Keras Community Newsletter - Feb 2026

[Yufeng Guo]

Keras Community Updates

February 2026

Welcome to the Newsletter

Welcome to the first Keras community newsletter, we're delighted to have you in the Keras community! We'll be publishing these newsletters following community meetings as a way to share updates more broadly. We hope you find it informative and interesting!

Since the last update, the team has been hard at work adding new models to KerasHub, additional Quantization support, LiteRT export features, and online content, as well as making updates to keras.io.

Happy machine learning,
Yufeng on behalf of the Keras team

New Models in KerasHub

New models have been added to KerasHub!

• Gemma Family: 

  • Translate Gemma

  • MedGemma 1.5 Variant

  • Function Gemma

• VideoPrism

• GPT-OSS and GPT-OSS-SAFEGUARD

• Qwen2.5-CODER, Qwen3-CODER, Qwen2.5-Math

• RWKV7

• SAM3(PCS)

KerasHub ↗︎

INT4 Sub-Channel Quantization

keras.quantizers.Int4QuantizationConfig

Previously, Keras INT4 quantization only supported "per-channel" quantization (one scale factor per output channel).

This release adds support for fine-grained quantization by grouping weights (e.g., blocks of 64 or 128 parameters) along the input dimension, which significantly reduces quantization error compared to per-channel approaches.

Learn more ↗︎

LiteRT Export

model.export(format='litert')

You can now use model.export to convert models to .tflite format. This enables you to use keras and keras-hub models on edge devices, IoT, Android, iOS, etc.

Docs ↗︎

Recent YouTube Videos

Keras Turns 10: A decade of deep learning

Join Francois and Matt as they reminisce about 10 years of Keras!

Watch on YouTube ↗︎

How to use KerasHub with Hugging Face

Learn how KerasHub allows you to mix and match model architectures with pre-trained weights from Hugging Face Hub seamlessly.

Watch on YouTube ↗︎

Keras Recommenders: ranking and retrieval

Introduction to Keras Recommenders (KerasRS), a library designed to help developers build reliable ranking and retrieval models with ease.

Watch on YouTube ↗︎

Keras YouTube Playlist

2026 Google TPU Research & Education Awards

TPUs, Funding, and GCP Credits

A new compute & funding opportunity for researchers, supporting researchers who are pushing the boundaries of ML systems, performance, and efficiency, and those working in high-impact applied science.

If you or a colleague are exploring what is possible with TPUs, consider applying!

Learn more ↗︎

New Guides

• Function calling guide (Thanks to Laxma Reddy)

  • Function Calling With KerasHub

• Quantization guides (Thanks to Jyotinder Singh)

  • Quantization in Keras

  • INT8 Quantization in Keras

  • INT4 Quantization in Keras

  • GPTQ Quantization in Keras

  • Writing Quantization-Compatible Layers in Keras

Keras.io Website updates

Links from APIs to examples

APIs now link directly to relevant guides and examples.

Book chapters are now linked

Chapters from Deep Learning with Python are linked in guides where relevant.

Copy button for code snippets

Easier copy paste is now just a click away!

Tex equation support

You can now write LaTeX in markdown for guides and examples.

Call for Contributions

Looking for a way to contribute to Keras? Start here!

• https://github.com/keras-team/keras-hub/issues/1836

• https://github.com/keras-team/keras/issues/19519

• https://github.com/keras-team/keras-rs/issues 

Keras Github

You're receiving this email because you're subscribed to the Keras Users group. To stop receiving these emails, update your subscription settings in the group page.