[Bug] Wan 2.2 unknown generation

[Naster17]

Git commit

rev: 636d3cb
commit: master-504-636d3cb

Using Wan2.2 5B Vulkan, Fedora 43, Messa 25.x.x, linux kernel 6.16.5

Im trying multiple combinations of commands, parameters, etc. But every time im receiving this: (originally .avi change file extension to .avi cuz github dont support avi files)

---

./bin/Release/sd-cli -M vid_gen --diffusion-model  ..\..\ComfyUI\models\diffusion_models\wan2.2_ti2v_5B_fp16.safetensors --vae ..\..\ComfyUI\models\vae\wan2.2_vae.safetensors --t5xxl ..\..\ComfyUI\models\text_encoders\umt5-xxl-encoder-Q8_0.gguf  -p "a lovely cat" --cfg-scale 6.0 --sampling-method euler -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 480 -H 832 --diffusion-fa --offload-to-cpu --video-frames 33 --flow-shift 3.0

---

build/bin/sd-cli -M vid_gen --diffusion-model  ~/Downloads/models/SD/Wan2.2-TI2V-5B-Q8_0.gguf --vae ~/Downloads/models/SD/wan2.2_vae.safetensors --t5xxl ~/Downloads/models/SD/umt5-xxl-encoder-Q8_0.gguf  -p "a lovely cat watching to camera"  -v -W 512 -H 512 --video-frames 24 --color -t 12 --diffusion-fa --vae-tiling --flow-shift 3.0 -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" --sampling-method euler

---

# Also with diferent combinations of -fa -W -H --sampling-method  --vae-tiling

output.mp4

Operating System & Version

Fedora 43, Linux 6.16.5

GGML backends

Vulkan

Command-line arguments used

build/bin/sd-cli -M vid_gen --diffusion-model ~/Downloads/models/SD/Wan2.2-TI2V-5B-Q8_0.gguf --vae ~/Downloads/models/SD/wan2.2_vae.safetensors --t5xxl ~/Downloads/models/SD/umt5-xxl-encoder-Q8_0.gguf -p "a lovely cat watching to camera" -v -W 512 -H 512 --video-frames 24 --color -t 12 --diffusion-fa --vae-tiling --flow-shift 3.0 -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" --sampling-method euler

Steps to reproduce

...

What you expected to happen

Broken video output

What actually happened

Broken video output

Logs / error messages / stack trace

No response

Additional context / environment details

No response