keras-sd/decoder-tflite

This repository hosts the TFLite version of decoder part of KerasCV Stable Diffusion.

Stable Diffusion consists of text encoder, diffusion model, decoder, and some glue codes to handl inputs and outputs of each part. The TFLite version of decoder in this repository is built not only with the decpder itself but also TensorFlow operations that takes latent from diffusion model and generates final images.

TFLite conversion was based on the SavedModel from this repository, and TensorFlow version >= 2.12-nightly was used.

• NOTE: Dynamic range quantization was used.
• NOTE: TensorFlow version < 2.12-nightly will fail for the conversion process.
• NOTE: For those who wonder how SavedModel is constructed, find it in keras-sd-serving repository.