upsert

app.py

@@ -46,7 +46,7 @@ client = chromadb.Client()
 collection = client.get_or_create_collection(name="embeddings_mxbai") 
 
 for batch in tqdm(batched_ds, desc="Processing dataset batches"): 
-    collection.add(
+    collection.upsert(
         ids=batch["id"],
         metadatas=batch["metadata"],
         documents=batch["document"],