From d1ed616a047a4dc4e767a7dd320b17c008261abc Mon Sep 17 00:00:00 2001 From: bryan davis Date: Mon, 2 Feb 2026 15:20:31 -0600 Subject: [PATCH] feat: turn off ocr on docling serve chunking to avoid issues with read-only docker containers --- haiku_rag_slim/haiku/rag/chunkers/docling_serve.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/haiku_rag_slim/haiku/rag/chunkers/docling_serve.py b/haiku_rag_slim/haiku/rag/chunkers/docling_serve.py index 3c429a01..e463c955 100644 --- a/haiku_rag_slim/haiku/rag/chunkers/docling_serve.py +++ b/haiku_rag_slim/haiku/rag/chunkers/docling_serve.py @@ -64,6 +64,9 @@ class DoclingServeChunker(DocumentChunker): def _build_chunking_data(self) -> dict[str, str]: """Build form data for chunking request.""" return { + "convert_do_ocr": "false", + "convert_force_ocr": "false", + "convert_ocr_engine": "tesseract", "chunking_max_tokens": str(self.config.processing.chunk_size), "chunking_tokenizer": self.config.processing.chunking_tokenizer, "chunking_merge_peers": str(