cuda out of memory

[learnuser1]

How do I set up to make the GPU memory required for inference smaller?

[AMujtaba57]

@learnuser1 , A single GPU with 16GB VRAM is sufficient for inference.

[iaekyx]

@learnuser1 , A single GPU with 16GB VRAM is sufficient for inference.

really? I tried A single GPU with 16GB VRAM but out of memory😭