Update rag.py

rag.py

@@ -16,11 +16,19 @@ from llama_index.core.agent import ReActAgent
 from chromadb import PersistentClient
 
 # writable cache directories
-os.environ["TRANSFORMERS_CACHE"] = "/app/.cache"
-os.environ["HF_HOME"] = "/app/.cache"
-os.environ["XDG_CACHE_HOME"] = "/app/.cache"  # Some libs use this
+CACHE_DIR = "/app/.cache"
+HUGGINGFACE_CACHE = os.path.join(CACHE_DIR, "huggingface")
+VECTORDb_DIR = "/app/vectordb"
+STREAMLIT_CONFIG = "/app/.streamlit"
 
-os.makedirs("/app/.cache", exist_ok=True)
+os.makedirs(HUGGINGFACE_CACHE, exist_ok=True)
+os.makedirs(VECTORDb_DIR, exist_ok=True)
+os.makedirs(STREAMLIT_CONFIG, exist_ok=True)
+
+os.environ["TRANSFORMERS_CACHE"] = CACHE_DIR
+os.environ["HF_HOME"] = CACHE_DIR
+os.environ["XDG_CACHE_HOME"] = CACHE_DIR
+os.environ["STREAMLIT_HOME"] = STREAMLIT_CONFIG
 
 @st.cache_resource
 def setup_rag_system(debug=False):
@@ -40,7 +48,11 @@ def setup_rag_system(debug=False):
     )
 
     # Embeddings
-    embedding_model = HuggingFaceEmbedding(model_name="sentence-transformers/all-MiniLM-L6-v2")
+    embedding_model = HuggingFaceEmbedding(
+        model_name="sentence-transformers/all-MiniLM-L6-v2",
+        cache_folder=HUGGINGFACE_CACHE
+    )
+
 
     # Persisted vector DBs
     persist_dirs = [
@@ -80,7 +92,7 @@ def setup_rag_system(debug=False):
         hybrid_retriever = RecursiveRetriever(
             "vector",
             retriever_dict={"vector": vector_retriever, "bm25": bm25_retriever},
-            verbose=True
+            verbose=False
         )
         hybrid_retrievers.append(hybrid_retriever)
 
@@ -166,7 +178,7 @@ def setup_rag_system(debug=False):
     agent = ReActAgent(
         tools=retriever_tools,
         llm=llm,
-        verbose=True,
+        verbose=False,
         max_iterations=20,
         system_prompt=system_prompt
     )