Does the "Paste Prompt" button change the GPU Weights?

[toyssamurai]

I am not 100% sure this is the cause of my problem, but checking my log, I could only come to the conclusion that using the "Paste Prompt" button (the small angled down arrow under the "Generate" button) changes the GPU weights as well:

Environment vars changed: {'stream': False, 'inference_memory': 8676.0, 'pin_shared_memory': False}
[GPU Setting] You will use 82.18% GPU memory (40000.00 MB) to load weights, and use 17.82% GPU memory (8676.00 MB) to do matrix computation.

Above is what I had in my log before I pressed that button. After I clicked on it, I got:

Environment vars changed: {'stream': False, 'inference_memory': -31528.0, 'pin_shared_memory': False}
------------------
[Low VRAM Warning] You just set Forge to use 100% GPU memory (80204.00 MB) to load model weights.
[Low VRAM Warning] This means you will have 0% GPU memory (0.00 MB) to do matrix computation. Computations may fallback to CPU or go Out of Memory.
[Low VRAM Warning] In many cases, image generation will be 10x slower.
[Low VRAM Warning] To solve the problem, you can set the 'GPU Weights' (on the top of page) to a lower value.
[Low VRAM Warning] If you cannot find 'GPU Weights', you can click the 'all' option in the 'UI' area on the left-top corner of the webpage.
[Low VRAM Warning] Make sure that you know what you are testing.
------------------

FYI, I used a cloud GPU, and once I restarted, the GPU changed from a 80Gb A100 to a 48Gb A6000. The prompt I was trying to recall from the button came from the session BEFORE the restart.

Immediately after I saw that Low VRAM warning, I got the connection was lost error from the web UI, which kind of makes sense if I ran out of VRAM on the backend.