File size: 1,215 Bytes
fbd604a 5263caf fbd604a 12d8157 fbd604a 5263caf 3c022b7 fbd604a 5263caf fbd604a 5263caf fbd604a 5263caf 8c4c24f 7ffaab0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 |
import torch
import gradio as gr
from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler
from diffusers.utils import export_to_video
import os
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
def generate_video(prompt):
# load pipeline
pipe = DiffusionPipeline.from_pretrained("damo-vilab/text-to-video-ms-1.7b", torch_dtype=torch.float16,
variant="fp16")
pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
# optimize for GPU memory
pipe.enable_model_cpu_offload()
pipe.enable_vae_slicing()
# generate
video_frames = pipe(prompt, num_inference_steps=25, num_frames=200).frames
# get absolute path to current working directory
current_directory = os.getcwd()
# create directory to store video
video_directory = os.path.join(current_directory, "generated_videos")
os.makedirs(video_directory, exist_ok=True)
# convert to video
video_path = export_to_video(video_frames, os.path.join(video_directory, "generated_video.mp4"))
return video_path
demo = gr.Interface(fn=generate_video, inputs="text", outputs="video")
demo.launch(share=True)
|