From 667af5ae46c9a09ec4563d8fb0263c8c54a20493 Mon Sep 17 00:00:00 2001
From: Ankur Goyal <ankur@impira.com>
Date: Sun, 7 Aug 2022 18:37:40 -0700
Subject: [PATCH] Bug fixes + Copy formatting from transformers repo

---
 README.md                               |  6 ++++++
 pipeline_document_question_answering.py |  6 ++++--
 pyproject.toml                          |  3 +++
 qa_helpers.py                           |  5 ++++-
 setup.cfg                               | 18 ++++++++++++++++++
 5 files changed, 35 insertions(+), 3 deletions(-)
 create mode 100644 pyproject.toml
 create mode 100644 setup.cfg

diff --git a/README.md b/README.md
index 4666008..7687f38 100644
--- a/README.md
+++ b/README.md
@@ -13,3 +13,9 @@ This is a fine-tuned version of the multi-modal [LayoutLM](https://aka.ms/layout
 The LayoutLM model was developed at Microsoft ([paper](https://arxiv.org/abs/1912.13318)) as a general purpose tool for understanding documents. This model is a fine-tuned checkpoint of [LayoutLM-Base-Cased](https://huggingface.co/microsoft/layoutlm-base-uncased), using both the [SQuAD2.0](https://huggingface.co/datasets/squad_v2) and [DocVQA](https://www.docvqa.org/) datasets.
 
 ## Getting started with the model
+
+
+
+## About us
+
+This model was created by the team at [Impira](https://www.impira.com/).
diff --git a/pipeline_document_question_answering.py b/pipeline_document_question_answering.py
index 02f6684..6048eeb 100644
--- a/pipeline_document_question_answering.py
+++ b/pipeline_document_question_answering.py
@@ -1,4 +1,4 @@
-# NOTE: This code is currently under review for inclusion in the main 
+# NOTE: This code is currently under review for inclusion in the main
 # huggingface/transformers repository:
 # https://github.com/huggingface/transformers/pull/18414
 from typing import List, Optional, Tuple, Union
@@ -189,7 +189,9 @@ class DocumentQuestionAnsweringPipeline(Pipeline):
             - **answer** (`str`) -- The answer to the question.
         """
         if isinstance(question, str):
-            inputs = {"question": question, "image": image, "word_boxes": word_boxes}
+            inputs = {"question": question, "image": image}
+            if word_boxes is not None:
+                inputs["word_boxes"] = word_boxes
         else:
             inputs = image
         return super().__call__(inputs, **kwargs)
diff --git a/pyproject.toml b/pyproject.toml
new file mode 100644
index 0000000..291558c
--- /dev/null
+++ b/pyproject.toml
@@ -0,0 +1,3 @@
+[tool.black]
+line-length = 119
+target-version = ['py35']
diff --git a/qa_helpers.py b/qa_helpers.py
index 5f47f37..2b5c675 100644
--- a/qa_helpers.py
+++ b/qa_helpers.py
@@ -1,4 +1,4 @@
-# NOTE: This code is currently under review for inclusion in the main 
+# NOTE: This code is currently under review for inclusion in the main
 # huggingface/transformers repository:
 # https://github.com/huggingface/transformers/pull/18414
 
@@ -15,6 +15,7 @@ if is_vision_available():
     from PIL import Image
 
     from transformers.image_utils import load_image
+
     VISION_LOADED = True
 else:
     Image = None
@@ -24,10 +25,12 @@ else:
 TESSERACT_LOADED = False
 if is_pytesseract_available():
     import pytesseract
+
     TESSERACT_LOADED = True
 else:
     pytesseract = None
 
+
 def decode_spans(
     start: np.ndarray, end: np.ndarray, topk: int, max_answer_len: int, undesired_tokens: np.ndarray
 ) -> Tuple:
diff --git a/setup.cfg b/setup.cfg
new file mode 100644
index 0000000..cc02383
--- /dev/null
+++ b/setup.cfg
@@ -0,0 +1,18 @@
+[isort]
+default_section = FIRSTPARTY
+ensure_newline_before_comments = True
+force_grid_wrap = 0
+include_trailing_comma = True
+known_first_party = transformers
+
+line_length = 119
+lines_after_imports = 2
+multi_line_output = 3
+use_parentheses = True
+
+[flake8]
+ignore = E203, E501, E741, W503, W605
+max-line-length = 119
+
+[tool:pytest]
+doctest_optionflags=NUMBER NORMALIZE_WHITESPACE ELLIPSIS