[LCM-LoRA] unexpected keys while loading LCM LoRA with `peft`

[sayakpaul]

Code to reproduce:

from diffusers import LCMScheduler, StableDiffusionXLPipeline
import torch

pipeline = StableDiffusionXLPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16).to("cuda")
pipeline.load_lora_weights("latent-consistency/lcm-lora-sdxl")
pipeline.fuse_lora()
pipeline.scheduler = LCMScheduler.from_config(pipeline.scheduler.config)

_ = pipe(
    prompt="a dog",
    num_inference_steps=4,
    guidance_scale=1.0,
)

Unexpected keys can be found here: https://github.com/huggingface/diffusers/actions/runs/8105285148/job/22153378360#step:8:45.

Cc: @younesbelkada @pacman100

[younesbelkada]

thanks for reproting @sayakpaul what are the unexpected keys you get?

[sayakpaul]

@younesbelkada sorry. Just updated the PR description.

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[sayakpaul]

@younesbelkada a gentle ping here.

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.