Update app.py

app.py

@@ -1,9 +1,7 @@
 import streamlit as st
-import pandas as pd
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 from sentence_transformers import SentenceTransformer
-import faiss
-import numpy as np
+import pickle
 
 st.title("Fin$mart Chatbot")
 st.markdown("Ask financial questions and get answers based on expert knowledge.")
@@ -18,37 +16,25 @@ def load_models():
 
 tokenizer, model, embedder = load_models()
 
-# Load data
-@st.cache_data
-def load_data():
-    df = pd.read_csv("train_data.csv").head(100)
-    df['content'] = "Q: " + df['question'] + " A: " + df['answer']  # Combine Q+A for better retrieval
-    return df
-
-data = load_data()
-
-# Build vector store
+# Load vector store from existing pickle file
 @st.cache_resource
-def build_vector_store(texts):
-    embeddings = embedder.encode(texts, show_progress_bar=True)
-    dim = embeddings[0].shape[0]
-    index = faiss.IndexFlatL2(dim)
-    index.add(np.array(embeddings))
-    return index, embeddings
+def load_vector_store():
+    with open("vectorstore.pkl", "rb") as f:
+        index, texts, _ = pickle.load(f)  # We ignore embeddings if not needed
+    return index, texts
 
-texts = data['content'].tolist()
-index, embeddings = build_vector_store(texts)
+index, texts = load_vector_store()
 
 # Chat interface
 prompt = st.chat_input("Ask something about finance...")
 
 if prompt:
-    # Embed and retrieve top 3 relevant pieces of text
+    # Embed query and retrieve top 3 results
     q_embed = embedder.encode([prompt])
-    _, I = index.search(np.array(q_embed), k=3)
+    _, I = index.search(q_embed, k=3)
     context = " ".join([texts[i] for i in I[0]])
 
-    # Prepare the prompt for flan-t5-base
+    # Build input for Flan-T5
     input_text = (
         f"You are a helpful financial assistant. Use the information provided below to answer the user's question.\n\n"
         f"Context: {context}\n\n"
@@ -60,10 +46,10 @@ if prompt:
     outputs = model.generate(**inputs, max_length=150)
     answer = tokenizer.decode(outputs[0], skip_special_tokens=True)
 
-    # Display the answer
+    # Display response
     st.markdown(f"**Answer:** {answer}")
 
-    # Optionally show the context used
+    # Show retrieved context
     with st.expander("Context Used"):
         for i in I[0]:
-            st.write(texts[i])
+            st.write(texts[i])