diff options
| -rw-r--r-- | rag/parser/pdf.py | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/rag/parser/pdf.py b/rag/parser/pdf.py index a5e4665..1680a47 100644 --- a/rag/parser/pdf.py +++ b/rag/parser/pdf.py @@ -8,7 +8,8 @@ from langchain_community.document_loaders import PyPDFLoader def parser(filepath: Path): content = PyPDFLoader(filepath).load() splitter = RecursiveCharacterTextSplitter( - chunk_size=os.environ["CHUNK_SIZE"], chunk_overlap=os.environ["CHUNK_OVERLAP"] + chunk_size=int(os.environ["CHUNK_SIZE"]), + chunk_overlap=int(os.environ["CHUNK_OVERLAP"]), ) chunks = splitter.split_documents(content) return chunks |