From f4fc6cfd4a1a40fe163f07d890e31d1de4ebb7b0 Mon Sep 17 00:00:00 2001 From: Peter Staar Date: Thu, 14 Nov 2024 07:45:36 +0100 Subject: [PATCH] added TableFormerMode.ACCURATE as default in cli Signed-off-by: Peter Staar --- docling/cli/main.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/docling/cli/main.py b/docling/cli/main.py index a1089a8a..db7bc4c9 100644 --- a/docling/cli/main.py +++ b/docling/cli/main.py @@ -24,6 +24,7 @@ from docling.datamodel.pipeline_options import ( PdfPipelineOptions, TesseractCliOcrOptions, TesseractOcrOptions, + TableFormerMode, ) from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption @@ -225,7 +226,8 @@ def convert( do_table_structure=True, ) pipeline_options.table_structure_options.do_cell_matching = True # do_cell_matching - + pipeline_options.table_structure_options.mode = TableFormerMode.ACCURATE + format_options: Dict[InputFormat, FormatOption] = { InputFormat.PDF: PdfFormatOption( pipeline_options=pipeline_options,