❤️ Please Consider Supporting My Work
I love building open tools and LoRA models for everyone. If you find this useful, consider supporting me with a coffee. It helps me create even better and more models!
👉 https://ko-fi.com/vizsumit
Tips:
Lora Weight: 0.65-0.9
check images for prompt
avoid making your prompt too complex
Description
FAQ
Comments (8)
Great vector graphics tool!
V2 works like charm, quality is also better now
I'm trying to load this lora weights with python diffusion library but I'm facing errors, can you share the loading code. @vizsumit
this was compressed using resize_lora option maybe thats why it is happening
@vizsumit Thanks for the great work! I also have issues. It's CLIP:
e.g.
size mismatch for text_model.encoder.layers.11.self_attn.q_proj.lora_linear_layer.up.weight: copying a param with shape torch.Size([768, 16]) from checkpoint, the shape in current model is torch.Size([768, 23]).
Do you have any insight on how to fix this?
I am using:
model_id = "sd-legacy/stable-diffusion-v1-5"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")
pipe.load_lora_weights("Vector_illustration_V2.safetensors")
@vizsumit Hello, thanks for sharing the LoRA. I am also encountering an issue when loading it. You mentioned it's probably due to compression. How can we download the non-compressed version?
not work with diffusers library , i give errors , please help
Check a video with the steps to use the Vectorial Illustration model in Fooocus.