Update app.py

app.py

@@ -1,43 +1,47 @@
 from transformers import AutoProcessor, MusicgenForConditionalGeneration
+from IPython.display import Audio
+import scipy
 import torch
 import streamlit as st
 
-def mu_gen(prompt):
-    processor = AutoProcessor.from_pretrained("facebook/musicgen-small")
-    model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")
 
-    device = torch.device("cpu")
-    model.to(device)
+def mu_gen(prompt):
+  processor = AutoProcessor.from_pretrained("facebook/musicgen-small")
+  model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")
 
-    inputs = processor(
-        text=[str(prompt)],
-        padding=True,
-        return_tensors="pt",
-    )
+  device = torch.device("cpu")
+  model.to(device)
 
-    inputs = {key: value.to(device) for key, value in inputs.items()}
+  inputs = processor(
+      text = [str(prompt)],     # This line is correct
+      padding=True,
+      return_tensors="pt",
+  )
+  
+  inputs = {key: value.to(device) for key, value in inputs.items()}
 
     # Generate audio on CPU
-    audio_values = model.generate(**inputs, max_new_tokens=256)
-    sampling_rate = model.config.audio_encoder.sampling_rate
+  audio_values = model.generate(**inputs, max_new_tokens=256)
+  sampling_rate = model.config.audio_encoder.sampling_rate
+
+    # Create an Audio object from the generated audio
+  result = Audio(audio_values[0].numpy(), rate=sampling_rate)
+
+  return result
 
-    return audio_values, sampling_rate
 
 def main():
-    st.title("Text to Music Generator")
+    st.title("Text to music")
 
     # Input text prompt
-    prompt = st.text_input("Enter a text prompt", "")
+    title = st.text_input('Write a prompt (จะใช้เวลาค่อนข้างมากในการสร้างเนื่องจากใช้ CPU ในการรันโมเดล)', "")
 
-    if st.button("Generate Music"):
-        if prompt:
-            # Call the mu_gen function to generate music
-            generated_music, sampling_rate = mu_gen(prompt)
+    if st.button('Generate Image'):
+        # Call the pic_mo function to generate an image
+        generated_music = mu_gen(prompt)
 
-            # Display the generated audio
-            st.audio(generated_music[0].numpy(), format="audio/wav", sample_rate=sampling_rate)
-        else:
-            st.warning("Please enter a text prompt.")
+        # Display the generated image
+        st.image(generated_music, caption='Generated Music', use_column_width=True)
 
-if __name__ == "__main__":
-    main()
+if __name__ == '__main__':
+    main()