This forum is for reporting errors with the Training process. If you want to get tips, or better understand the Training process, then you should look in the Training Discussion forum.
Please mark any answers that fixed your problems so others can find the solutions.
After achieving some decent results following the tutorial to the letter, i tried my first train with Dlight. I tried with a batch-size of 12 and it crashed after about 3 hours due to not enough VRAM. I had no other program running.
After that, i tried with a batch-size of 10, in the same conditions, but the same thing is happening.
I am running the nvidia-smi and during the batch-size 10 training it shows a constant 2886 MiB out of 4096 MiB.
The amount of VRAM that faceswap requires will not change over time. Most likely something else (Google chrome/game/desktop etc) is stealing VRAM which is causing FS to fail.