Error with RX 580 8Gb

[WhiteAssassins]

I have this error "No Generated Images"

[RafaelEstevamReis]

Can you append the logs?

1. You can access the longs in the logs icon (second last)

2.Or in your installation folder, then: Data > logs > [date time of the error/session]

Inside that folder, you can find "general.txt" and "sd.txt" These are the logs

[WhiteAssassins]

general.txt sd.txt

[RafaelEstevamReis]

At [00000406] [02-08-2023 07:57:45] ... AssertionError: Torch not compiled with CUDA enabled - Implementation: OptimizedSd - Force Kill: False the logs says you used OptimizedSD

This implementation is only for CUDA

On configuration, you should go to Settings and change to ONNX This instructions are on the AMD Guide

Try to redo all steps on the guide, and don't forget the model conversion

Each of these implementation require a different model, booth StableDiffusion for CUDA uses checkpoints Stable Diffusion (ONNX - DirectML - For AMD GPUs) uses ONNX files

[WhiteAssassins]

ONNX Implementetion doents work

El vie, 10 feb 2023, 7:17 a. m., Rafael Estevam @.***> escribió:

At [00000406] [02-08-2023 07:55:04] the logs says you used StableDiffusion - OptimizedSD - CUDA

This implementation is only for CUDA

On configuration, you should go to Settings and change to ONNX This instructions are on the AMD Guide https://github.com/n00mkrad/text2image-gui/blob/main/docs/Amd.md

Try to redo all steps on the guide, and don't forget the model conversion

Each of these implementation require a different model, booth StableDiffusion for CUDA uses checkpoints Stable Diffusion (ONNX - DirectML - For AMD GPUs) uses ONNX files

— Reply to this email directly, view it on GitHub https://github.com/n00mkrad/text2image-gui/issues/80#issuecomment-1425732744, or unsubscribe https://github.com/notifications/unsubscribe-auth/AG47RXZCKFMGKYO2AZ6LCYDWWYWWHANCNFSM6AAAAAAUVG7EJM . You are receiving this because you authored the thread.Message ID: @.***>

[n00mkrad]

Need logs.