Update app.py

app.py

@@ -1,11 +1,24 @@
 import torch
 import gradio as gr
-from diffusers import AnimateDiffPipeline, MotionAdapter, DPMSolverMultistepScheduler, AutoencoderKL, SparseControlNetModel, EulerAncestralDiscreteScheduler
+from diffusers import AnimateDiffPipeline, MotionAdapter, DPMSolverMultistepScheduler, AutoencoderKL, SparseControlNetModel
 from diffusers.utils import export_to_gif, load_image
+from transformers import pipeline
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
+# 한글-영어 번역 모델 로드
+translator = pipeline("translation", model="Helsinki-NLP/opus-mt-ko-en")
+
+def translate_korean_to_english(text):
+    if any('\u3131' <= char <= '\u3163' or '\uac00' <= char <= '\ud7a3' for char in text):
+        translated = translator(text)[0]['translation_text']
+        return translated
+    return text
+
 def generate_video(prompt, negative_prompt, num_inference_steps, conditioning_frame_indices, controlnet_conditioning_scale):
+    prompt = translate_korean_to_english(prompt)
+    negative_prompt = translate_korean_to_english(negative_prompt)
+
     motion_adapter = MotionAdapter.from_pretrained("guoyww/animatediff-motion-adapter-v1-5-3", torch_dtype=torch.float16).to(device)
     controlnet = SparseControlNetModel.from_pretrained("guoyww/animatediff-sparsectrl-scribble", torch_dtype=torch.float16).to(device)
     vae = AutoencoderKL.from_pretrained("stabilityai/sd-vae-ft-mse", torch_dtype=torch.float16).to(device)
@@ -43,52 +56,18 @@ def generate_video(prompt, negative_prompt, num_inference_steps, conditioning_fr
     export_to_gif(video, "output.gif")
     return "output.gif"
 
-def generate_simple_video(prompt):
-    adapter = MotionAdapter.from_pretrained("guoyww/animatediff-motion-adapter-v1-5-2", torch_dtype=torch.float16).to(device)
-    pipe = AnimateDiffPipeline.from_pretrained("SG161222/Realistic_Vision_V6.0_B1_noVAE", motion_adapter=adapter, torch_dtype=torch.float16).to(device)
-    pipe.scheduler = EulerAncestralDiscreteScheduler(
-        beta_schedule="linear",
-        beta_start=0.00085,
-        beta_end=0.012,
-    )
-    
-    pipe.enable_free_noise()
-    pipe.vae.enable_slicing()
-    pipe.enable_model_cpu_offload()
-
-    frames = pipe(
-        prompt,
-        num_frames=128,
-        num_inference_steps=100,
-        guidance_scale=15.0,
-        decode_chunk_size=1,
-    ).frames[0]
-    
-    export_to_gif(frames, "simple_output.gif")
-    return "simple_output.gif"
-
-demo1 = gr.Interface(
+demo = gr.Interface(
     fn=generate_video,
     inputs=[
-        gr.Textbox(label="Prompt", value="an aerial view of a cyberpunk city, night time, neon lights, masterpiece, high quality"),
-        gr.Textbox(label="Negative Prompt", value="low quality, worst quality, letterboxed"),
+        gr.Textbox(label="Prompt (한글 또는 영어)", value="사이버펑크 도시의 공중 전망, 밤, 네온 불빛, 걸작, 고품질"),
+        gr.Textbox(label="Negative Prompt (한글 또는 영어)", value="저품질, 최악의 품질, 레터박스"),
         gr.Slider(label="Number of Inference Steps", minimum=1, maximum=200, step=1, value=100),
         gr.Textbox(label="Conditioning Frame Indices", value="[0, 8, 15]"),
         gr.Slider(label="ControlNet Conditioning Scale", minimum=0.1, maximum=2.0, step=0.1, value=1.0)
     ],
     outputs=gr.Image(label="Generated Video"),
-    title="Generate Video with AnimateDiffSparseControlNetPipeline",
-    description="Generate a video using the AnimateDiffSparseControlNetPipeline."
+    title="AnimateDiffSparseControlNetPipeline을 사용한 비디오 생성",
+    description="AnimateDiffSparseControlNetPipeline을 사용하여 비디오를 생성합니다. 한글 또는 영어로 프롬프트를 입력할 수 있습니다."
 )
 
-demo2 = gr.Interface(
-    fn=generate_simple_video,
-    inputs=gr.Textbox(label="Prompt", value="An astronaut riding a horse on Mars."),
-    outputs=gr.Image(label="Generated Simple Video"),
-    title="Generate Simple Video with AnimateDiff",
-    description="Generate a simple video using the AnimateDiffPipeline."
-)
-
-demo = gr.TabbedInterface([demo1, demo2], ["Advanced Video Generation", "Simple Video Generation"])
-
 demo.launch()