add instrument as inputs

app.py

@@ -1,3 +1,4 @@
+import typing
 from pathlib import Path
 
 import numpy as np
@@ -24,7 +25,7 @@ class AudioPalette:
             "Slow": "low"
         }
     
-    def generate(self, input_image: PIL.Image.Image, ngrok_endpoint: str):
+    def generate(self, input_image: PIL.Image.Image, instrument: typing.Union[str, None], ngrok_endpoint: str):
         pace = self.pace_model.predict(input_image)
         print("Pace Prediction Done")
 
@@ -33,7 +34,13 @@ class AudioPalette:
 
         generated_text = generated_text if generated_text is not None else ""
         # prompt = f"Generate a soundtrack for {generated_text} with {pace} beats and the instrument of choice is the guitar, High quality"
-        prompt = f"A soundtrack for {generated_text} with {self.pace_map[pace]} beats per minute. High quality."
+
+        if instrument is None:
+            prompt = f"A soundtrack for {generated_text} with {self.pace_map[pace]} beats per minute. High quality."
+        else:
+            prompt = f"A soundtrack for {generated_text} with {self.pace_map[pace]} beats per minute with a {instrument}. High quality."
+        
+        print("Generated Prompt:", prompt)
 
         audio_file = self.audio_generation.generate(prompt, ngrok_endpoint)
         print("Audio Generation Done")
@@ -54,6 +61,12 @@ def main():
                 show_label=True,
                 container=True
             ),
+            gr.Radio(
+                choices=["Piano", "Drums", "Guitar", "Violin", "Flute"],
+                label="Instrument",
+                show_label=True,
+                container=True
+            ),
             gr.Textbox(
                 lines=1,
                 placeholder="ngrok endpoint",
@@ -72,7 +85,7 @@ def main():
                 show_label=True,
                 container=True,
                 type="text",
-                visible=True
+                visible=False
             ),
             gr.Textbox(
                 lines=1,