Upload folder using huggingface_hub

hf_gradio_ai_app.py

@@ -73,7 +73,8 @@ def initialize_ai_components():
         
         The user will describe, in plain English, the type of query they would like to run on this clinical table.
         Do your best to provide a SQL query that would return the data they are looking for.
-        If the user's prompt doesn't seem like a valid query request, just inform them that you cannot help with a task that is not query generation, and explain why their prompt doesn't look relevant to your function.
+        The user may also ask for the schema of the table, or for a description of the columns.
+        If the user's prompt doesn't seem like a valid query request or a valid question about the schema, inform them that you cannot help with their request, and explain why their prompt doesn't look relevant to your function.
 
         User Question: {user_input}
 

hf_gradio_ai_app_Ben_Clinical_Query_draft3.py

@@ -0,0 +1,185 @@
+# gradio_ai_chatbot_dotenv.py
+#
+# To run this script:
+# 1. Create a .env file in the same directory with your OPENAI_API_KEY.
+#    Example .env file content:
+#    OPENAI_API_KEY="sk-yourActualOpenAIapiKeyGoesHere"
+# 2. Install the required packages:
+#    pip install gradio langchain openai langchain_openai python-dotenv
+# 3. Run the script from your terminal:
+#    python gradio_ai_chatbot_dotenv.py
+#
+# The script will output a local URL and potentially a public Gradio link.
+
+import gradio as gr
+from langchain_openai import ChatOpenAI
+from langchain.prompts import ChatPromptTemplate
+import os
+from dotenv import load_dotenv
+
+# --- Load environment variables from .env file ---
+load_dotenv()
+
+# --- Global variables and Initial Setup ---
+OPENAI_API_KEY_GLOBAL = os.getenv("OPENAI_API_KEY")
+LANGCHAIN_LLM = None
+LANGCHAIN_PROMPT_TEMPLATE = None
+INITIAL_AI_SETUP_MESSAGE = "" # To store status/error from initial setup
+
+def initialize_ai_components():
+    """
+    Initializes LangChain components (LLM and prompt template) using the API key
+    from environment variables. Updates global variables and sets a status message.
+    """
+    global LANGCHAIN_LLM, LANGCHAIN_PROMPT_TEMPLATE, OPENAI_API_KEY_GLOBAL, INITIAL_AI_SETUP_MESSAGE
+    
+    if not OPENAI_API_KEY_GLOBAL:
+        INITIAL_AI_SETUP_MESSAGE = "<p style='color:red; font-weight:bold;'>ERROR: OpenAI API Key not found. Please ensure it's in your .env file or environment variables.</p>"
+        print("ERROR: OpenAI API Key not found. Make sure it's in your .env file or environment.")
+        return False # Indicate failure
+    
+    try:
+        # Initialize the LangChain LLM (OpenAI model)
+        LANGCHAIN_LLM = ChatOpenAI(openai_api_key=OPENAI_API_KEY_GLOBAL, model_name="gpt-4o-mini")
+        
+        # Define the prompt template for the LLM
+        prompt_template_str = """
+        You are a helpful, friendly, and insightful AI assistant.
+        You will be given access to a dataset that contains a single table. This table contains sample clinical information that was collected during clinical encounters. 
+        The rows in the table are per-encounter. Patients who have more frequent clinical encounters will therefore have more rows in the table.
+        The table has the following columns:
+			index: int64
+			ENCOUNTER_ID: int64
+			CLINICAL_NOTES: string
+			BIRTHDATE: string
+			FIRST: string
+			START: string
+			STOP: string
+			PATIENT_ID: int64
+			ENCOUNTERCLASS: string
+			CODE: int64
+			DESCRIPTION: string
+			BASE_ENCOUNTER_COST: float64
+			TOTAL_CLAIM_COST: float64
+			PAYER_COVERAGE: float64
+			REASONCODE: float64
+			REASONDESCRIPTION: string
+			PATIENT_AGE: int64
+			DESCRIPTION_OBSERVATIONS: string
+			DESCRIPTION_CONDITIONS: string
+			DESCRIPTION_MEDICATIONS: string
+			DESCRIPTION_PROCEDURES: string
+			CLINICAL_NOTES-embeddings: string
+        
+        The user will describe, in plain English, the type of query they would like to run on this clinical table.
+        Do your best to provide a SQL query that would return the data they are looking for.
+        The user may also ask for the schema of the table, or for a description of the columns.
+        If the user's prompt doesn't seem like a valid query request or a valid question about the schema, inform them that you cannot help with their request, and explain why their prompt doesn't look relevant to your function.
+
+        User Question: {user_input}
+
+        AI Response:
+        """
+        LANGCHAIN_PROMPT_TEMPLATE = ChatPromptTemplate.from_template(prompt_template_str)
+        
+        INITIAL_AI_SETUP_MESSAGE = "<p style='color:green; font-weight:bold;'>AI Components Initialized Successfully! Ready to chat. My purpose is to take in natural language questions about the clinical dataset and to generate a relevant SQL query.</p>"
+        print("AI Components Initialized Successfully!")
+        return True # Indicate success
