Update retrieval.py

retrieval.py

@@ -80,7 +80,7 @@ class EmbedFunction:
 EMBED_MODEL_NAME = "sentence-transformers/all-MiniLM-L6-v2"
 embed_function = EmbedFunction(EMBED_MODEL_NAME)
 
-# Use a temporary directory for persistent storage
+# Use a temporary directory for persistent storage.
 temp_dir = tempfile.mkdtemp()
 print("Using temporary persist_directory:", temp_dir)
 
@@ -97,6 +97,15 @@ collection = client.get_or_create_collection(
     embedding_function=embed_function
 )
 
+# Force initialization: add a dummy document and perform a dummy query.
+try:
+    collection.add(documents=["dummy"], ids=["dummy"])
+    _ = collection.query(query_texts=["dummy"], n_results=1)
+    # Optionally, remove the dummy document if needed (Chromadb might not support deletion, so you can ignore it)
+    print("Dummy initialization successful.")
+except Exception as init_err:
+    print("Dummy initialization failed:", init_err)
+
 def index_pubmed_docs(docs: List[str], prefix: str = "doc"):
     """
     Adds documents to the Chromadb collection with unique IDs.