Missing easy way to set various renderers/samplers

[paciox]

Describe the bug

(https://www.reddit.com/r/StableDiffusion/comments/wwm2at/sampler_vs_steps_comparison_low_to_mid_step_counts/?utm_medium=android_app&utm_source=share)

Check here, now

Now I launch it like this:
import sys
import torch
from torch import autocast
from diffusers import StableDiffusionPipeline

model_id = "CompVis/stable-diffusion-v1-4"
device = "cuda"


pipe = StableDiffusionPipeline.from_pretrained(model_id, use_auth_token=True)
pipe = pipe.to(device)

image1 = pipe("SOME PROPMT", num_inference_steps=50, width=512, height=512)["sample"][0] 
  image1.save(f"aa.png")

NOw what if I want to set the others?

Reproduction

No response

Logs

No response

System Info

Windows 10, SD with 1.4 weights

[anton-l]

Hi @paciox! The recommended way to set a custom/alternative sampler for a pipeline is the following:

Load the scheduler's hyperparameters from another scheduler's config. Here we're loading the default scheduler's (PNDM's) params like beta_start, beta_end into LMS:

from diffusers import StableDiffusionPipeline, LMSDiscreteScheduler

model_id = "CompVis/stable-diffusion-v1-4"
scheduler = LMSDiscreteScheduler(model_id, subfolder="scheduler", use_auth_token=True)
pipe = StableDiffusionPipeline.from_pretrained(model_id, scheduler=scheduler, use_auth_token=True)
pipe = pipe.to("cuda")

[patrickvonplaten]

This is also related to: https://github.com/huggingface/diffusers/issues/336

[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.