Out of memory when training

[bobboff]

I always get CUDA out of memory when trying to train a new subject in A1111 using the Train tab. My understanding was that my VRAM should be sufficient but apparently not. Are there any settings I can change or anything I can do to get it to work? Card is a GTX1660 Super 6gb

Have tried loading with --medvram and as normal but makes no difference.

Thanks

[enn-nafnlaus]

Restart the server. There's a memory leak in training. Best to restart it after each interrupt, though I can usually get away with every other interrupt.