Update app.py

app.py

@@ -1,45 +1,46 @@
 import gradio as gr
 import speech_recognition as sr
+import torch
 from transformers import pipeline
 
-# Initialize the Hugging Face pipeline for speech-to-text
-speech_to_text = pipeline("automatic-speech-recognition", model="openai/whisper-large")
-
-# Function to process user input (name and email)
-def capture_user_info(audio_input):
-    # Convert speech to text using Hugging Face's ASR model
-    name = speech_to_text(audio_input)["text"]
-    
-    # Now, ask for the email
-    email_prompt = "Please provide your email address."
-    
-    # Return the name and email prompt
-    return name, email_prompt
-
-# Function to capture email input after the name
-def capture_email(audio_input):
-    # Convert speech to text for the email
-    email = speech_to_text(audio_input)["text"]
-    return email
-
-# Define Gradio Interface
+# Load ASR model (Whisper)
+device = "cuda" if torch.cuda.is_available() else "cpu"
+speech_to_text = pipeline("automatic-speech-recognition", model="openai/whisper-base", device=0 if device == "cuda" else -1)
+
+# Initialize Speech Recognition
+recognizer = sr.Recognizer()
+
+# Function to Capture Name
+def capture_name(audio):
+    try:
+        text = speech_to_text(audio)["text"]
+        return f"Name Captured: {text}", "Please provide your email address."
+    except Exception as e:
+        return f"Error: {str(e)}", ""
+
+# Function to Capture Email
+def capture_email(audio):
+    try:
+        text = speech_to_text(audio)["text"]
+        return f"Email Captured: {text}"
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+# Gradio Interface
 def gradio_interface():
     with gr.Blocks() as demo:
-        gr.Markdown("### Welcome to Biryani Hub")
-        
+        gr.Markdown("### 🎙️ Welcome to Biryani Hub")
+
         with gr.Column():
             gr.Markdown("#### Step 1: Tell me your name")
-            audio_input_name = gr.Audio(source="microphone", type="filepath", label="Speak your name")
+            audio_input_name = gr.Audio(type="filepath", label="Record your Name")
             name_output = gr.Textbox(label="Your Name:")
-            
-            # Step 1: Capture Name
-            audio_input_name.change(capture_user_info, inputs=audio_input_name, outputs=[name_output, email_prompt_output])
-            
-            gr.Markdown("#### Step 2: Please provide your email address")
-            audio_input_email = gr.Audio(source="microphone", type="filepath", label="Speak your email")
+            email_prompt_output = gr.Textbox(label="Next Step:", interactive=False)
+            audio_input_name.change(capture_name, inputs=audio_input_name, outputs=[name_output, email_prompt_output])
+
+            gr.Markdown("#### Step 2: Provide your email")
+            audio_input_email = gr.Audio(type="filepath", label="Record your Email")
             email_output = gr.Textbox(label="Your Email:")
-            
-            # Step 2: Capture Email
             audio_input_email.change(capture_email, inputs=audio_input_email, outputs=email_output)
 
     return demo