Create app.py

app.py

@@ -0,0 +1,69 @@
+import speech_recognition as sr
+from pydub import AudioSegment
+import gradio as gr
+from os import path
+import requests
+import openai
+
+r = sr.Recognizer()
+
+def record_text(audio_file):
+    sound = audio_file
+    sound_type = sound.split(".")
+    if sound_type[-1] == 'mp3':
+        input_file = sound
+        output_file = "con_sound.wav"
+        try:
+            # convert mp3 file to wav file 
+            sound = AudioSegment.from_mp3(input_file) 
+            sound.export(output_file, format="wav")
+            sound = "con_sound.wav"
+        except Exception as e:
+            print(f"Error converting MP3 to WAV: {e}")
+            return ""
+    
+    MyText = ""
+    with sr.AudioFile(sound) as source:
+        r.adjust_for_ambient_noise(source)
+        print("Converting audio file to text..")
+        audio2 = r.record(source, duration=None)  # Use record instead of listen
+        try:
+            MyText = r.recognize_google(audio2, language="en-US", key=None, show_all=False)
+            MyText = MyText.lower()
+            print("Converted audio is:\n" + MyText + '.')
+        except sr.UnknownValueError:
+            print("Google Speech Recognition could not understand audio")
+    
+    return MyText
+
+def message_and_history(audio_file, history, api_key):
+    history = history or []
+    input_text = "Type and press Enter"
+    output_text = record_text(audio_file)
+    
+    if len(input_text) == 0:
+        input_text = "Apply proper punctuation on the given paragraph."
+        history.append((input_text, output_text))
+    else:
+        history.append((input_text, output_text))
+    
+    return history, history
+
+prompt = "Type and press Enter"
+block = gr.Blocks(theme=gr.themes.Glass(primary_hue="slate"))
+with block:
+    gr.Markdown("""<h1><center>Audio Recognition - Ask and Learn about an Audio</center></h1> """)
+    with gr.Row():
+        with gr.Column(scale=0.5):
+            aud_input = gr.Audio(type="filepath", label="Upload Audio")
+            api_input = gr.Textbox(label="Enter Api-key")
+            upload_button = gr.Button(value="Upload & Start Chat", interactive=True, variant="primary")
+        with gr.Column():
+            chatbot = gr.Chatbot(label="Ask questions about the audio")
+            message = gr.Textbox(label="User", placeholder=prompt)
+            state = gr.State()
+            
+    upload_button.click(message_and_history, inputs=[aud_input, state, api_input], outputs=[chatbot, state])
+    message.submit(message_and_history, inputs=[aud_input, state, api_input], outputs=[chatbot, state])
+    message.submit(lambda: None, None, message, queue=False)
+block.launch(share=True)