Update app.py

app.py

@@ -3,62 +3,47 @@ from gradio_client import Client, handle_file
 import os
 
 # Define your Hugging Face token (make sure to set it as an environment variable)
-HF_TOKEN = os.getenv("HF_TOKEN")  # Replace with your actual token if not using env variable
+HF_TOKEN = os.getenv("HF_TOKEN")  # Replace with your actual token if not using an environment variable
 
 # Initialize the Gradio Client for the specified API
 client = Client("on1onmangoes/CNIHUB10724v9", hf_token=HF_TOKEN)
 
 # Authentication function
 def login(username, password):
-    if username == "your_username" and password == "your_password":  # Update with actual credentials
-        return True
-    else:
-        return False
-
-# Function to handle different API calls based on user input
-def handle_api_call(username, password, message=None, client_name="rosariarossi", 
-                    system_prompt="You are an expert assistant", num_retrieved_docs=10, 
-                    num_docs_final=9, temperature=0, max_new_tokens=1024, 
-                    top_p=1, top_k=20, penalty=1.2, 
-                    pdf_file=None, query=None, question=None):
-    
-    if not login(username, password):
-        return "Invalid credentials! Please try again."
-
-    if message:
-        # Handle chat message
-        chat_result = client.predict(
-            message=message,
-            client_name=client_name,
-            system_prompt=system_prompt,
-            num_retrieved_docs=num_retrieved_docs,
-            num_docs_final=num_docs_final,
-            temperature=temperature,
-            max_new_tokens=max_new_tokens,
-            top_p=top_p,
-            top_k=top_k,
-            penalty=penalty,
-            api_name="/chat"
-        )
-        return chat_result
-    elif pdf_file:
-        # Handle PDF file
-        pdf_result = client.predict(
-            pdf_file=handle_file(pdf_file),
-            client_name=client_name,
-            api_name="/process_pdf2"
-        )
-        return pdf_result[1]  # Returning the string result from the PDF processing
-    elif query:
-        # Handle search query
-        search_result = client.predict(query=query, api_name="/search_with_confidence")
-        return search_result
-    elif question:
-        # Handle question for RAG
-        rag_result = client.predict(question=question, api_name="/answer_with_rag")
-        return rag_result
-    else:
-        return "No valid input provided!"
+    return username == "your_username" and password == "your_password"  # Update with actual credentials
+
+# Function to handle chat API call
+def stream_chat_with_rag(message, client_name, system_prompt, num_retrieved_docs, num_docs_final, temperature, max_new_tokens, top_p, top_k, penalty):
+    response = client.predict(
+        message=message,
+        client_name=client_name,
+        system_prompt=system_prompt,
+        num_retrieved_docs=num_retrieved_docs,
+        num_docs_final=num_docs_final,
+        temperature=temperature,
+        max_new_tokens=max_new_tokens,
+        top_p=top_p,
+        top_k=top_k,
+        penalty=penalty,
+        api_name="/chat"
+    )
+    return response
+
+# Function to handle PDF processing API call
+def process_pdf(pdf_file, client_name):
+    return client.predict(
+        pdf_file=handle_file(pdf_file),
+        client_name=client_name,
+        api_name="/process_pdf2"
+    )[1]  # Return only the result string
+
+# Function to handle search API call
+def search_api(query):
+    return client.predict(query=query, api_name="/search_with_confidence")
+
+# Function to handle RAG API call
+def rag_api(question):
+    return client.predict(question=question, api_name="/answer_with_rag")
 
