Error on generate_image.py while diffusion runs

[uls0]

Hello! im using the mjtimblin/epaper-slow-generative-art fork the only different thing is that im not connecting the e-ink paper because i want to test with the hdmi monitor.

im getting this error while running ./bin/run.sh

---------[diffusion]---------
Traceback (most recent call last):
File "/root/epaper-slow-generative-art/src/generate_image.py", line 227 in <module>
     main()
     File "/root/epaper-slow-generative-art/src/generate_image.py", line 196, in main
           generate_sd_image(prompt, sd_image_filepath)
     File "/root/epaper-slow-generative-art/src/generate_image.py", line 54, in generate_sd_image
                                 raise Exception('Image generation failed')
Exception: image generation failed

i've already try to fix it editing the screen size, the steps (down to 3), changing the resolution of the monitor but nothing works.

could you point me in the right direction?

[mjtimblin]

Hi @uls0 That error message is triggered whenever the stable diffusion binary exits with a non-zero code. I admit that I don't know much about the inner workings of the OnnxStream project, but can you try calling the stable diffusion binary directly to see if you get an error message? e.g.

[path to sd binary] \
  --models-path "[path to weights dir]" \
  --steps 5 \
  --rpi \
  --neg-prompt "ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, extra limbs, disfigured, deformed, body out of frame, bad anatomy, watermark, signature, cut off, low contrast, underexposed, overexposed, bad art, beginner, amateur, distorted face, blurry, draft, grainy" \
  --prompt "a photo of a cute kitten playing with a ball of yarn" \
  --output "output.png"

In the future, please open issues on mjtimblin/epaper-slow-generative-art when using my fork. I'm grateful for @rvdveen's work, and I don't want him to be responsible for any of my code changes or bugs I may introduce 😄