Update app.py

app.py

@@ -45,20 +45,20 @@ def create_retriever_from_chroma(vectorstore_path="docs/chroma/", search_type='m
 
     
     # Check if vectorstore exists
-    if os.path.exists(vectorstore_path) and os.listdir(vectorstore_path):
+    #if os.path.exists(vectorstore_path) and os.listdir(vectorstore_path):
         # Load the existing vectorstore
-        vectorstore = Chroma(persist_directory=vectorstore_path,embedding_function=embeddings)
-    else:
+    #    vectorstore = Chroma(persist_directory=vectorstore_path,embedding_function=embeddings)
+    #else:
         # Load documents from the specified data path
-        loader = DirectoryLoader('./data/', glob="./*.txt", loader_cls=TextLoader)
-        docs = loader.load()
-        text_splitter = RecursiveCharacterTextSplitter(chunk_size=chunk_size, chunk_overlap=chunk_overlap)
-        split_docs = text_splitter.split_documents(docs)
+    loader = DirectoryLoader('./data/', glob="./*.txt", loader_cls=TextLoader)
+    docs = loader.load()
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=chunk_size, chunk_overlap=chunk_overlap)
+    split_docs = text_splitter.split_documents(docs)
 
         
 
         # Create the vectorstore
-        vectorstore = Chroma.from_documents(
+    vectorstore = Chroma.from_documents(
             documents=split_docs, embedding=embeddings, persist_directory=vectorstore_path
         )
     
@@ -94,7 +94,7 @@ def main():
 
     st.markdown("Hi, I am Qwen, chat mmodel, based on respublic of Lithuania law document. Write you question and press enter to start chat.")
     
-    retriever = create_retriever_from_chroma(vectorstore_path="docs/chroma/", search_type='mmr', k=12, chunk_size=300, chunk_overlap=20)
+    retriever = create_retriever_from_chroma(vectorstore_path="docs/chroma/", search_type='mmr', k=9, chunk_size=300, chunk_overlap=20)
     if user_question := st.text_input("Ask a question about your documents:"):
         handle_userinput(user_question,retriever)