From 7691d4ea4677cb56aa68f8534bee716ebe73a829 Mon Sep 17 00:00:00 2001 From: Greg Hogue Date: Tue, 23 Sep 2025 14:13:32 -0400 Subject: [PATCH] reduce default chunksize for reactome embedding --- src/data_generation/reactome/__init__.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/data_generation/reactome/__init__.py b/src/data_generation/reactome/__init__.py index ce36442..f62cd38 100644 --- a/src/data_generation/reactome/__init__.py +++ b/src/data_generation/reactome/__init__.py @@ -53,11 +53,13 @@ def upload_to_chromadb( embeddings_instance: Embeddings if hf_model is None: # Use OpenAI embeddings_instance = OpenAIEmbeddings( + chunk_size=500, show_progress_bar=True, ) elif hf_model.startswith("openai/text-embedding-"): embeddings_instance = OpenAIEmbeddings( model=hf_model[len("openai/") :], + chunk_size=500, show_progress_bar=True, ) elif "HUGGINGFACEHUB_API_TOKEN" in os.environ: