search

pkuliyi2015 / multidiffusion-upscaler-for-automatic1111

Tiled Diffusion and VAE optimize, licensed under CC BY-NC-SA 4.0
Other
4.71k stars 333 forks source link

Upscale problem #333

Open sebaxakerhtc opened 10 months ago

sebaxakerhtc commented 10 months ago

Hello and thank you for this awesome project.

I have a problem with upscaling img2img with tiled diffusion.

A half a year ago I used it for upscaling up to 16K (maybe more, but I don't need it)

Now, when I try to upscale to 2K, [Tiled Diffusion] OK - about 70%-92% it gives me error at [Tiled VAE]: Executing Decoder Task Queue step:

RuntimeError: [enforce fail at ..\c10\core\impl\alloc_cpu.cpp:72] data. DefaultCPUAllocator: not enough memory: you tried to allocate 375160832 bytes.

Temporary moved to Ultimate SD Upscaler, but your extension much better. Please, any suggesions.

A1111:

Python 3.10.11 (tags/v3.10.11:7d4cc5a, Apr  5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)]
Version: v1.6.0-2-g4afaaf8a
Commit hash: 4afaaf8a020c1df457bcf7250cb1c7f609699fa7

GPU: RTX A4000 16GB VRAM RAM: 32GB DDR5

Full output: output.log

SafetyLucas commented 6 days ago

Ensure that your page file is correct for 32 GB of RAM