Update app.py

app.py

@@ -7,7 +7,6 @@ from transformers import MusicgenForConditionalGeneration, AutoProcessor
 from scipy.io import wavfile
 import ffmpeg
 
-# Function to generate video frames
 def generate_video(image, prompt, negative_prompt, video_length):
     generator = torch.manual_seed(8888)
     device = torch.device("mps" if torch.backends.mps.is_available() else "cpu")
@@ -34,18 +33,18 @@ def generate_video(image, prompt, negative_prompt, video_length):
 
     return frames
 
-# Function to export frames to video
 def export_frames_to_video(frames, output_file):
     frames_np = [np.array(frame) for frame in frames]
     clip = ImageSequenceClip(frames_np, fps=30)
     clip.write_videofile(output_file, codec='libx264', audio=False)
 
-# Function to generate music
 def generate_music(prompt, unconditional=False):
     model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")
     device = "cuda:0" if torch.cuda.is_available() else "cpu"
     model.to(device)
 
+    # Simulate progress for music generation
+    st.progress(0)  # Initialize progress bar
     if unconditional:
         unconditional_inputs = model.get_unconditional_inputs(num_samples=1)
         audio_values = model.generate(**unconditional_inputs, do_sample=True, max_new_tokens=256)
@@ -56,12 +55,14 @@ def generate_music(prompt, unconditional=False):
             padding=True,
             return_tensors="pt",
         )
-        audio_values = model.generate(**inputs.to(device), do_sample=True, guidance_scale=3, max_new_tokens=256)
+        # Simulate progress by updating the progress bar
+        for i in range(1, 6):  # Assuming 5 steps for demonstration
+            audio_values = model.generate(**inputs.to(device), do_sample=True, guidance_scale=3, max_new_tokens=256)
+            st.progress(i / 5)  # Update progress bar
 
     sampling_rate = model.config.audio_encoder.sampling_rate
     return audio_values[0].cpu().numpy(), sampling_rate
 
-# Function to combine audio and video
 def combine_audio_video(audio_file, video_file, output_file):
     audio = ffmpeg.input(audio_file)
     video = ffmpeg.input(video_file)
@@ -89,8 +90,8 @@ if st.sidebar.button("Generate Video and Music"):
         
         # Video generation with progress bar
         st.write("Generating video...")
-        frames = generate_video(image, prompt, negative_prompt, video_length)
-        export_frames_to_video(frames, "output_video.mp4")
+        video_frames = generate_video(image, prompt, negative_prompt, video_length)
+        export_frames_to_video(video_frames, "output_video.mp4")
         st.video("output_video.mp4")
 
     # Music generation with progress bar