PaddlePaddle provides a deep learning framework for training models across devices and platforms.
qkeras
QKeras provides quantization tools for deep learning with TensorFlow Keras.

About qkeras
QKeras is a quantization extension to Keras for TensorFlow that provides drop-in replacements for parameter-creating, activation and arithmetic layers to build quantized networks. Quantizing requires replacing layers such as Dense and Conv2D with QDense and QConv2D and quantizing math operations. Design follows Keras principles and aims to stay minimally intrusive.
Pricing is not clearly stated.
Key features
- Quantized recurrent cells including QSimpleRNN, QLSTM and QGRU
- Quantized separable, depthwise and transpose convolution variants
- Quantized activation functions including binary, ternary and quantized_bits with bit controls
- QAveragePooling2D that quantizes the pooling result
- QBatchNormalization is still in its experimental stage
- Not all functionality is safe with layer wrappers
- QActivation save and restore interface has not been fully tested
qkeras at a glance
- Website
- github.com/google/qkeras
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Details verified
- Oct 22, 2026 · Official source
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
qkeras alternatives
Tools and sites you could use instead, ranked by how closely they match.
The MegEngine framework supports deep learning and automatic differentiation, according to its repository description.
The linfa project provides a machine learning framework written in Rust.
FATE supports federated learning through an industrial grade framework.
Mosec serves machine learning models with dynamic batching and CPU or GPU pipelines for inference workloads.
The fuku-ml project provides a machine learning library described as simple to use.
Popular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Tell us and an editor will check it.
Claim the listing and suggest updates yourself. An editor checks every claim first.











