![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
A couple of days after trying to troubleshoot AI-Toolkit in order to train my own Illustrious models an "update" for the app popped up and I remember thinking, "No fucking way." As it turns out, yes way. It now launches and runs flawlessly. As an added bonus, it runs far quicker than I anticipated at ~1.4s/iteration. The last time I trained LoRAs locally was SD 1.5 on the RTX3070 which took something like 12-hours. The RTX4090 is chewing through larger SDXL trainings in a little over an hour.