 # Create the Gradio Blocks interface
 with gr.Blocks() as app:
@@ -69,109 +54,130 @@ with gr.Blocks() as app:
         password_input = gr.Textbox(label="Password", placeholder="Enter your password", type="password")
     
     with gr.Tab("Chat"):
-        message_input = gr.Textbox(label="Message", placeholder="Type your message here")
-        
-        gr.Markdown("### Client Options")
-        client_name_dropdown = gr.Dropdown(
-            label="Select Client",
-            choices=["rosariarossi", "bianchifiordaliso", "lorenzoverdi"],
-            value="rosariarossi"
-        )
-        
-        system_prompt_input = gr.Textbox(
-            label="System Prompt",
-            placeholder="Enter system prompt here",
-            value="You are an expert assistant"
-        )
-        
-        num_retrieved_docs_slider = gr.Slider(
-            label="Number of Initial Documents to Retrieve",
-            minimum=1,
-            maximum=100,
-            step=1,
-            value=10
-        )
+        chatbot = gr.Chatbot(fill_height=True)  # Create a chatbot interface
         
-        num_docs_final_slider = gr.Slider(
-            label="Number of Final Documents to Retrieve",
-            minimum=1,
-            maximum=100,
-            step=1,
-            value=9
-        )
+        # Create an accordion for additional input parameters
+        additional_inputs_accordion = gr.Accordion(label="⚙️ Parameters", open=False, render=False)
         
-        temperature_slider = gr.Slider(
-            label="Temperature",
-            minimum=0,
-            maximum=2,
-            step=0.1,
-            value=0
-        )
+        with additional_inputs_accordion:
+            client_name_dropdown = gr.Dropdown(
+                label="Select Client",
+                choices=["rosariarossi", "bianchifiordaliso", "lorenzoverdi"],
+                value="rosariarossi",
+                render=False
+            )
+            system_prompt_input = gr.Textbox(
+                value="You are an expert assistant",
+                label="System Prompt",
+                render=False
+            )
+            num_retrieved_docs_slider = gr.Slider(
+                label="Number of Initial Documents to Retrieve",
+                minimum=1,
+                maximum=10,
+                step=1,
+                value=10,
+                render=False
+            )
+            num_docs_final_slider = gr.Slider(
+                label="Number of Final Documents to Retrieve",
+                minimum=1,
+                maximum=10,
+                step=1,
+                value=9,
+                render=False
+            )
+            temperature_slider = gr.Slider(
+                label="Temperature",
+                minimum=0.2,
+                maximum=1,
+                step=0.1,
+                value=0,
+                render=False
+            )
+            max_new_tokens_slider = gr.Slider(
+                label="Max New Tokens",
+                minimum=128,
+                maximum=8192,
+                step=1,
+                value=1024,
+                render=False
+            )
+            top_p_slider = gr.Slider(
+                label="Top P",
+                minimum=0.0,
+                maximum=1.0,
+                step=0.1,
+                value=1.0,
+                render=False
+            )
+            top_k_slider = gr.Slider(
+                label="Top K",
+                minimum=1,
+                maximum=20,
+                step=1,
+                value=20,
+                render=False
+            )
+            penalty_slider = gr.Slider(
+                label="Repetition Penalty",
+                minimum=0.0,
+                maximum=2.0,
+                step=0.1,
+                value=1.2,
+                render=False
+            )
         
-        max_new_tokens_slider = gr.Slider(
-            label="Max New Tokens",
-            minimum=1,
-            maximum=2048,
-            step=1,
-            value=1024
+        chat_interface = gr.ChatInterface(
+            fn=stream_chat_with_rag,
+            chatbot=chatbot,
+            fill_height=True,
+            additional_inputs_accordion=additional_inputs_accordion,
+            additional_inputs=[
+                client_name_dropdown,
+                system_prompt_input,
+                num_retrieved_docs_slider,
+                num_docs_final_slider,
+                temperature_slider,
+                max_new_tokens_slider,
+                top_p_slider,
+                top_k_slider,
+                penalty_slider,
+            ],
         )
-        
-        top_p_slider = gr.Slider(
-            label="Top P",
-            minimum=0,
-            maximum=1,
-            step=0.01,
-            value=1
-        )
-        
-        top_k_slider = gr.Slider(
-            label="Top K",
-            minimum=1,
-            maximum=100,
-            step=1,
-            value=20
-        )
-        
-        penalty_slider = gr.Slider(
-            label="Repetition Penalty",
-            minimum=1,
-            maximum=5,
-            step=0.1,
-            value=1.2
-        )
-        
-        chat_output = gr.Textbox(label="Chat Response", interactive=False)
 
     with gr.Tab("Process PDF"):
         pdf_input = gr.File(label="Upload PDF File")
         pdf_output = gr.Textbox(label="PDF Result", interactive=False)
 
