R
RunPod11mo ago
Raios

Image Generation problem

NansException: A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check. why this happens when i try to generate a pic?
4 Replies
ashleyk
ashleyk11mo ago
1. Which template are you using? 2. Which GPU type are you using? 3. Are you using Secure Cloud or Community cloud and which region? 4. Did you perhaps start your pod without a GPU?
Raios
RaiosOP11mo ago
Stable Diffusion, RTX A4000, ive backed up files on cloud but im using community storage, EU no its running with gpu i need to go on text2img type something and then after i go back to img generation it works but why this happens?
ashleyk
ashleyk11mo ago
Which Stable Diffusion template, there is more than one?
Raios
RaiosOP11mo ago
runpod/stable-diffusion:web-ui-10.2.1
Want results from more Discord servers?
Add your server