import gradio as gr import torch import os import spaces import uuid from diffusers import AnimateDiffPipeline, MotionAdapter, EulerDiscreteScheduler from diffusers.utils import export_to_video from huggingface_hub import hf_hub_download from safetensors.torch import load_file from PIL import Image # Constants bases = { "ToonYou": "frankjoshua/toonyou_beta6", "epiCRealism": "emilianJR/epiCRealism" } step_loaded = None base_loaded = "ToonYou" motion_loaded = None # Ensure model and scheduler are initialized in GPU-enabled function if not torch.cuda.is_available(): raise NotImplementedError("No GPU detected!") device = "cuda" dtype = torch.float16 pipe = AnimateDiffPipeline.from_pretrained(bases[base_loaded], torch_dtype=dtype).to(device) pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing", beta_schedule="linear") # Function @spaces.GPU(enable_queue=True) def generate_image(prompt, base, motion, step): global step_loaded global base_loaded print(prompt, base, step) if step_loaded != step: repo = "ByteDance/AnimateDiff-Lightning" ckpt = f"animatediff_lightning_{step}step_diffusers.safetensors" pipe.unet.load_state_dict(load_file(hf_hub_download(repo, ckpt), device=device), strict=False) step_loaded = step if base_loaded != base: pipe.unet.load_state_dict(torch.load(hf_hub_download(bases[base], "unet/diffusion_pytorch_model.bin"), map_location=device), strict=False) base_loaded = base if motion_loaded != motion: pipe.unload_lora_weights() pipe.load_lora_weights(hf_hub_download("guoyww/animatediff", motion), adapter_name="motion") pipe.set_adapters(["motion"], [0.7]) motion_loaded = motion output = pipe(prompt=prompt, guidance_scale=1.0, num_inference_steps=step) name = str(uuid.uuid4()).replace("-", "") path = f"/tmp/{name}.mp4" export_to_video(output.frames[0], path, fps=10) return path # Gradio Interface with gr.Blocks(css="style.css") as demo: gr.HTML("

AnimateDiff-Lightning ⚡

") gr.HTML("

Lightning-fast text-to-video generation

https://huggingface.co/ByteDance/AnimateDiff-Lightning

") with gr.Group(): with gr.Row(): prompt = gr.Textbox( label='Prompt (English)' ) with gr.Row(): select_base = gr.Dropdown( label='Base model', choices=[ "ToonYou", "epiCRealism", ], value=base_loaded, interactive=True ) select_motion = gr.Dropdown( label='Motion LoRAs', choices=[ ("None", None), ("Zoom in", "v2_lora_ZoomIn.ckpt"), ("Zoom out", "v2_lora_ZoomOut.ckpt"), ], value=None, interactive=True ) select_step = gr.Dropdown( label='Inference steps', choices=[ ('1-Step', 1), ('2-Step', 2), ('4-Step', 4), ('8-Step', 8)], value=4, interactive=True ) submit = gr.Button( scale=1, variant='primary' ) video = gr.Video( label='AnimateDiff-Lightning', autoplay=True, height=512, width=512, elem_id="video_output" ) prompt.submit( fn=generate_image, inputs=[prompt, select_base, select_motion, select_step], outputs=video, ) submit.click( fn=generate_image, inputs=[prompt, select_base, select_motion, select_step], outputs=video, ) demo.queue().launch()