Update strolr_amia.py

strolr_amia.py

@@ -76,21 +76,21 @@
 
 ## DATABASE
 
-COLLECTION_NAME = "strolr_test"
-CONNECTION_STRING = PGVector.connection_string_from_db_params(
-     driver=os.environ.get("PGVECTOR_DRIVER", "psycopg2"),
-     host=os.environ.get("PGVECTOR_HOST", "vectordb.cfowaqqqovp0.us-east-2.rds.amazonaws.com"),
-     port=int(os.environ.get("PGVECTOR_PORT", "5432")),
-     database=os.environ.get("PGVECTOR_DATABASE", "postgres"),
-     user=os.environ.get("PGVECTOR_USER", "postgres"),
-     password=os.environ.get("PGVECTOR_PASSWORD", "temporary"),
-)
-
-conn = psycopg2.connect(
-    host="vectordb.cfowaqqqovp0.us-east-2.rds.amazonaws.com",
-    database="postgres",
-    user="postgres",
-    password="temporary")
+#COLLECTION_NAME = "strolr_test"
+#CONNECTION_STRING = PGVector.connection_string_from_db_params(
+#     driver=os.environ.get("PGVECTOR_DRIVER", "psycopg2"),
+#     host=os.environ.get("PGVECTOR_HOST", "vectordb.cfowaqqqovp0.us-east-2.rds.amazonaws.com"),
+#     port=int(os.environ.get("PGVECTOR_PORT", "5432")),
+#     database=os.environ.get("PGVECTOR_DATABASE", "postgres"),
+#     user=os.environ.get("PGVECTOR_USER", "postgres"),
+#     password=os.environ.get("PGVECTOR_PASSWORD", "temporary"),
+#)
+
+#conn = psycopg2.connect(
+#    host="vectordb.cfowaqqqovp0.us-east-2.rds.amazonaws.com",
+#    database="postgres",
+#    user="postgres",
+#    password="temporary")
 
 
 # Create a string for downloadable chat history
@@ -136,12 +136,17 @@ def format_response(responses):
 def load_chain_with_sources():
 
     embeddings = OpenAIEmbeddings()
+   # store = PGVector(
+   #     collection_name=COLLECTION_NAME,
+   #     connection_string=CONNECTION_STRING,
+   #     embedding_function=embeddings,
+   # )
+    connection = "postgresql+psycopg://langchain:[email protected]:5432/strolrdb"
+    collection_name = "strolr_docs"
     store = PGVector(
-        collection_name=COLLECTION_NAME,
-        connection_string=CONNECTION_STRING,
-        embedding_function=embeddings,
-    )
-
+    embeddings=embeddings,
+    collection_name=collection_name,
+    connection=connection)
     retriever = store.as_retriever(search_type="similarity_score_threshold", search_kwargs = {"k":3, "score_threshold":0.8})
     llm = ChatOpenAI(temperature = 0.8, model = "gpt-4o-mini")