This Lora is using a brand new unique way of creating Lora's.
This result in a better contrast, likeness, flexibility and morphology while being way smaller in size than my traditional Lora training.
Use a Lora strength between 0.85 to 1.0 depending of the checkpoint but for SDXL it's mostly 1.
This Lora doesn't do well with other checkpoints. Use with the original SDXL checkpoint.
Disclaimer: This artificial intelligence model ("the Model") is developed solely for artistic expression and technological demonstration, not for creating or distributing illicit or inappropriate content. The creator expressly disapproves of such use. By downloading or using the Model, you ("the User") agree not to misuse it in any manner contrary to these intentions and accept all responsibility for compliance with these terms.
Description
Initial Release
FAQ
Comments (5)
WOWW THANK U SO MUCHHH SO EXCITED TO TRY THIS LATER, WAS NOT EXPECTING THIS ESPECIALLY SO SOON SO I DONT HAVE MUCH TO TIP BUT MARK MY WORDS I WILL TIP EVERDAY THANK U
Iman Vellani and Chloe Grace Moretz please 🥺
hey man nice its a dreambooth extraction?
Getting this error when I try to import to magespace - can this be updated for?
Error message: Cannot load because text_model.encoder.layers.0.self_attn.q_proj.lora_linear_layer.up.weight expected shape tensor(..., device='meta', size=(768, 41), dtype=torch.float16), but got torch.Size([768, 56]). If you want to instead overwrite randomly initialized weights, please make sure to pass both `low_cpu_mem_usage=False` and `ignore_mismatched_sizes=True`. For more information, see also: https://github.com/huggingface/diffusers/issues/1619#issuecomment-1345604389 as an example.
How did you train this? What were your parameters? Best LORA on here