{"id":14459,"date":"2024-02-21T15:31:38","date_gmt":"2024-02-21T15:31:38","guid":{"rendered":"http:\/\/TheNextWeb=1404126"},"modified":"2024-02-21T15:31:38","modified_gmt":"2024-02-21T15:31:38","slug":"google-deepmind-has-a-new-family-of-open-ai-models-for-devs-gemma","status":"publish","type":"post","link":"https:\/\/www.londonchiropracter.com\/?p=14459","title":{"rendered":"Google DeepMind has a new family of open AI models for devs: Gemma"},"content":{"rendered":"\n

Just months after Google DeepMind unveiled Gemini<\/a> \u2014 its most capable AI<\/a> model ever \u2014 the London-based lab has released its compact offspring: Gemma.<\/p>\n

Named after the Latin word for \u201cprecious stone,\u201d Gemma is a new family of open models for developers and researchers.<\/p>\n

\u201cDemonstrating strong performance across benchmarks for language understanding and reasoning, Gemma is available worldwide starting today,\u201d Sundar Pichai, the CEO of Google, said on Twitter<\/a>.<\/span><\/p>\n

Gemma comes in two sizes \u2014 2 billion and 7 billion parameters. Each of them<\/a> has been released with pre-trained and instruction-tuned variants.<\/span><\/span><\/p>\n

\n
\n
<\/div>\n

\n

The <3 of EU tech<\/p>\n

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!<\/p>\n<\/div>\n<\/div>\n<\/div>\n

The lightweight models are descendants of Gemini. As a result, Gemma has inherited technical and infrastructure components from its parent, which enables \u201cbest-in-class performance,\u201d Google<\/a> said. <\/span><\/p>\n

As evidence, the tech titan revealed some eye-catching comparisons with Llama-2, a family of large language models (LLMs) released by Meta a year ago.<\/span><\/p>\n

\"Chart
<\/i><\/a>According to Google, Gemma\u2019s performance surpasses that of Meta\u2019s Llama 2. Credit: Google<\/figcaption>