American tech giant Google has recently unveiled Gemma, a “family of lightweight, state-of-the-art open models”. The models were developed by Google DeepMind with the help of multiple teams at Google.
“Today, we’re excited to introduce a new generation of open models from Google to assist developers and researchers in building AI responsibly”, the company stated in a press release.
Gemma is built on the same technology as Gemini, Google’s” largest and most capable AI model”. The models come in two weight sizes: Gemma 2B and Gemma 7B with each size implementing pre-trained and instruction-tuned variants.
Additionally, the company has also released several tools to help developers innovate new AI applications. Gemma comes packaged with “Ready-to-use Colab and Kaggle notebooks”. The model also provides extensive cross-device compatibility as it works on laptops, desktops, IoT, mobile, and cloud.
See Related: Polygon Teams Up With Google Cloud To Advance Web 3
Google’s Collaboration With NVIDIA
Another notable aspect of Gemma is its optimization for NVIDIA GPUs as part of Google’s collaboration with NVIDIA.
The rapid advancement of generative AI has given rise to many safety and ethical concerns. Google has addressed this issue by stating, “We’re also releasing a new Responsible Generative AI Toolkit together with Gemma to help developers and researchers prioritize building safe and responsible AI applications”. The toolkit includes powerful safety classifiers, a debugging tool, and general guidelines for building responsible AI applications.