diff options
| author | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2024-04-05 01:57:31 +0200 |
|---|---|---|
| committer | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2024-04-05 01:57:31 +0200 |
| commit | 07741e09f4569d27aca3ffa111178bc324af1eab (patch) | |
| tree | 17d2dcb35c3b7cc4e4e8a2602e3a4f50e671f634 /rag/parser/pdf.py | |
| parent | 901212a2fd9658e9b51a4e977c68a47ef7b840b7 (diff) | |
Add .env for parser
Diffstat (limited to 'rag/parser/pdf.py')
| -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 |