Meet Gemma: Open, fast language models from Gemini research
Gemma is a new family of small, useful language models made from ideas and tools behind the Gemini project.
These models come in two sizes, 2 billion and 7 billion, and you can use them as-is or further train them for your needs.
They read and write text well, and often beat other open models on many common tasks, so they feel surprisingly capable for their size.
We focused on safety and responsible release, because that matters to people and teams building apps.
The code and model files are open, which helps researchers and hobbyists to learn and build, while still paying attention to risks.
Gemma aims to be practical, friendly to use, and easy to run even on modest hardware, so more people can try it.
It's not magic, but it's a big step toward making smart language tech more open and usable.
Try it, explore, and show what you make — just use it carefully and thoughtfully.
Read article comprehensive review in Paperium.net:
Gemma: Open Models Based on Gemini Research and Technology
🤖 This analysis and review was primarily generated and structured by an AI . The content is provided for informational and quick-review purposes.
Top comments (0)