Update app.py

app.py

@@ -2,7 +2,6 @@ import os
 import gradio as gr
 import openai
 import speech_recognition as sr
-import time
 
 # Set OpenAI API key
 openai.api_key = os.environ.get("OPENAI_API_KEY")
@@ -38,27 +37,32 @@ def transcribe_audio_realtime(audio):
     except sr.RequestError:
         return "Could not request results from the speech recognition service"
 
-def practice_pronunciation(audio):
-    original_text = generate_text()
+def practice_pronunciation(audio, text_to_read):
+    if not text_to_read:
+        text_to_read = generate_text()
     transcription = transcribe_audio_realtime(audio)
-    feedback = get_pronunciation_feedback(original_text, transcription)
-    return original_text, transcription, feedback
+    feedback = get_pronunciation_feedback(text_to_read, transcription)
+    return text_to_read, transcription, feedback
 
 # Gradio interface
-demo = gr.Interface(
-    fn=practice_pronunciation,
-    inputs=[
-        gr.Audio(type="filepath")
-    ],
-    outputs=[
-        gr.Textbox(label="Text to Read"),
-        gr.Textbox(label="Your Transcription"),
-        gr.Textbox(label="Pronunciation Feedback")
-    ],
-    title="Pronunciation Practice Tool",
-    description="Read the generated text aloud. The system will transcribe your speech and provide pronunciation feedback.",
-    live=True
-)
+with gr.Blocks() as demo:
+    gr.Markdown("# Pronunciation Practice Tool")
+    gr.Markdown("Generate a text to read, then record yourself reading it. The system will provide pronunciation feedback.")
+    
+    with gr.Row():
+        text_to_read = gr.Textbox(label="Text to Read")
+        generate_button = gr.Button("Generate New Text")
+    
+    audio_input = gr.Audio(type="filepath", label="Record your voice")
+    
+    with gr.Row():
+        transcription_output = gr.Textbox(label="Your Transcription")
+        feedback_output = gr.Textbox(label="Pronunciation Feedback")
+    
+    submit_button = gr.Button("Submit")
+    
+    generate_button.click(generate_text, outputs=text_to_read)
+    submit_button.click(practice_pronunciation, inputs=[audio_input, text_to_read], outputs=[text_to_read, transcription_output, feedback_output])
 
 # Launch the app
 if __name__ == "__main__":