Skip to content

Guide for quantization, conversation of the tensorflow model to tensorflow lite

License

Notifications You must be signed in to change notification settings

lusinlu/tensorflow-lite-guide

Repository files navigation

Guide on tflite quantization and converting

Repository contains examples on how to do quantization aware training, and converting to tflite freezed graphdef, keras model, or checkpoint file.

Requirements

tensorflow 1.4 or higher python 3

pip3 install -r requirements.txt

Running quantization aware training

Script contains training/exporting/testing of the dummy model for the image-to-image transformation

python3 quantization_aware_training.py

for the export add --mode export, and for the testing --mode test

Converting to tflite trained model

there is a 3 different options for obtaining tflite model - from graphdef, from the keras model and from the checkpoint. Also you can choose to do post training quantization for reducing the model size, which will convert weights to uint8.

python3 convert_tf_model.py --model_type "type of the trained model" --model_path "path to the trained model"

Useful resources

Official documentation of the Tensorflow Lite
Blog post with explanation of examples

About

Guide for quantization, conversation of the tensorflow model to tensorflow lite

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages