result not as expected

[lamfan]

swift cli: swift run StableDiffusionSample "a photo of an astronaut riding a horse on mars" --resource-path <output-mlpackages-directory>/Resources/ --seed 93 --output-path </path/to/output/image>

model: https://huggingface.co/apple/coreml-stable-diffusion-1-4-palettized/blob/main/coreml-stable-diffusion-1-4-palettized_original_compiled.zip

the result is a picture with lot of small face (around 10x12 small square), not as a man riding a horse, why?

I clone this repo and conda to setup a correct env, what do I miss? please help, thanks!

macOS Sonoma 14.0

[czkoko]

I think it's possible that you are using Intel CPU.

[lamfan]

yes, is that possible to run on Intel CPU?

[czkoko]

Adding this parameter should generate a normal picture, but it is quite slow. You can try it. --compute-units cpuOnly

[lamfan]

Adding this parameter should generate a normal picture, but it is quite slow. You can try it. --compute-units cpuOnly

Thank you so much @czkoko

It take around 60x times to complete the image.

[czkoko]

Intel cpu can use diffusionBee app

[lamfan]

Intel cpu can use diffusionBee app

Thanks, will try to use it!