keras-hub

New Models

• Translate Gemma: A multimodal variant of the Gemma 3 model fine-tuned for high-quality machine translation across 55 languages and capable to translate many more languages both directions, supporting both text and image inputs.
• SAM3 (Segment Anything Model 3): A next-generation computer vision model that introduces Promptable Concept Segmentation (PCS), allowin...

What's Changed

• Fix reversible embedding quantization by @pctablet505 in https://github.com/keras-team/keras-hub/pull/2476
• Add FunctionGemma checkpoints to kerasHub by @laxmareddyp in https://github.com/keras-team/keras-hub/pull/2480
• Models should reference ReversibleEmbedding from Keras core by @JyotinderSingh in https://github.com/keras-team/keras-hub/pull/2482
• Revert "Models shoul...

What's Changed

• Fix float16 overflow in Gemma3 by addressing precision-related instabilities.

Full Changelog: https://github.com/keras-team/keras-hub/compare/v0.25.0...v0.25.1

Summary:

New Models:

We've integrated new open-weight models to expand the capabilities of KerasHub, featuring specialized tools for function calling and safety, as well as high-performance open-source reasoning models:

• FunctionGemma: We have added support for FunctionGemma, a lightweight model from Google built on the Gemma 3 270M architecture. Designed specifically fo...

What's Changed

• Add Gemma3 Conversion script to port weights from HF directly by @laxmareddyp in https://github.com/keras-team/keras-hub/pull/2445
• Fix ESM attention for TFLite compatibility by @pctablet505 in https://github.com/keras-team/keras-hub/pull/2466
• Fix PARSeq decoder for TFLite compatibility by @pctablet505 in https://github.com/keras-team/keras-hub/pull/2467
• Fix SAM tests...