Google Releases EmbeddingGemma 2, Outperforming Rivals Twice Its Size
Google has launched EmbeddingGemma 2, a 740-million parameter open model capable of converting text, images, video, audio, and code into vectors. Requiring only 191 MB of RAM, it runs efficiently on-device. This matters because it enables developers to build offline retrieval-augmented generation applications locally without transmitting sensitive data to external servers.
Source: The Decoder