>>11337900
>but what gpu do you use?
A 3090, but this toml uses like 7gb vram so it feels kinda pointless lol, guess I need to get into finetuning and locon extraction now
>How long was that lora bake?
Don't remember exactly, this lora is epoch 2/15 and the metadata says it took 6min to bake. Though so far I haven't had any dataset take more than 7 epochs to converge