Update app.py

app.py

@@ -1,31 +1,36 @@
 import streamlit as st
+import os
 
-# ✅ Correct imports (new structure)
+# ✅ Imports
 from langchain_community.document_loaders import PyPDFLoader, TextLoader
 from langchain_text_splitters import RecursiveCharacterTextSplitter
 from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain_community.vectorstores import FAISS
-from langchain_community.llms import HuggingFacePipeline
-
 from langchain.chains import RetrievalQA
+
 from transformers import pipeline
+from langchain_community.llms import HuggingFacePipeline
 
 
 # -------------------------------
-# Load Documents
+# Load Documents (SAFE PATH)
 # -------------------------------
 def load_documents(uploaded_files):
     documents = []
+
     for file in uploaded_files:
-        with open(file.name, "wb") as f:
+        file_path = os.path.join("/tmp", file.name)
+
+        with open(file_path, "wb") as f:
             f.write(file.getbuffer())
 
         if file.name.endswith(".pdf"):
-            loader = PyPDFLoader(file.name)
+            loader = PyPDFLoader(file_path)
         else:
-            loader = TextLoader(file.name)
+            loader = TextLoader(file_path)
 
         documents.extend(loader.load())
+
     return documents
 
 
@@ -41,43 +46,42 @@ def split_documents(documents):
 
 
 # -------------------------------
-# Create Vector Store
+# Create Vector Store (LOCAL)
 # -------------------------------
 def create_vectorstore(chunks):
     embeddings = HuggingFaceEmbeddings(
-        model_name="sentence-transformers/all-MiniLM-L6-v2"
+        model_name="sentence-transformers/all-MiniLM-L6-v2"  # works without token
     )
     return FAISS.from_documents(chunks, embeddings)
 
 
 # -------------------------------
-# Load Local LLM (FREE)
+# Load LOCAL LLM (VERY LIGHT)
 # -------------------------------
 def load_llm():
     pipe = pipeline(
-        "text2text-generation",   # ✅ FIXED
-        model="google/flan-t5-base",
-        max_length=512
+        "text2text-generation",
+        model="sshleifer/tiny-t5",   # 🔥 super light, no auth needed
+        max_length=256
     )
     return HuggingFacePipeline(pipeline=pipe)
 
 
 # -------------------------------
-# Build QA Chain (with strict prompt)
+# Build QA Chain
 # -------------------------------
 def build_qa(vectorstore):
     llm = load_llm()
     retriever = vectorstore.as_retriever()
 
-    qa = RetrievalQA.from_chain_type(
+    return RetrievalQA.from_chain_type(
         llm=llm,
         retriever=retriever
     )
-    return qa
 
 
 # -------------------------------
-# Streamlit UI
+# UI
 # -------------------------------
 st.set_page_config(page_title="RAG Chatbot", layout="wide")
 st.title("📄 Chat with Your Documents (RAG)")
@@ -94,12 +98,15 @@ if uploaded_files:
         vectorstore = create_vectorstore(chunks)
         qa_chain = build_qa(vectorstore)
 
-    st.success("Documents ready!")
+    st.success("✅ Documents ready!")
 
     query = st.text_input("Ask a question from your documents")
 
     if query:
         with st.spinner("Generating answer..."):
-            result = qa_chain.run(query)
-            st.write("### Answer:")
-            st.write(result)
+            try:
+                result = qa_chain.run(query)
+                st.write("### Answer:")
+                st.write(result)
+            except Exception as e:
+                st.error(str(e))