Trained on Js2Prony: https://civarchive.com/models/280321?modelVersionId=315675
This is a LyCORIS. It can be used just like a LoRA: <lora:name:1>
Description
FAQ
Comments (10)
looks good
You gave up your 870.3MB XL lora. Why?
I have switched from LoRA to LyCORIS. Training a LyCORIS requires more VRAM. I have to reduce rank.
@silvermoong Oh well, I didn't realise it was LyCORIS as it was enabled in the same way.
I think you remembered that you were going to make a 1.5 freng lora, I'm not trying to rush you, it's just that I see that it's been two months since you last made a 1.5 lora, and I'm worried that the more time that passes, the more rusty you're going to get with the 1.5 lora process. -- translation software
@silvermoong Of course I'm not familiar with lora production matters, so if I'm wrong, just wait for your list's turn for 1.5 freng as originally planned . Finally, I think XL is a real pain in the arse, and if a few different spec checkpoints come along later that are as good or even better than Animagine XL or pony XL, everyone will have to rebuild the artist lora a few more times. -- translation software
@aertt6950 here you go: https://civitai.com/models/331318?modelVersionId=371191
I used a new set parameters (V8) this time. V7.3 requires resizing images which needs lots of works. This is the main reason I'm slow on making 1.5 LoRAs.
@silvermoong Thank you very much! I didn't realise that even with a 1.5 lora you could make it extra-large :450.3MB .......
If memory serves me right, this is the first time I've ever downloaded a 1.5 Artist lora with more than 150+MB. Thankfully I got a new hard drive last week -- translation software
Oh, sorry, another mistake. It's not lora, it's lycoris.
Hey, I get a message saying <lora:DPOXL:1> not found. Is there a specific Lora i need to get when I use this prompt?
DPO is not required.
Here is the one that I use: https://civitai.com/models/237775?modelVersionId=269354
@silvermoong Thanks!