News for hackers. Stuff that matters.

A comment on PaddleOCR: Lightweight, 80 Langauge OCR. Parent comment

Re:PaddleOCR: Lightweight, 80 Langauge OCR by julien_c on Friday September 10, 2021 @08:08AM (#28480376)
FWIW this is extractive question answering so the output of the model can only be a span of the input paragraph.

SQuAD is the prototypical example of a dataset for this task, see https://rajpurkar.github.io/SQuAD-explorer/

[ Parent | on HN ]
Re:PaddleOCR: Lightweight, 80 Langauge OCR by hdjjhhvvhga on Friday September 10, 2021 @09:55AM (#28481604)
The problem is it fails even simplest tests (Q: "Adam was not born in 2012. When was Adam born?" A: "2012"; "Mary says Adam was born in 2012. John says Adam was born in 2013. And in fact the latter date is correct." A: "2012").
[ Parent | on HN ]