DocumentExtractor - Invalid buffer size

All features of docking are working fine except the new feature DocumentExtractor. There is no issue with the file,
its the same file used in the example - https://github.com/docling-project/docling/blob/main/tests/data_scanned/qr_bill_example.jpg

I've tested on 2 Mac M1 machines and google Colab. Default config, should be using 2B param model, so memory shouldn't be an issue.

from docling.datamodel.base_models import InputFormat
from docling.document_extractor import DocumentExtractor

dict_templ = {
"bill_no": "string",
"total": "number",
}
file_path = "qr_bill_example.jpg"

extractor = DocumentExtractor(allowed_formats=[InputFormat.IMAGE, InputFormat.PDF])
result = extractor.extract(file_path, template=dict_templ)
print(result)