+        pdf_button = gr.Button("Process PDF")
+        pdf_button.click(
+            process_pdf,
+            inputs=[pdf_input, client_name_dropdown],
+            outputs=pdf_output
+        )
+
     with gr.Tab("Search"):
         query_input = gr.Textbox(label="Enter Search Query")
         search_output = gr.Textbox(label="Search Confidence Result", interactive=False)
 
+        search_button = gr.Button("Search")
+        search_button.click(
+            search_api,
+            inputs=query_input,
+            outputs=search_output
+        )
+
     with gr.Tab("Answer with RAG"):
         question_input = gr.Textbox(label="Enter Question for RAG")
         rag_output = gr.Textbox(label="RAG Answer Result", interactive=False)
 
-    api_button = gr.Button("Submit")
-
-    # Bind the button click to the handle_api_call function
-    api_button.click(
-        handle_api_call,
-        inputs=[
-            username_input, password_input,
-            message_input, client_name_dropdown, 
-            system_prompt_input, num_retrieved_docs_slider, 
-            num_docs_final_slider, temperature_slider, 
-            max_new_tokens_slider, top_p_slider, 
-            top_k_slider, penalty_slider,
-            pdf_input, query_input, question_input
-        ],
-        outputs=[
-            chat_output, pdf_output, search_output, rag_output
-        ]
-    )
+        rag_button = gr.Button("Get Answer")
+        rag_button.click(
+            rag_api,
+            inputs=question_input,
+            outputs=rag_output
+        )
 
 # Launch the app
 app.launch()
@@ -182,6 +188,193 @@ app.launch()
 
 
 
