Update app.py

app.py

@@ -8,6 +8,8 @@ from langchain.prompts import PromptTemplate
 from langchain.chains.question_answering import load_qa_chain
 from langchain.vectorstores import Chroma
 from langchain.retrievers import mmr_retriever
+from utills import load_txt_documents , split_docs, chroma_db,
+
 
 # Initialize variables and paths
 script_dir = os.path.dirname(os.path.abspath(__file__))
@@ -32,7 +34,7 @@ def load_embeddings():
 hf = load_embeddings()
 
 @st.cache_data
-def load_documents(data_path):
+def load_txt_documents(data_path):
     documents = []
     for filename in os.listdir(data_path):
         if filename.endswith('.txt'):
@@ -40,7 +42,7 @@ def load_documents(data_path):
             documents.extend(TextLoader(file_path).load())
     return documents
 
-documents = load_documents(data_path)
+documents = load_txt_documents(data_path)
 
 def split_docs(documents, chunk_size, overlap):
     # Your implementation here
@@ -48,25 +50,20 @@ def split_docs(documents, chunk_size, overlap):
 
 docs = split_docs(documents, 450, 20)
 
-@st.cache_resource
-def create_chroma_db(docs, hf):
-    return Chroma(docs, hf)
 
-chroma_db = create_chroma_db(docs, hf)
 
-@st.cache_resource
-def create_retriever(chroma_db):
-    return mmr_retriever(chroma_db, "mmr", 6)
+chroma_db = chroma_db(docs, hf)
+
+
+retriever = retriever_from_chroma(chroma_db,"mmr",6)
 
-retriever = create_retriever(chroma_db)
 
-# Set up LlamaCpp model
 callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
 
 @st.cache_resource
-def load_llm():
+def load_llm(model_path):
     return LlamaCpp(
-        model_path='qwen2-0_5b-instruct-q4_0.gguf',
+        model_path=model_path,
         n_gpu_layers=0,
         temperature=0.0,
         top_p=0.5,