unknown model architecture: 'gemma-embedding'

[admiralakber]

Hi there,

Thanks for the Rust bindings.

When trying to load the recent EmbeddingGemma 300m model I get the following error using library version 0.1.121.

llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'gemma-embedding'

This model has been supported in upstream llama.cpp in the following release: https://github.com/ggml-org/llama.cpp/releases/tag/b6384

[MarcusDunn]

a PR would be awesome. Generally updates are pretty simple.

[admiralakber]

Done, PR made. I brought us up to the latest release of the time of me doing the work too.
There were some API changes, but nothing too bad.

Builds and tests work, and testing my fork/branch in my project utilising embeddinggemma all working now too.

[MarcusDunn]

Awesome! Thanks for the PR. I'll merge and cut a release soon as checks are passing.