[Stable Diffusion] Prompt Sharing and Learning Thread

[modine2021]

anyone know how to fix this? of of a sudden Comfy decided it dont want to render no more. but it worked before. i get this error using Wan sampler AND Ksampler now::



WanVideoSampler (happens with KSampler also)
Command '['M:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\runtime\\tcc\\tcc.exe', 'X:\\temp\\tmpmaemgsyk\\__triton_launcher.c', '-O3', '-shared', '-Wno-psabi', '-o', 'X:\\temp\\tmpmaemgsyk\\__triton_launcher.cp313-win_amd64.pyd', '-fPIC', '-D_Py_USE_GCC_BUILTIN_ATOMICS', '-lcuda', '-lpython313', '-LM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\lib', '-LM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\lib\\x64', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\include', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\include', '-IX:\\temp\\tmpmaemgsyk', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Include']' returned non-zero exit status 1.

 

[modine2021]

anyone know how to fix this? of of a sudden Comfy decided it dont want to render no more. but it worked before. i get this error using Wan sampler AND Ksampler now::



WanVideoSampler (happens with KSampler also)
Command '['M:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\runtime\\tcc\\tcc.exe', 'X:\\temp\\tmpmaemgsyk\\__triton_launcher.c', '-O3', '-shared', '-Wno-psabi', '-o', 'X:\\temp\\tmpmaemgsyk\\__triton_launcher.cp313-win_amd64.pyd', '-fPIC', '-D_Py_USE_GCC_BUILTIN_ATOMICS', '-lcuda', '-lpython313', '-LM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\lib', '-LM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\lib\\x64', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\include', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Lib\\site-packages\\triton\\backends\\nvidia\\include', '-IX:\\temp\\tmpmaemgsyk', '-IM:\\AI\\SwarmUI\\dlbackend\\python_embeded\\Include']' returned non-zero exit status 1.

EDIT: never mind. fixed by reinstalling triton

 

[JhonLui]

EDIT: never mind. fixed by reinstalling triton

I've noticed that anytime there is an update involving Torch and/or Torch components, you have to reinstall triton, as it will be deactivated.

 

[rogue_69]

Appreciate this thread, thanks to all who have contributed. I'm releasing a fairly massive Twine game soon-ish that uses AI images, as a necessary evil, despite my dislike of most all things AI. This thread definitely helped some.

You're setting yourself up for failure if you "dislike most things AI" while also needing to use AI as a necessary evil. Either use it and embrace it, or don't use it. Otherwise you'll feel shame about it, and won't really enjoy your results no matter how things turn out. Not picking on you, just trying to give some life advice.

 

[rogue_69]

Wan 2.2 starting off with the image and then using "hard cut" to get to the first actual frame of the video. Then used the second to last frame of each video to start a new one. I feel like the character consistency was acceptable.

 

[HinsdaleDays]

You're setting yourself up for failure if you "dislike most things AI" while also needing to use AI as a necessary evil. Either use it and embrace it, or don't use it. Otherwise you'll feel shame about it, and won't really enjoy your results no matter how things turn out. Not picking on you, just trying to give some life advice.

I get what you're saying, but we're talking about AI porn here. I think talking about feeling "shame" around this is a little bit in the rear view mirror