Update app.py

app.py

@@ -1,29 +1,28 @@
-# Install required libraries.
-# Files, API key, etc
 import os
 from pathlib import Path
-from dotenv import load_dotenv
+from typing import Optional, List, Any
 import gradio as gr
 
-# LangChain imports
+# Fix OpenMP environment variable issue BEFORE other imports
+os.environ["OMP_NUM_THREADS"] = "1"
+
+# NO langchain.chains anymore (hopefully this fixes it)
 from langchain_huggingface import HuggingFaceEmbeddings
 from langchain_community.document_loaders import UnstructuredMarkdownLoader
 from langchain_text_splitters import MarkdownHeaderTextSplitter
 from langchain_community.vectorstores import Chroma
-from langchain.chains.combine_documents import create_stuff_documents_chain
-from langchain.chains import create_retrieval_chain
 from langchain_core.prompts import PromptTemplate
-from huggingface_hub import InferenceClient
+from langchain_core.runnables import RunnablePassthrough
+from langchain_core.output_parsers import StrOutputParser
 from langchain_core.language_models.llms import LLM
-from typing import Optional, List, Any
+from huggingface_hub import InferenceClient
 
-# Fix OpenMP environment variable issue
-os.environ["OMP_NUM_THREADS"] = "1"
+# Get HF token
+hf_token = os.getenv("HUGGINGFACEHUB_API_TOKEN")
 
-# Load environment variables first
-# load_dotenv() # This will need commented out on HuggingFace. To run the code yourself on your device, keep this and comment out the api_key var
-# via os.getenv
-api_key = os.getenv("HUGGINGFACEHUB_API_TOKEN")
+if not hf_token:
+    print("ERROR: Token not found in environment variables!")
+    exit(1)
 
 print("Loading portfolio data...")
 
@@ -66,13 +65,6 @@ prompt = PromptTemplate(
     partial_variables={"system_prompt": system_prompt_content}
 )
 
-# Get HF token
-hf_token = os.getenv("HUGGINGFACEHUB_API_TOKEN")
-
-if not hf_token:
-    print("ERROR: Token not found in environment variables!")
-    exit(1)
-
 print("Initializing model...")
 
 # Initialize InferenceClient
@@ -103,9 +95,23 @@ class HFInferenceClientLLM(LLM):
 # Create LLM instance
 llm = HFInferenceClientLLM(client=client, model="meta-llama/Llama-3.2-3B-Instruct")
 
-# Create the document processing chain
-combine_docs_chain = create_stuff_documents_chain(llm, prompt)
-qa_chain = create_retrieval_chain(vectorstore.as_retriever(), combine_docs_chain)
+# Create retriever
+retriever = vectorstore.as_retriever()
+
+# Build RAG chain manually - NO langchain.chains imports needed!
+def format_docs(docs):
+    return "\n\n".join([d.page_content for d in docs])
+
+# Create RAG chain using RunnablePassthrough
+rag_chain = (
+    {
+        "context": retriever | format_docs,
+        "input": RunnablePassthrough()
+    }
+    | prompt
+    | llm
+    | StrOutputParser()
+)
 
 print("RAG system ready!")
 
@@ -114,27 +120,25 @@ print("RAG system ready!")
 def chat_with_portfolio(message, history):
     """Process user message and return AI response"""
     try:
-        response = qa_chain.invoke({"input": message})
-        return response["answer"]
+        response = rag_chain.invoke(message)
+        return response
     except Exception as e:
         return f"Error: {str(e)}"
 
+
 # Aesthetics - Fixed CSS for proper icon sizing
 custom_css = """
 @import url('https://fonts.googleapis.com/css2?family=Quicksand:wght@400;600&display=swap');
 
-/* Global Font and Body */
 * { 
     font-family: 'Quicksand', sans-serif !important; 
 }
 
-/* The Main Background */
 .gradio-container {
     background: linear-gradient(135deg, #a78bfa 0%, #6366f1 50%, #3b82f6 100%) !important;
     background-attachment: fixed !important;
 }
 
-/* Glassmorphism for the Chat bubbles */
 .message-wrap {
     background: rgba(255, 255, 255, 0.15) !important;
     backdrop-filter: blur(12px) !important;
@@ -147,7 +151,6 @@ custom_css = """
     color: white !important;
 }
 
-/* Input Box Styling */
 .input-wrap textarea, 
 textarea {
     background: rgba(255, 255, 255, 0.1) !important;
@@ -161,7 +164,6 @@ textarea::placeholder {
     color: rgba(255, 255, 255, 0.6) !important;
 }
 
-/* FIX FOR HUGE ICONS - Constrain all buttons and icons */
 button {
     max-width: 100px !important;
     max-height: 44px !important;
@@ -176,14 +178,12 @@ button svg {
     max-height: 20px !important;
 }
 
-/* Specific fix for send button */
 .submit-btn, 
 button[aria-label="Submit"] {
     max-width: 60px !important;
     width: 60px !important;
 }
 
-/* Example buttons styling */
 .examples button {
     background: rgba(255, 255, 255, 0.1) !important;
     border: 1px solid rgba(255, 255, 255, 0.2) !important;
@@ -196,12 +196,10 @@ button[aria-label="Submit"] {
     background: rgba(255, 255, 255, 0.2) !important;
 }
 
-/* Hide Gradio footer */
 footer { 
     display: none !important; 
 }
 
-/* Chatbot container styling */
 .chatbot {
     background: rgba(255, 255, 255, 0.05) !important;
     border-radius: 20px !important;
@@ -209,7 +207,6 @@ footer {
 """
 
 with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
-    # Title with specific white color for contrast
     gr.Markdown(
         "<h1 style='text-align: center; color: white; text-shadow: 2px 2px 4px rgba(0,0,0,0.3);'>🐈‍⬛ Chat with Adda-Bot</h1>"
     )
@@ -228,4 +225,4 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()