Closed tinkle-bell closed 2 days ago
There are still PRs open in llama.cpp for proper gemma support (e.g. https://github.com/ggerganov/llama.cpp/pull/8197) which I'm waiting on before starting a backend update :)
The PR https://github.com/ggerganov/llama.cpp/pull/8197 is merged, do you have any ETA for this feature?
I've started on a binary update, but it's held up at the moment due to an issue in the MacOS build pipeline.
Gemma2 is supported now that 0.14.0 is released 🥳
Background & Description
I think the backend should be updated to support for gemma2.
API & Usage
No response
How to implement
Update the backend.