Update app.py

app.py

@@ -377,69 +377,24 @@ Sure! The Responses are as follows:
 """
 
 
-# phi_custom_template = """
-#  <|system|>
-# Fornisci risposte concise dal documento in questo formato:
-
-# Nome del documento: (nome)
-# Numero di pagina: (pagina)
-# Contenuto effettivo: (contenuto)
-# Alla fine, fornisci la risposta finale:
-
-# Risposta: (risposta)
-# Se non trovata, rispondi con "Questa domanda è oltre le mie conoscenze".
-#  <|end|>
-
-# <|user|>
-# {context}
-# Question: {question}<|end|>
-
-# <|assistant|>
-# Sure! The Responses are as follows:
-# """
-
-
-
-# def generate_bot_response(history, choice, retrieval_mode, model_choice):
-#     if not history:
-#         return
-
-#     # Select the model
-#     # selected_model = chat_model if model_choice == "LM-1" else phi_pipe
-#     selected_model = chat_model if model_choice == "LM-1" else (chat_model1 if model_choice == "LM-3" else phi_pipe)
-
-
-#     response, addresses = generate_answer(history[-1][0], choice, retrieval_mode, selected_model)
-#     history[-1][1] = ""
-
-#     for character in response:
-#         history[-1][1] += character
-#         yield history  # Stream each character as it is generated
-#         time.sleep(0.05)  # Add a slight delay to simulate streaming
-
-#     yield history  # Final yield with the complete response
-
-
-# Modified chatbot response function to also return the confidence plot
-def generate_bot_response(chatbot, choice, retrieval_mode, model_choice):
-    # Your chatbot logic to generate the response
-    response = f"This is a sample response based on your query."
-
-    # Get the confidence level (you can customize this for your actual confidence calculation)
-    confidence_level = get_confidence_level(response)
-
-    # Plot the confidence as a bar graph
-    confidence_plot = plot_confidence_bar(confidence_level)
-
-    # Return both the chatbot response and the confidence plot
-    return chatbot + [[None, response]], confidence_plot
-
-
+def generate_bot_response(history, choice, retrieval_mode, model_choice):
+    if not history:
+        return
 
+    # Select the model
+    # selected_model = chat_model if model_choice == "LM-1" else phi_pipe
+    selected_model = chat_model if model_choice == "LM-1" else (chat_model1 if model_choice == "LM-3" else phi_pipe)
 
 
+    response, addresses = generate_answer(history[-1][0], choice, retrieval_mode, selected_model)
+    history[-1][1] = ""
 
+    for character in response:
+        history[-1][1] += character
+        yield history  # Stream each character as it is generated
+        time.sleep(0.05)  # Add a slight delay to simulate streaming
 
+    yield history  # Final yield with the complete response
 
 
 def generate_tts_response(history, tts_choice):
@@ -1517,193 +1472,69 @@ def fetch_google_flights(departure_id="JFK", arrival_id="BHM", outbound_date=cur
 #     return prompt[0] if prompt else current_text
 
 
-#confidence Level 
-import numpy as np
-import matplotlib.pyplot as plt
-
-# Function to calculate the confidence level (you can modify this to use real confidence scores)
-def get_confidence_level(response):
-    # Simulate a confidence score between 0 and 1 for demonstration purposes
-    confidence_level = np.random.rand()
-    return confidence_level
-
-# Function to plot confidence score as a bar graph
-def plot_confidence_bar(confidence_level):
-    fig, ax = plt.subplots()
-
-    categories = ['Confidence Level']
-    values = [confidence_level]
-
-    # Create a bar chart
-    ax.bar(categories, values, color='blue')
-    ax.set_ylim(0, 1)  # Confidence level is between 0 and 1
-    ax.set_title('Chatbot Confidence Level')
-    ax.set_ylabel('Confidence Score')
-
-    # Save the plot as an image
-    temp_image_path = "confidence_plot.png"
-    plt.savefig(temp_image_path)
-    plt.close(fig)
-
-    return temp_image_path
-
-
-
-# with gr.Blocks(theme='gradio/soft') as demo:
-
-   
-#     with gr.Row():
-#         with gr.Column():
-#             state = gr.State()
-
-#             chatbot = gr.Chatbot([], elem_id="RADAR:Channel 94.1", bubble_full_width=False)
-#             choice = gr.Radio(label="Select Style", choices=["Details", "Conversational"], value="Conversational",interactive=False,visible=False)
-#             retrieval_mode = gr.Radio(label="Retrieval Mode", choices=["VDB", "KGF"], value="VDB",interactive=False,visible=False)
-#             model_choice = gr.Dropdown(label="Choose Model", choices=["LM-2"], value="LM-2")
-
-#             # Link the dropdown change to handle_model_choice_change
-#             model_choice.change(fn=handle_model_choice_change, inputs=model_choice, outputs=[retrieval_mode, choice, choice])
-
-#             # gr.Markdown("<h1 style='color: red;'>Talk to RADAR</h1>", elem_id="voice-markdown")
-
-#             chat_input = gr.Textbox(show_copy_button=True, interactive=True, show_label=False, label="ASK Radar !!!")
-#             tts_choice = gr.Radio(label="Select TTS System", choices=["Alpha", "Beta"], value="Alpha")
-        
-#             retriever_button = gr.Button("Retriever")
-
-#             clear_button = gr.Button("Clear")
-#             clear_button.click(lambda: [None, None], outputs=[chat_input, state])
-
-#             # gr.Markdown("<h1 style='color: red;'>Radar Map</h1>", elem_id="Map-Radar")
-#             # location_output = gr.HTML()
-#             audio_output = gr.Audio(interactive=False, autoplay=True)
-
-#             def stop_audio():
-#                 audio_output.stop()
-#                 return None
-
-               
-    
-           
-
-#             retriever_sequence = (
-#                 retriever_button.click(fn=stop_audio, inputs=[], outputs=[audio_output], api_name="api_stop_audio_recording")
-#                 .then(fn=add_message, inputs=[chatbot, chat_input], outputs=[chatbot, chat_input], api_name="api_addprompt_chathistory")
-#                 # First, generate the bot response
-#                 .then(fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot], api_name="api_generate_bot_response")
-#                 # Then, generate the TTS response based on the bot's response
-#                 .then(fn=generate_tts_response, inputs=[chatbot, tts_choice], outputs=[audio_output], api_name="api_generate_tts_response")
-#                 .then(fn=clear_textbox, inputs=[], outputs=[chat_input], api_name="api_clear_textbox")
-#             )
-
-
-
-
-
-
-            
 
