Skip to content

Model files: EmbeddingGemma 2 (GGUF)

Choose a tag to compare

@ScriptSage1 ScriptSage1 released this 13 Nov 02:46
· 31 commits to main since this release

Model files that Juno downloads for search by meaning. This is not an app release.

File Size SHA-256 Used for
embeddinggemma-2-text-q8_0.gguf 303,957,280 bytes 111e84992a0406bacb70d11dddd8874001c783cc56b82c462ec82a2f5b8d08b1 Search by meaning (text)
embeddinggemma-2-mmproj.gguf 980,895,232 bytes 74bf65c860d292a1eebafd7ca85b44532a2528f03017f9204fe1695916d2c66f Picture search (loaded only while pictures are indexed)

Source and licence. These are GGUF conversions of EmbeddingGemma 2 by Google DeepMind, for use with llama.cpp. The text model is quantised to 8 bits. The model is licensed under Apache 2.0 (see LICENSE-Apache-2.0.txt and the model card). Use of the model must follow the Gemma Prohibited Use Policy. Juno is not affiliated with Google.

Juno checks the SHA-256 of each file after downloading it.