-
Notifications
You must be signed in to change notification settings - Fork 6.4k
Description
Describe the bug
I use diffuser to generate the image with dreamshaper_lightning model, but the result is really strange.
Model url:https://civitai.com/models/112902?modelVersionId=354657
Reproduction
`import torch
from diffusers import StableDiffusionXLPipeline, EulerDiscreteScheduler, DPMSolverMultistepScheduler
from transformers import CLIPTextModel
base_model_path = "dreamshaperXL_lightningDPMSDE.safetensors"
pipe = StableDiffusionXLPipeline.from_single_file(base_model_path,torch_dtype=torch.float16, variant="fp16")
pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config, use_karras_sigmas=True, algorithm_type="dpmsolver++")
pipe = pipe.to("cuda")
prompt = "a girl"
negative_prompt = "NSFW,"
generator = torch.Generator(torch.device("cuda")).manual_seed(-1)
num_inference_steps = 4
guidance_scale = 2
image = pipe(prompt=prompt,
negative_prompt=negative_prompt,
width=512,
height=512,
num_inference_steps=num_inference_steps,
guidance_scale=guidance_scale).images[0]
image.save("a girl.png")
`
Logs
No response
System Info
diffusers
version: 0.24.0- Platform: Linux-4.14.0_1-0-0-45-x86_64-with-glibc2.18
- Python version: 3.10.13
- PyTorch version (GPU?): 2.0.1+cu117 (True)
- Huggingface_hub version: 0.20.1
- Transformers version: 4.30.2
- Accelerate version: 0.21.0
- xFormers version: 0.0.20