mradermacher

/

model_requests

like 132

All GLM-4 based models are currently on halt as the latest llama.cpp version seam to generate broken quants for thous type of models. We are currently waiting for a fix from the llama.cpp team.

See https://huggingface.co/mradermacher/model_requests/discussions/844#67fe8fb6046aebf7c8dcdcf5 for more information.

We need to wait for https://github.com/ggml-org/llama.cpp/pull/12957 to get merged.

Or more preciesely the follow up pull request https://github.com/ggml-org/llama.cpp/pull/13021 to be merged

as a sidenote, it should be safe to queue in advance - they are currently held in trhe queue on kaos (and i queued this one)