[Bug] GGUF error,unsupported src0 type: q3_K

[zwukong]

Logs

[INFO ] model.cpp:1528 - loading tensors completed, taking 1.28s (process: 0.00s, read: 1.05s, memcpy: 0.00s, convert: 0.00s, copy_to_backend: 0.06s)
[DEBUG] lora.hpp:86 - finished loaded lora
[DEBUG] lora.hpp:526 - (1200 / 1200) LoRA tensors will be applied
[DEBUG] ggml_extend.hpp:1600 - lora compute buffer size: 7214.28 MB(VRAM)
[DEBUG] lora.hpp:526 - (1200 / 1200) LoRA tensors will be applied
D:\a\stable-diffusion.cpp\stable-diffusion.cpp\ggml\src\ggml-cuda\getrows.cu:204: ggml_cuda_get_rows_switch_src0_type: unsupported src0 type: q3_K

Operating System & Version

windows11

GGML backends

CUDA

Command-line arguments used

sd.exe -M vid_gen --diffusion-model M:\ComfyUI\models\diffusion_models\Wan2.2-T2V-A14B-HighNoise-Q3_K_M_2.gguf --high-noise-diffusion-model M:\ComfyUI\models\diffusion_models\Wan2.2_T2V_A14B_4steps_25-09-28_Dyno_High_lightx2v-Q3_K_M.gguf --vae M:\ComfyUI\models\vae\wan_2.1_vae.safetensors --t5xxl M:\ComfyUI\models\text_encoders\umt5-xxl-encoder-Q4_K_M.gguf -p "a lovely catlora:|low_noise|Wan22_A14B_T2V_LOW_Lightning_4steps_lora_250928_rank64_fp16:1lora:|high_noise|Wan22_A14B_T2V_HIGH_Lightning_4steps_lora_250928_rank128_fp16:1" --cfg-scale 3.5 --sampling-method euler --steps 4 --high-noise-cfg-scale 3.5 --high-noise-sampling-method euler --high-noise-steps 4 -v -n "色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走" -W 832 -H 480 --diffusion-fa --offload-to-cpu --lora-model-dir M:\ComfyUI\models\loras --video-frames 33 --flow-shift 3.0

[leejet]

This PR #969 should fix the issue.