-#             chat_input.submit(fn=stop_audio, inputs=[], outputs=[audio_output], api_name="api_stop_audio_recording").then(
-#                 fn=add_message, inputs=[chatbot, chat_input], outputs=[chatbot, chat_input], api_name="api_addprompt_chathistory"
-#             ).then(
-#                 # First, generate the bot response
-#                 fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot], api_name="api_generate_bot_response"
-#             ).then(
-#                 # Then, generate the TTS response based on the bot's response
-#                 fn=generate_tts_response, inputs=[chatbot, tts_choice], outputs=[audio_output], api_name="api_generate_tts_response"
-#             ).then(
-#                 fn=clear_textbox, inputs=[], outputs=[chat_input], api_name="api_clear_textbox"
-#             )
 
 
 
-
-
-
-
-#             audio_input = gr.Audio(sources=["microphone"], streaming=True, type='numpy', every=0.1)
-#             audio_input.stream(transcribe_function, inputs=[state, audio_input], outputs=[state, chat_input], api_name="api_voice_to_text")
-            
-#         #     gr.Markdown("<h1 style='color: red;'>Example Prompts</h1>", elem_id="Example-Prompts")
-#         #     gr.Examples(examples=examples, fn=insert_prompt,inputs=chat_input, outputs=chat_input)
-
-#         # with gr.Column():
-#         #     weather_output = gr.HTML(value=fetch_local_weather())
-#         #     news_output = gr.HTML(value=fetch_local_news())
-#         #     events_output = gr.HTML(value=fetch_local_events())
-
-#         # with gr.Column():
-           
-          
-#         #     # Call update_images during the initial load to display images when the interface appears
-#         #     initial_images = update_images()
-            
-#         #     # Displaying the images generated using Flux API directly
-#         #     image_output_1 = gr.Image(value=initial_images[0], label="Image 1", elem_id="flux_image_1", width=400, height=400)
-#         #     image_output_2 = gr.Image(value=initial_images[1], label="Image 2", elem_id="flux_image_2", width=400, height=400)
-#         #     image_output_3 = gr.Image(value=initial_images[2], label="Image 3", elem_id="flux_image_3", width=400, height=400)
-    
-#         #     # Refresh button to update images
-#         #     refresh_button = gr.Button("Refresh Images")
-#         #     refresh_button.click(fn=update_images, inputs=None, outputs=[image_output_1, image_output_2, image_output_3])
-            
-
-
-
-
-           
-
-# demo.queue()
-# demo.launch(show_error=True)
-
-
-
-# Rest of your Gradio code
 with gr.Blocks(theme='gradio/soft') as demo:
 
+   
     with gr.Row():
         with gr.Column():
             state = gr.State()
 
             chatbot = gr.Chatbot([], elem_id="RADAR:Channel 94.1", bubble_full_width=False)
-            choice = gr.Radio(label="Select Style", choices=["Details", "Conversational"], value="Conversational", interactive=False, visible=False)
-            retrieval_mode = gr.Radio(label="Retrieval Mode", choices=["VDB", "KGF"], value="VDB", interactive=False, visible=False)
+            choice = gr.Radio(label="Select Style", choices=["Details", "Conversational"], value="Conversational",interactive=False,visible=False)
+            retrieval_mode = gr.Radio(label="Retrieval Mode", choices=["VDB", "KGF"], value="VDB",interactive=False,visible=False)
             model_choice = gr.Dropdown(label="Choose Model", choices=["LM-2"], value="LM-2")
 
