From 6711e91dbf8ff631a04d7cff41dd92ff58b77502 Mon Sep 17 00:00:00 2001 From: Yuhong Sun Date: Sat, 14 Dec 2024 17:23:00 -0800 Subject: [PATCH] Seed Spacing (#3474) --- backend/onyx/seeding/load_docs.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/backend/onyx/seeding/load_docs.py b/backend/onyx/seeding/load_docs.py index 36eccd6dc647..c7c78293d212 100644 --- a/backend/onyx/seeding/load_docs.py +++ b/backend/onyx/seeding/load_docs.py @@ -9,6 +9,7 @@ from onyx.access.models import default_public_access from onyx.configs.constants import DEFAULT_BOOST from onyx.configs.constants import DocumentSource from onyx.configs.constants import KV_DOCUMENTS_SEEDED_KEY +from onyx.configs.constants import RETURN_SEPARATOR from onyx.configs.model_configs import DEFAULT_DOCUMENT_ENCODER_MODEL from onyx.connectors.models import Document from onyx.connectors.models import IndexAttemptMetadata @@ -71,7 +72,7 @@ def _create_indexable_chunks( source_links={0: preprocessed_doc["url"]}, section_continuation=False, source_document=document, - title_prefix=preprocessed_doc["title"], + title_prefix=preprocessed_doc["title"] + RETURN_SEPARATOR, metadata_suffix_semantic="", metadata_suffix_keyword="", mini_chunk_texts=None,