Unknown model architecture: qwen3

Open aropb opened this issue 7 months ago • 1 comments

Description

LLamaSharp 0.23.0 CUDA Windows LLM: Qwen3-8B-Q5_K_M.gguf

Errors: llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'qwen3' llama_model_load_from_file_impl: failed to load model

llama.cpp: https://github.com/ggml-org/llama.cpp/pull/12828

Apr 29 '25 07:04 aropb

Because supporting qwen3 requires a relatively new version of llama.cpp, but currently the latest version of llama.cpp supported by LLamaSharp only reaches b026.

Apr 29 '25 08:04 nipeone

wait https://github.com/SciSharp/LLamaSharp/pull/1179

May 03 '25 02:05 sangyuxiaowu

The new version that will come out soon works with qwen3 (updated binary). You will have to wait until the PR is merged.

May 09 '25 10:05 zsogitbe