cleaned up repo

app2.py

@@ -1,120 +0,0 @@
-import gradio as gr
-import src.srf_bot as sb
-import prompts.system_prompts as sp
-from langchain_core.messages import HumanMessage
-
-# Initialize the chatbot
-chatbot = sb.SRFChatbot()
-
-# Dictionary to store passages with identifiers
-retrieved_passages = {}
-
-# Define the respond function
-def respond(query, history):
-    formatted_query = [HumanMessage(content=query)]
-    # Invoke the chatbot
-    result = chatbot.graph.invoke({"messages": formatted_query}, chatbot.config)
-    
-    # Extract the assistant's response
-    response = result["messages"][-1].content
-
-    # Retrieve passages from your vector database based on the query
-    # For the example, we'll use dummy passages
-    passages = [
-        "This is the full text of Passage 1.",
-        "This is the full text of Passage 2.",
-        "This is the full text of Passage 3."
-    ]
-
-    # Store passages with identifiers
-    passage_ids = []
-    for idx, passage in enumerate(passages):
-        identifier = f"Passage {idx+1}"
-        retrieved_passages[identifier] = passage
-        passage_ids.append(identifier)
-
-    # Reference passages in the response
-    linked_response = f"{response}\n\nReferences:"
-    for pid in passage_ids:
-        linked_response += f" [{pid}]"
-
-    # Append to history
-    history.append((query, linked_response))
-    return history, ""
-
-# Function to get passage content based on selection
-def get_passage_content(passage_id):
-    return retrieved_passages.get(passage_id, "Passage not found.")
-
-# Function to update the system prompt
-def update_system_prompt(selected_prompt):
-    # Update the chatbot's system prompt
-    chatbot.reset_system_prompt(selected_prompt)
-    # Update the displayed system prompt text
-    return sp.system_prompt_templates[selected_prompt]
-
-# Gradio interface
-with gr.Blocks() as demo:
-    gr.Markdown("# SRF Chatbot")
-
-    with gr.Row():
-        with gr.Column(scale=4):
-            # Chatbot interface
-            chatbot_output = gr.Chatbot()
-            user_input = gr.Textbox(placeholder="Type your question here...", label="Your Question")
-            submit_button = gr.Button("Submit")
-
-        with gr.Column(scale=1):
-            # Dropdown to select system prompts
-            system_prompt_dropdown = gr.Dropdown(
-                choices=list(sp.system_prompt_templates.keys()),
-                label="Select Chatbot Instructions",
-                value=list(sp.system_prompt_templates.keys())[0]
-            )
-            # Display the selected system prompt
-            system_prompt_display = gr.Textbox(
-                value=sp.system_prompt_templates[list(sp.system_prompt_templates.keys())[0]],
-                label="Current Chatbot Instructions",
-                lines=5,
-                interactive=False
-            )
-            
-            # Update system prompt display when a new prompt is selected
-            system_prompt_dropdown.change(
-                fn=update_system_prompt,
-                inputs=[system_prompt_dropdown],
-                outputs=[system_prompt_display]
-            )
-            
-            # Passage selection and display
-            gr.Markdown("### References")
-            passage_selector = gr.Dropdown(label="Select a passage to view", choices=[])
-            passage_display = gr.Markdown()
-
-    # Update the chatbot when the submit button is clicked
-    submit_button.click(
-        fn=respond,
-        inputs=[user_input, chatbot_output],
-        outputs=[chatbot_output, user_input]
-    )
-
-    # Update the passage selector options when the chatbot output changes
-    def update_passage_selector(chat_history):
-        # Get the latest passages
-        choices = list(retrieved_passages.keys())
-        return gr.update(choices=choices)
-    
-    chatbot_output.change(
-        fn=update_passage_selector,
-        inputs=[chatbot_output],
-        outputs=[passage_selector]
-    )
-
-    # Display the selected passage
-    passage_selector.change(
-        fn=get_passage_content,
-        inputs=[passage_selector],
-        outputs=[passage_display]
-    )
-
-demo.launch()