             # Link the dropdown change to handle_model_choice_change
             model_choice.change(fn=handle_model_choice_change, inputs=model_choice, outputs=[retrieval_mode, choice, choice])
 
+            # gr.Markdown("<h1 style='color: red;'>Talk to RADAR</h1>", elem_id="voice-markdown")
+
             chat_input = gr.Textbox(show_copy_button=True, interactive=True, show_label=False, label="ASK Radar !!!")
             tts_choice = gr.Radio(label="Select TTS System", choices=["Alpha", "Beta"], value="Alpha")
-            
+        
             retriever_button = gr.Button("Retriever")
+
             clear_button = gr.Button("Clear")
             clear_button.click(lambda: [None, None], outputs=[chat_input, state])
 
+            # gr.Markdown("<h1 style='color: red;'>Radar Map</h1>", elem_id="Map-Radar")
+            # location_output = gr.HTML()
             audio_output = gr.Audio(interactive=False, autoplay=True)
 
             def stop_audio():
                 audio_output.stop()
                 return None
 
-            # Add new output for the confidence bar graph
-            confidence_plot_output = gr.Image(label="Confidence Level Bar Graph")
+               
+    
+           
 
             retriever_sequence = (
                 retriever_button.click(fn=stop_audio, inputs=[], outputs=[audio_output], api_name="api_stop_audio_recording")
                 .then(fn=add_message, inputs=[chatbot, chat_input], outputs=[chatbot, chat_input], api_name="api_addprompt_chathistory")
-                # First, generate the bot response and confidence score
-                .then(fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot, confidence_plot_output], api_name="api_generate_bot_response")
+                # First, generate the bot response
+                .then(fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot], api_name="api_generate_bot_response")
                 # Then, generate the TTS response based on the bot's response
                 .then(fn=generate_tts_response, inputs=[chatbot, tts_choice], outputs=[audio_output], api_name="api_generate_tts_response")
                 .then(fn=clear_textbox, inputs=[], outputs=[chat_input], api_name="api_clear_textbox")
             )
 
+
+
+
+
+
+            
+
             chat_input.submit(fn=stop_audio, inputs=[], outputs=[audio_output], api_name="api_stop_audio_recording").then(
                 fn=add_message, inputs=[chatbot, chat_input], outputs=[chatbot, chat_input], api_name="api_addprompt_chathistory"
             ).then(
-                # First, generate the bot response and confidence score
-                fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot, confidence_plot_output], api_name="api_generate_bot_response"
+                # First, generate the bot response
+                fn=generate_bot_response, inputs=[chatbot, choice, retrieval_mode, model_choice], outputs=[chatbot], api_name="api_generate_bot_response"
             ).then(
                 # Then, generate the TTS response based on the bot's response
                 fn=generate_tts_response, inputs=[chatbot, tts_choice], outputs=[audio_output], api_name="api_generate_tts_response"
@@ -1711,8 +1542,46 @@ with gr.Blocks(theme='gradio/soft') as demo:
                 fn=clear_textbox, inputs=[], outputs=[chat_input], api_name="api_clear_textbox"
             )
 
+
+
+
+
+
+
             audio_input = gr.Audio(sources=["microphone"], streaming=True, type='numpy', every=0.1)
             audio_input.stream(transcribe_function, inputs=[state, audio_input], outputs=[state, chat_input], api_name="api_voice_to_text")
+            
+        #     gr.Markdown("<h1 style='color: red;'>Example Prompts</h1>", elem_id="Example-Prompts")
+        #     gr.Examples(examples=examples, fn=insert_prompt,inputs=chat_input, outputs=chat_input)
+
+        # with gr.Column():
+        #     weather_output = gr.HTML(value=fetch_local_weather())
+        #     news_output = gr.HTML(value=fetch_local_news())
+        #     events_output = gr.HTML(value=fetch_local_events())
+
+        # with gr.Column():
+           
+          
+        #     # Call update_images during the initial load to display images when the interface appears
+        #     initial_images = update_images()
+            
+        #     # Displaying the images generated using Flux API directly
+        #     image_output_1 = gr.Image(value=initial_images[0], label="Image 1", elem_id="flux_image_1", width=400, height=400)
+        #     image_output_2 = gr.Image(value=initial_images[1], label="Image 2", elem_id="flux_image_2", width=400, height=400)
+        #     image_output_3 = gr.Image(value=initial_images[2], label="Image 3", elem_id="flux_image_3", width=400, height=400)
+    
+        #     # Refresh button to update images
+        #     refresh_button = gr.Button("Refresh Images")
+        #     refresh_button.click(fn=update_images, inputs=None, outputs=[image_output_1, image_output_2, image_output_3])
+            
+
+
+
+
+           
 
 demo.queue()
-demo.launch(show_error=True)
+demo.launch(show_error=True)
+
+
+