Publish your Keras models on Kaggle and Hugging Face- Google Developers Blog
May 7, 2024Martin Görner

Summary
This post introduces the ability to upload and share fine-tuned Keras models directly to Kaggle and Hugging Face. It details the integration of `keras.save_to_preset()` and `keras_nlp.upload_preset()` for seamless model distribution. The post highlights the benefits of Keras as a preferred model format due to its consistent API across JAX, TensorFlow, and PyTorch, and showcases the success of Gemma models on Kaggle. It also explains how KerasNLP models are now first-class citizens on Hugging Face, allowing direct loading and uploading.
More from the Google Developer Platform trail
- Autonomous LLM post-training with Tunix on TPUs- Google Developers Blog
- HeyGen x Google Cloud: Bringing Avatar IV to TPUs- Google Developers Blog
- Run Ray on TPU, Part 2: Ray AI libraries- Google Developers Blog
- Scaling Agentic RL: High-Throughput Agentic Training with Tunix- Google Developers Blog
- Run Ray on TPU, Part 1: The foundations- Google Developers Blog