Starble comments

Repositories
Issues
Comments

Results 1 comments of


                                            Starble

Qwen3-8B and other models generate garbage output / repeat tokens (GGGGGG...) in llama.cpp via LM Studio Vulkan backend

I have the same issue too, however this is for GLM-4-32B using Q_4_K_S from [bartowski's quant](https://huggingface.co/bartowski/THUDM_GLM-4-32B-0414-GGUF), however some investigation at different HF discussions appears the issue specifically for GLM-4-32B comes...