+
+
+
+# import gradio as gr
+# from gradio_client import Client, handle_file
+# import os
+
+# # Define your Hugging Face token (make sure to set it as an environment variable)
+# HF_TOKEN = os.getenv("HF_TOKEN")  # Replace with your actual token if not using env variable
+
+# # Initialize the Gradio Client for the specified API
+# client = Client("on1onmangoes/CNIHUB10724v9", hf_token=HF_TOKEN)
+
+# # Authentication function
+# def login(username, password):
+#     if username == "your_username" and password == "your_password":  # Update with actual credentials
+#         return True
+#     else:
+#         return False
+
+# # Function to handle different API calls based on user input
+# def handle_api_call(username, password, message=None, client_name="rosariarossi", 
+#                     system_prompt="You are an expert assistant", num_retrieved_docs=10, 
+#                     num_docs_final=9, temperature=0, max_new_tokens=1024, 
+#                     top_p=1, top_k=20, penalty=1.2, 
+#                     pdf_file=None, query=None, question=None):
+    
+#     if not login(username, password):
+#         return "Invalid credentials! Please try again."
+
+#     if message:
+#         # Handle chat message
+#         chat_result = client.predict(
+#             message=message,
+#             client_name=client_name,
+#             system_prompt=system_prompt,
+#             num_retrieved_docs=num_retrieved_docs,
+#             num_docs_final=num_docs_final,
+#             temperature=temperature,
+#             max_new_tokens=max_new_tokens,
+#             top_p=top_p,
+#             top_k=top_k,
+#             penalty=penalty,
+#             api_name="/chat"
+#         )
+#         return chat_result
+#     elif pdf_file:
+#         # Handle PDF file
+#         pdf_result = client.predict(
+#             pdf_file=handle_file(pdf_file),
+#             client_name=client_name,
+#             api_name="/process_pdf2"
+#         )
+#         return pdf_result[1]  # Returning the string result from the PDF processing
+#     elif query:
+#         # Handle search query
+#         search_result = client.predict(query=query, api_name="/search_with_confidence")
+#         return search_result
+#     elif question:
+#         # Handle question for RAG
+#         rag_result = client.predict(question=question, api_name="/answer_with_rag")
+#         return rag_result
+#     else:
+#         return "No valid input provided!"
+
+# # Create the Gradio Blocks interface
+# with gr.Blocks() as app:
+#     gr.Markdown("### Login")
+    
+#     with gr.Row():
+#         username_input = gr.Textbox(label="Username", placeholder="Enter your username")
+#         password_input = gr.Textbox(label="Password", placeholder="Enter your password", type="password")
+    
+#     with gr.Tab("Chat"):
+#         message_input = gr.Textbox(label="Message", placeholder="Type your message here")
+        
+#         gr.Markdown("### Client Options")
+#         client_name_dropdown = gr.Dropdown(
+#             label="Select Client",
+#             choices=["rosariarossi", "bianchifiordaliso", "lorenzoverdi"],
+#             value="rosariarossi"
+#         )
+        
+#         system_prompt_input = gr.Textbox(
+#             label="System Prompt",
+#             placeholder="Enter system prompt here",
+#             value="You are an expert assistant"
+#         )
+        
+#         num_retrieved_docs_slider = gr.Slider(
+#             label="Number of Initial Documents to Retrieve",
+#             minimum=1,
+#             maximum=100,
+#             step=1,
+#             value=10
+#         )
+        
+#         num_docs_final_slider = gr.Slider(
+#             label="Number of Final Documents to Retrieve",
+#             minimum=1,
+#             maximum=100,
+#             step=1,
+#             value=9
+#         )
+        
+#         temperature_slider = gr.Slider(
+#             label="Temperature",
+#             minimum=0,
+#             maximum=2,
+#             step=0.1,
+#             value=0
+#         )
+        
+#         max_new_tokens_slider = gr.Slider(
+#             label="Max New Tokens",
+#             minimum=1,
+#             maximum=2048,
+#             step=1,
+#             value=1024
+#         )
+        
+#         top_p_slider = gr.Slider(
+#             label="Top P",
+#             minimum=0,
+#             maximum=1,
+#             step=0.01,
+#             value=1
+#         )
+        
+#         top_k_slider = gr.Slider(
+#             label="Top K",
+#             minimum=1,
+#             maximum=100,
+#             step=1,
+#             value=20
+#         )
+        
+#         penalty_slider = gr.Slider(
+#             label="Repetition Penalty",
+#             minimum=1,
+#             maximum=5,
+#             step=0.1,
+#             value=1.2
+#         )
+        
+#         chat_output = gr.Textbox(label="Chat Response", interactive=False)
+
+#     with gr.Tab("Process PDF"):
+#         pdf_input = gr.File(label="Upload PDF File")
+#         pdf_output = gr.Textbox(label="PDF Result", interactive=False)
+
+#     with gr.Tab("Search"):
+#         query_input = gr.Textbox(label="Enter Search Query")
+#         search_output = gr.Textbox(label="Search Confidence Result", interactive=False)
+
+#     with gr.Tab("Answer with RAG"):
+#         question_input = gr.Textbox(label="Enter Question for RAG")
+#         rag_output = gr.Textbox(label="RAG Answer Result", interactive=False)
+
+#     api_button = gr.Button("Submit")
+
+#     # Bind the button click to the handle_api_call function
+#     api_button.click(
+#         handle_api_call,
+#         inputs=[
+#             username_input, password_input,
+#             message_input, client_name_dropdown, 
+#             system_prompt_input, num_retrieved_docs_slider, 
+#             num_docs_final_slider, temperature_slider, 
+#             max_new_tokens_slider, top_p_slider, 
+#             top_k_slider, penalty_slider,
+#             pdf_input, query_input, question_input
+#         ],
+#         outputs=[
+#             chat_output, pdf_output, search_output, rag_output
+#         ]
+#     )
+
+# # Launch the app
+# app.launch()
+
+
+
+
+
+
+
 # import gradio as gr
 # from gradio_client import Client, handle_file
 # import os