cagliostrolab

/

animagine-xl-3.0

like 720

I hope this message finds you well. First and foremost, I want to express my gratitude for creating such an amazing anime-style image generation model! Animagine is undoubtedly a top-tier model.

Now, I have some inquiries regarding the usage of Diffusers in conjunction with LCM and LoRA. The key points of my questions are as follows:

1. Can LCM (https://huggingface.co/Linaqruf/lcm-lora-sdxl-rank1) be used with version 3.0 of Animagine?
2. After loading the model, is it possible to toggle the activation of LCM on and off each time the generation function is executed (as shown below)?

#Load Model
pipe = ...(model)

def generate(prompt, n_prompt, use_lcm)
    if use_lcm == True:
        #Enable LCM
    image = pipe(prompt, negative_prompt=n_prompt...) #Generate Image
    if use_lcm == True:
        #Disable LCM
    return image

3. If I want to enable both LCM and LoRA (https://huggingface.co/collections/Linaqruf/animagine-xl-lora-656323c28fb38d71f76be91e) for generation, are there any specific considerations to keep in mind?
4. What code is required to use the Upscaler as seen in your demo?

Thank you for your time and for creating such an exceptional model.