Update app.py

app.py

@@ -1,5 +1,6 @@
 import streamlit as st
 import tempfile
+import os
 
 from langchain_community.document_loaders import PyPDFLoader, TextLoader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
@@ -8,7 +9,7 @@ from langchain.vectorstores import FAISS
 from langchain.llms import HuggingFacePipeline
 from langchain.chains import RetrievalQA
 
-from transformers import pipeline
+from transformers.pipelines import pipeline   # ✅ FIXED IMPORT
 
 # -------------------------------
 # Page Config
@@ -24,18 +25,23 @@ def load_documents(uploaded_files):
     documents = []
 
     for file in uploaded_files:
-        # Save file safely using temp file
-        with tempfile.NamedTemporaryFile(delete=False, suffix=file.name) as tmp:
+        file_extension = os.path.splitext(file.name)[1]
+
+        # Save safely as temp file
+        with tempfile.NamedTemporaryFile(delete=False, suffix=file_extension) as tmp:
             tmp.write(file.getbuffer())
             temp_path = tmp.name
 
-        # Load based on type
-        if file.name.endswith(".pdf"):
-            loader = PyPDFLoader(temp_path)
-        else:
-            loader = TextLoader(temp_path)
+        try:
+            if file_extension.lower() == ".pdf":
+                loader = PyPDFLoader(temp_path)
+            else:
+                loader = TextLoader(temp_path)
+
+            documents.extend(loader.load())
 
-        documents.extend(loader.load())
+        except Exception as e:
+            st.error(f"❌ Error loading file: {e}")
 
     return documents
 
@@ -52,7 +58,7 @@ def split_documents(documents):
 
 
 # -------------------------------
-# Cached Embeddings (IMPORTANT)
+# Cached Embeddings
 # -------------------------------
 @st.cache_resource
 def get_embeddings():
@@ -70,13 +76,13 @@ def create_vectorstore(chunks):
 
 
 # -------------------------------
-# Cached LLM (IMPORTANT)
+# Cached LLM (FIXED)
 # -------------------------------
 @st.cache_resource
 def load_llm():
     pipe = pipeline(
-        "text-generation",
-        model="google/flan-t5-small",  # lightweight model
+        "text2text-generation",   # ✅ CORRECT TASK
+        model="google/flan-t5-small",
         max_length=256
     )
     return HuggingFacePipeline(pipeline=pipe)
@@ -108,6 +114,11 @@ uploaded_files = st.file_uploader(
 if uploaded_files:
     with st.spinner("📄 Processing documents..."):
         docs = load_documents(uploaded_files)
+
+        if not docs:
+            st.error("❌ No valid documents loaded.")
+            st.stop()
+
         chunks = split_documents(docs)
         vectorstore = create_vectorstore(chunks)
         qa_chain = build_qa(vectorstore)
@@ -121,7 +132,9 @@ if uploaded_files:
 
     if query:
         with st.spinner("🤖 Generating answer..."):
-            result = qa_chain.run(query)
-
-        st.markdown("### 🧠 Answer:")
-        st.write(result)
+            try:
+                result = qa_chain.run(query)
+                st.markdown("### 🧠 Answer:")
+                st.write(result)
+            except Exception as e:
+                st.error(f"❌ Error generating answer: {e}")