app_multi_tab.py

@@ -1,107 +0,0 @@
-import gradio as gr
-from prompts.system_prompts import get_systemprompt, system_prompt_templates
-
-# Define some pre-written templates for Tab 1
-templates = {
-    "Friendly Chatbot": "You are a helpful, friendly chatbot that engages in light-hearted conversations.",
-    "Technical Assistant": "You are a technical assistant specialized in answering questions related to Python programming.",
-    "Nutrition Advisor": "You provide evidence-based advice on nutrition and healthy eating habits.",
-}
-
-# Define some agentic workflows for Tab 2
-agentic_workflows = {
-    "Blog Post Generator": "This agent is designed to help generate a blog post based on user input.",
-    "Document Summarizer": "This agent summarizes long documents by extracting key points.",
-    "Task Manager": "This agent helps you organize tasks and provides step-by-step guidance."
-}
-
-# Chatbot logic for custom instructions or pre-written templates
-def chatbot_response(system_instructions, user_query):
-    if "friendly" in system_instructions.lower():
-        return f"Friendly Chatbot says: Hi there! 😊 How can I assist you today?"
-    elif "technical" in system_instructions.lower():
-        return f"Technical Assistant says: Sure! Here's some information on Python: {user_query}"
-    elif "nutrition" in system_instructions.lower():
-        return f"Nutrition Advisor says: Here's some advice about healthy eating: {user_query}"
-    else:
-        return f"Custom Chatbot says: {user_query}"
-
-# Chatbot conversation function
-def chatbot_conversation(system_instructions, chat_history, user_query):
-    response = chatbot_response(system_instructions, user_query)
-    chat_history.append((user_query, response))
-    return chat_history, ""
-
-# Chatbot conversation for predefined agentic workflows
-def agentic_chatbot_conversation(workflow_instructions, chat_history, user_query):
-    response = f"Agent Workflow ({workflow_instructions}) says: {user_query}"
-    chat_history.append((user_query, response))
-    return chat_history, ""
-
-# Function to update the interface when a selection is made from the dropdown (Tab 1)
-def update_interface(template_name, custom_instructions):
-    if template_name == "Custom Instructions":
-        return gr.update(visible=True), gr.update(visible=False)
-    else:
-        template_content = templates.get(template_name, "")
-        return gr.update(visible=False), gr.update(visible=True, value=template_content)
-
-# Build the Gradio interface with Tabs
-with gr.Blocks(css=".gradio-container {background-color: #F0F0F0;} .gr-button {background-color: #1E3A8A; color: white;} .gr-textbox textarea {font-size: 16px;} .gr-markdown {font-size: 18px; color: #1E3A8A;}") as demo:
-    
-    # Add Tabs
-    with gr.Tabs():
-        
-        # Tab 1: Custom Instructions or Pre-Written Templates
-        with gr.Tab("Custom Instructions Chatbot"):
-            gr.Markdown("""
-            <div style='background-color:#E0E0E0; padding: 20px; border-radius: 10px;'>
-                <h1 style='text-align: center; color: #1E3A8A;'>SRF Innovation Labs - AI Chatbot Use Case Explorer</h1>
-                <p style='font-size: 18px; text-align: center; color: #1E3A8A;'>
-                    This tool allows you to experiment with different system prompts, 
-                    giving you control over how the chatbot behaves. You can either use pre-defined templates 
-                    or write your own custom instructions.
-                </p>
-            </div>
-            """)
-
-            # Section to select system instructions from dropdown
-            gr.Markdown("<h2 style='color: #1E3A8A;'>Chatbot Setup</h2>")
-            template_name = gr.Dropdown(choices=["Custom Instructions"] + list(templates.keys()), label="Choose Instructions", value="Friendly Chatbot")
-            custom_instructions = gr.Textbox(label="Custom Instructions", visible=False, placeholder="Write your own instructions here...")
-            template_display = gr.Textbox(label="Template Content", interactive=False, visible=True)
-
-            # Chatbot interface
-            gr.Markdown("<h2 style='color: #1E3A8A;'>Chatbot Interaction</h2>")
-            chatbot = gr.Chatbot(label="Chatbot Conversation", height=300, show_label=False)
-            user_query = gr.Textbox(label="Your Query", placeholder="Ask a question or say something to the chatbot...")
-            submit_button = gr.Button("Send", elem_classes=["gr-button"])
-
-            # Update logic for Tab 1
-            template_name.change(fn=update_interface, inputs=[template_name, custom_instructions], outputs=[custom_instructions, template_display])
-            submit_button.click(fn=chatbot_conversation, inputs=[custom_instructions if template_name == "Custom Instructions" else template_display, chatbot, user_query], outputs=[chatbot, user_query])
-
-        # Tab 2: Predefined Agentic Workflows
-        with gr.Tab("Agentic Workflow Chatbots"):
-            gr.Markdown("""
-            <div style='background-color:#E0E0E0; padding: 20px; border-radius: 10px;'>
-                <h1 style='text-align: center; color: #1E3A8A;'>Agentic Workflow Explorer</h1>
-                <p style='font-size: 18px; text-align: center; color: #1E3A8A;'>
-                    Explore predefined agentic workflows that execute specific tasks, such as generating blog posts, 
-                    summarizing documents, or managing tasks.
-                </p>
-            </div>
-            """)
-
-            # Dropdown for selecting agentic workflows
-            workflow_name = gr.Dropdown(choices=list(agentic_workflows.keys()), label="Choose Agent Workflow", value="Blog Post Generator")
-            workflow_display = gr.Textbox(label="Workflow Description", interactive=False, visible=True)
-            workflow_chatbot = gr.Chatbot(label="Agent Workflow Conversation", height=300, show_label=False)
-            workflow_user_query = gr.Textbox(label="Your Query", placeholder="Ask the agent to perform a task...")
-            workflow_submit_button = gr.Button("Send", elem_classes=["gr-button"])
-
-            # Chatbot interaction for agentic workflows
-            workflow_submit_button.click(fn=agentic_chatbot_conversation, inputs=[workflow_name, workflow_chatbot, workflow_user_query], outputs=[workflow_chatbot, workflow_user_query])
-
-# Launch the app
-demo.launch()

