How was this made?
#1
by
HeyGoodEnough
- opened
I am wondering how to train such lora?
How many image pairs are required?
we did it with 100 pairs
Not working for me. Creating real broccoli on a head. Trigger word obviously used.
Where are you testing it? It might not work on comfy
Yes in latest ComfyUI. Could be be converted for Comfy?
It is not working on ComfyUI because at that time when this lora made we did not have ComfyUI support on our Kontext Trainer. It wont work on Comfy.
I see, Thank you. All fal and your loras won't work or just this particular one?
This one and some old ones. New ones generally work on Comfy because we have an option to choose Comfy compatible loras now.