+    except Exception as e:
+        INITIAL_AI_SETUP_MESSAGE = f"<p style='color:red; font-weight:bold;'>ERROR: Failed to initialize AI components. Error: {str(e)}. Please check your API key and model access.</p>"
+        LANGCHAIN_LLM = None
+        LANGCHAIN_PROMPT_TEMPLATE = None
+        print(f"ERROR: Failed to initialize AI components: {str(e)}")
+        return False # Indicate failure
+
+# --- Attempt to initialize AI components when the script loads ---
+AI_INITIALIZED_SUCCESSFULLY = initialize_ai_components()
+
+def ai_chat_response_function(user_message, chat_history):
+    """
+    This is the core function called by Gradio's ChatInterface.
+    It takes the user's message and the chat history, and returns the AI's response string.
+    """
+    if not AI_INITIALIZED_SUCCESSFULLY or not LANGCHAIN_LLM or not LANGCHAIN_PROMPT_TEMPLATE:
+        # Use the globally set error message from initialization
+        # Clean up HTML for plain error string if needed, or pass raw if Markdown supports it
+        error_msg_text = INITIAL_AI_SETUP_MESSAGE.replace("<p style='color:red; font-weight:bold;'>", "").replace("</p>", "")
+        return f"ERROR: AI is not ready. Status: {error_msg_text}"
+
+    # Proceed with generating response if components are ready
+    try:
+        # Create the LangChain chain (Prompt + LLM)
+        chain = LANGCHAIN_PROMPT_TEMPLATE | LANGCHAIN_LLM
+        
+        # Invoke the chain with the user's input
+        ai_response = chain.invoke({"user_input": user_message})
+        
+        # Return the content of the AI's response
+        return ai_response.content
+    except Exception as e:
+        print(f"Error during LangChain invocation: {e}") # Log for server-side debugging
+        return f"Sorry, an error occurred while trying to get a response: {str(e)}"
+
+# --- Gradio Interface Definition using gr.Blocks for layout control ---
+with gr.Blocks(theme=gr.themes.Soft(primary_hue=gr.themes.colors.blue, secondary_hue=gr.themes.colors.sky), title="AI Chatbot (Gradio)") as gradio_app:
+    gr.Markdown(
+        """
+        # 🤖 AI Chatbot with Gradio, LangChain & OpenAI
+        Powered by OpenAI's `gpt-4o-mini` model.
+        OpenAI API Key is loaded from your `.env` file.
+        """
+    )
+
+    # Display the initial AI setup status
+    gr.Markdown(INITIAL_AI_SETUP_MESSAGE)
+    
+    gr.Markdown("---") # Visual separator
+    gr.Markdown("## Chat Interface")
+
+    # Gradio ChatInterface for the main chat functionality
+    chat_interface_component = gr.ChatInterface(
+        fn=ai_chat_response_function, # The function that handles chat logic
+        chatbot=gr.Chatbot(
+            height=550,
+            show_label=False,
+            placeholder="AI's responses will appear here." if AI_INITIALIZED_SUCCESSFULLY else "AI is not available. Check setup status above.",
+            avatar_images=("https://raw.githubusercontent.com/svgmoji/svgmoji/main/packages/svgmoji__openmoji/svg/1F468-1F3FB-200D-1F9B0.svg", "https://raw.githubusercontent.com/gradio-app/gradio/main/gradio/icons/huggingface-logo.svg"),
+            type='messages'
+        ),
+        textbox=gr.Textbox(
+            placeholder="Type your message here and press Enter...",
+            show_label=False,
+            scale=7,
+            # Disable textbox if AI did not initialize successfully
+            interactive=AI_INITIALIZED_SUCCESSFULLY
+        ),
+        submit_btn="➡️ Send" if AI_INITIALIZED_SUCCESSFULLY else None, # Hide button if not ready
+        examples=[
+            "What is Paris, France known for?",
+            "Explain the concept of a Large Language Model (LLM) simply.",
+            "Can you give me a basic recipe for brownies?",
+            "Tell me an interesting fact about sunflowers."
+        ] if AI_INITIALIZED_SUCCESSFULLY else None, # Only show examples if AI is ready
+        title=None,
+        autofocus=True
+    )
+    
+    # If AI initialization failed, you might want to make the ChatInterface non-interactive.
+    # One way is to conditionally enable/disable components or hide buttons as done above.
+    if not AI_INITIALIZED_SUCCESSFULLY:
+        # Further disable parts of the chat interface if needed, though ChatInterface
+        # doesn't have a simple 'interactive=False' for the whole thing.
+        # Hiding buttons and disabling textbox is a good start.
+        # The error message in `ai_chat_response_function` will also prevent interaction.
+        pass
+
+
+# --- Main execution block to launch the Gradio app ---
+if __name__ == '__main__':
+    print("Attempting to launch Gradio App...")
+    if not OPENAI_API_KEY_GLOBAL:
+        print("WARNING: OpenAI API Key was not found in environment variables or .env file.")
+        print("The application UI will launch, but AI functionality will be disabled.")
+        print("Please create a .env file with your OPENAI_API_KEY.")
+    
+    gradio_app.launch(share=True, debug=True)