example_app.py

@@ -1,63 +0,0 @@
-import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()

src/srf_bot.py

@@ -102,9 +102,9 @@ class SRFChatbot:
     ):
         # Initialize the LLM and the system message
         ## THIS SHOULD BE REFACTORED AS THERE IS REPEITITION OF THE CODE IN RESET SYSTEM PROMPT TOO
-        self.chatbot_instructions_dropdown = chatbot_instructions_dropdown
-        self.chatbot_instructions = sp.system_prompt_templates[self.chatbot_instructions_dropdown]
-        self.system_message = SystemMessage(content=self.chatbot_instructions)
+        # self.chatbot_instructions_dropdown = chatbot_instructions_dropdown
+        # self.chatbot_instructions = sp.system_prompt_templates[self.chatbot_instructions_dropdown]
+        # self.system_message = SystemMessage(content=self.chatbot_instructions)
 
         self.llm = ChatOpenAI(model=model, temperature=temperature)
         self.tools = ToolManager().get_tools()
@@ -113,7 +113,7 @@ class SRFChatbot:
         # Build the graph
         self.graph = self.build_graph()
         # Get the configurable
-        self.config = self.get_configurable()
+        self.reset_system_prompt(chatbot_instructions_dropdown)
     
     def reset_system_prompt(self, chatbot_instructions_dropdown: str):
         # Update the dropdown
@@ -122,9 +122,13 @@ class SRFChatbot:
         self.chatbot_instructions = sp.system_prompt_templates[self.chatbot_instructions_dropdown]
         # Reset the system prompt
         self.system_message = SystemMessage(content=self.chatbot_instructions)
-        # Reset the configurable
+        # Get the configurable
         self.config = self.get_configurable()
-        return self.chatbot_instructions
+        # Reset the system prompt info in the graph
+        self.graph.update_state(values={"system_message": self.system_message, "system_message_dropdown": self.chatbot_instructions_dropdown},
+                                config=self.config)
+
+        # return self.chatbot_instructions
 
     def get_configurable(self):
         # This thread id is used to keep track of the chatbot's conversation
@@ -137,8 +141,8 @@ class SRFChatbot:
     # Add the system message onto the llm
     ## THIS SHOULD BE REFACTORED SO THAT THE STATE ALWAYS HAS THE DEFINITIVE SYSTEM MESSAGE THAT SHOULD BE IN USE
     def chatbot(self, state: AgentState):
-        messages = [self.system_message] + state["messages"]
-        return {"messages": [self.llm_with_tools.invoke(messages)], "system_message": self.system_message, "system_message_dropdown": self.chatbot_instructions_dropdown}
+        messages = [state['system_message']] + state["messages"]
+        return {"messages": [self.llm_with_tools.invoke(messages)]}
     
     def build_graph(self):
         # Add chatbot state