Instructions to use minimaxir/sdxl-wrong-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use minimaxir/sdxl-wrong-lora with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("minimaxir/sdxl-wrong-lora") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
ValueError: Checkpoint not supported
#4
by foolmoron - opened
Getting this error on the latest colab. Looks like the layer string is lora_unet_down_blocks_1_attentions_0_transformer_blocks_0_attn1_to_k.lora_down.weight but the loader code doesn't have a case for down_blocks so it fails.
https://github.com/huggingface/diffusers/blob/main/src/diffusers/loaders.py#L1098
Any ideas?
It might be grabbing the safetensors file by default instead of the bin?
Most likely yes.
I removed it for now just in case, until there is further testing (might have to wait until another diffusers release which has more support for safetensors)
Sounds good. I also found you can pass weight_name="pytorch_lora_weights.bin" to load_lora_weights
foolmoron changed discussion status to closed