Google has introduced Gemma, a new family of lightweight open models, just a week after unveiling the latest iteration of its Gemini models.
Gemma open models
Gemma is a family of lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models. Developed by Google DeepMind and other teams across Google, it is inspired by Gemini, and the name reflects the Latin gemma, meaning “precious stone.” Accompanying our model weights, we’re also releasing tools to support developer innovation, foster collaboration, and guide responsible use of Gemma models.
Gemma is available worldwide, starting today. Here are the key details to know:
- We’re releasing model weights in two sizes: Gemma 2B and Gemma 7B. Each size is released with pre-trained and instruction-tuned variants.
- A new Responsible Generative AI Toolkit provides guidance and essential tools for creating safer AI applications with Gemma.
- We’re providing toolchains for inference and supervised fine-tuning (SFT) across all major frameworks: JAX, PyTorch, and TensorFlow through native Keras 3.0.
- Ready-to-use Colab and Kaggle notebooks, alongside integration with popular tools such as Hugging Face, MaxText, NVIDIA NeMo and TensorRT-LLM, make it easy to get started with Gemma.
- Pre-trained and instruction-tuned Gemma models can run on your laptop, workstation, or Google Cloud with easy deployment on Vertex AI and Google Kubernetes Engine (GKE).
- Optimization across multiple AI hardware platforms ensures industry-leading performance, including NVIDIA GPUs and Google Cloud TPUs.
- Terms of use permit responsible commercial usage and distribution for all organizations, regardless of size.
![Gemma vs Llama-2](https://i0.wp.com/nosisnews.com/wp-content/uploads/2024/02/image-138.png?resize=1000%2C615&ssl=1)
![Gemma vs Llama-2](https://i0.wp.com/nosisnews.com/wp-content/uploads/2024/02/image-138.png?resize=1000%2C615&ssl=1)
In a press briefing, Janine Banks from Google emphasized the company’s commitment to openness and explained the distinction between open models and open source. Developers can use Gemma models for inferencing and fine-tuning, with the potential for various use cases due to improved generation quality over the past year.
Tris Warkentin, Google DeepMind’s product management director, highlighted the enhanced capabilities of smaller models, enabling AI applications to run locally on developers’ machines or on cloud platforms like GCP with Cloud TPUs. This development signifies a shift in AI application development possibilities, unlocking new opportunities for developers.
While Google’s competitors also offer open models with similar features, the real-world performance of the models remains to be seen. To complement Gemma’s release, Google is launching a responsible generative AI toolkit, providing guidance and essential tools for creating safer AI applications with Gemma.
Additionally, a debugging tool will be made available, enhancing the overall development and deployment experience for developers working with Gemma models.