Trying cache resource on vectordb build

app.py

@@ -17,12 +17,6 @@ from sklearn.model_selection import train_test_split
 
 # # Download dataset
 file_path = "dataset-tickets-multi-lang-4-20k.csv"
-# # Load the latest version
-# df = kagglehub.load_dataset(
-#   KaggleDatasetAdapter.PANDAS,
-#   "tobiasbueck/multilingual-customer-support-tickets",
-#   file_path,
-# )
 
 df = pd.read_csv(file_path)
 
@@ -47,13 +41,23 @@ documents = loader.load()
 
 # Get OpenAI setup
 openai_api_key = os.getenv("openai_token")
-embedding = OpenAIEmbeddings(openai_api_key=openai_api_key)
+# embedding = OpenAIEmbeddings(openai_api_key=openai_api_key)
 
-vectordb = Chroma.from_documents(
-    documents=documents,
-    embedding=embedding,
-    persist_directory=persist_directory
-)
+# vectordb = Chroma.from_documents(
+#     documents=documents,
+#     embedding=embedding,
+#     persist_directory=persist_directory
+# )
+
+@st.cache_resource
+def get_vectordb():
+    embedding = OpenAIEmbeddings(openai_api_key=os.getenv("openai_token"))
+    return Chroma.from_documents(
+        documents=documents,
+        embedding=embedding,
+        persist_directory=persist_directory)
+
+vectordb = get_vectordb()
 
 # @st.cache_resource
 # def get_vectordb():