Google Gemma

About

Google Gemma is a family of lightweight, state-of-the-art open-source large language models developed by Google DeepMind, leveraging core technology from Google's Gemini model family. This series offers various parameter scales, including 1B, 4B, 12B, and 27B, designed to cater to diverse application requirements from edge devices to high-performance servers. Gemma boasts robust multimodal understanding capabilities, supporting both text and image inputs, and can process ultra-long contexts of up to 128K tokens. Its efficient architecture allows it to run smoothly on a single GPU or even a personal laptop, significantly lowering the barrier for local deployment and development. Furthermore, Gemma provides extensive support for major deep learning frameworks such as JAX, PyTorch, and Keras, and can be conveniently deployed locally via platforms like Ollama and Hugging Face, offering researchers and developers a highly flexible tool for lightweight applications, rapid prototyping, and AI deployments in resource-constrained environments.