Update app.py

app.py

@@ -1,20 +1,17 @@
 import gradio as gr
 from transformers import pipeline
 from gtts import gTTS
-from moviepy.editor import CompositeVideoClip, ImageClip, AudioFileClip, VideoFileClip, concatenate_videoclips
+from moviepy.editor import CompositeVideoClip, ImageClip, AudioFileClip, concatenate_videoclips
 from diffusers import StableDiffusionPipeline
 import torch
-from PIL import Image
-import numpy as np
-from scipy.io.wavfile import write
+from PIL import Image, ImageDraw, ImageFont
+import scipy.io.wavfile
 import random
 
 # Load and Initialize Models
 script_generator = pipeline("text-generation", model="gpt2", truncation=True, max_length=100)
 image_generator = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", torch_dtype=torch.float16).to("cpu")
-# For example, use a specific available model if MusicGen isn't accessible
-music_generator = pipeline("text-to-audio", model="facebook/musicgen-melody", device="cpu")
-
+music_generator = pipeline("text-to-audio", model="facebook/musicgen-small", device="cpu")
 
 
 # Generate Comedy Script
@@ -29,7 +26,6 @@ def text_to_speech(script):
     tts.save(audio_file)
     return audio_file
 
-
 # Create Images Using Stable Diffusion
 def create_images_from_script(script):
     lines = script.split('. ')
@@ -41,13 +37,18 @@ def create_images_from_script(script):
         image_paths.append(img_path)
     return image_paths
 
-# Generate Fun Music Track
+# Update: Generate Fun Music Track using MusicGen
 def generate_fun_music(prompt, output_music_file="fun_music.wav"):
-    # Generate music based on the prompt
-    music = music_generator(prompt)[0]['generated_music']
+    # Generate music based on the prompt using MusicGen
+    response = music_generator(prompt)
+    
+    # Extract audio and sampling rate from the response
+    audio_data = response["audio"]
+    sampling_rate = response["sampling_rate"]
+    
     # Save the generated music to a file
-    with open(output_music_file, 'wb') as f:
-        f.write(music)
+    scipy.io.wavfile.write(output_music_file, rate=sampling_rate, data=audio_data)
+    
     return output_music_file
 
 # Create Video from Generated Images
@@ -91,6 +92,7 @@ def generate_kids_animation_with_music(theme, output_video_file="kids_animation.
         frame_path = f'/tmp/kids_temp_{i}.png'
         img.save(frame_path)
         clips.append(ImageClip(frame_path).set_duration(1).set_position(('center', 'center')))
+    
     final_video = CompositeVideoClip(clips, size=(800, 400))
     final_video = final_video.set_audio(AudioFileClip(music_file))
     final_video.write_videofile(output_video_file, fps=24)
@@ -138,3 +140,4 @@ app.launch()
 
 
 
+