Add Nougat (#25942)
* Add conversion script * Add NougatImageProcessor * Add crop margin * More improvements * Add docs, READMEs * Remove print statements * Include model_max_length * Add NougatTokenizerFast * Fix imports * Improve postprocessing * Improve image processor * Fix image processor * Improve normalize method * More improvements * More improvements * Add processor, improve docs * Simplify fast tokenizer * Remove test file * Fix docstrings * Use NougatProcessor in conversion script * Add is_levensthein_available * Add tokenizer tests * More improvements * Use numpy instead of opencv * Add is_cv2_available * Fix cv2_available * Add is_nltk_available * Add image processor tests, improve crop_margin * Add integration tests * Improve integration test * Use do_rescale instead of hacks, thanks Amy * Remove random_padding * Address comments * Address more comments * Add import * Address more comments * Address more comments * Address comment * Address comment * Set max_model_input_sizes * Add tests * Add requires_backends * Add Nougat to exotic tests * Use to_pil_image * Address comment regarding nltk * Add NLTK * Improve variable names, integration test * Add test * refactor, document, and test regexes * remove named capture groups, add comments * format * add non-markdown fixed tokenization * format * correct flakyness of args parse * add regex comments * test functionalities for crop_image, align long axis and expected output * add regex tests * remove cv2 dependency * test crop_margin equality between cv2 and python * refactor table regexes to markdown add newline * change print to log, improve doc * fix high count tables correction * address PR comments: naming, linting, asserts * Address comments * Add copied from * Update conversion script * Update conversion script to convert both small and base versions * Add inference example * Add more info * Fix style * Add require annotators to test * Define all keyword arguments explicitly * Move cv2 annotator * Add tokenizer init method * Transfer checkpoints * Add reference to Donut * Address comments * Skip test * Remove cv2 method * Add copied from statements * Use cached_property * Fix docstring * Add file to not doctested --------- Co-authored-by: Pablo Montalvo <pablo.montalvo.leroux@gmail.com>
This commit is contained in:
NielsRogge
0
tests/models/nougat/__init__.py
Normal file
0
tests/models/nougat/__init__.py
Normal file
194
tests/models/nougat/test_image_processing_nougat.py
Normal file
194
tests/models/nougat/test_image_processing_nougat.py
Normal file
@@ -0,0 +1,194 @@
|
||||
# coding=utf-8
|
||||
# Copyright 2023 HuggingFace Inc.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
|
||||
import unittest
|
||||
|
||||
import numpy as np
|
||||
from huggingface_hub import hf_hub_download
|
||||
|
||||
from transformers.testing_utils import require_torch, require_vision
|
||||
from transformers.utils import cached_property, is_torch_available, is_vision_available
|
||||
|
||||
from ...test_image_processing_common import ImageProcessingTestMixin, prepare_image_inputs
|
||||
|
||||
|
||||
if is_torch_available():
|
||||
import torch
|
||||
|
||||
if is_vision_available():
|
||||
from PIL import Image
|
||||
|
||||
from transformers import NougatImageProcessor
|
||||
|
||||
|
||||
class NougatImageProcessingTester(unittest.TestCase):
|
||||
def __init__(
|
||||
self,
|
||||
parent,
|
||||
batch_size=7,
|
||||
num_channels=3,
|
||||
image_size=18,
|
||||
min_resolution=30,
|
||||
max_resolution=400,
|
||||
do_crop_margin=True,
|
||||
do_resize=True,
|
||||
size=None,
|
||||
do_thumbnail=True,
|
||||
do_align_long_axis: bool = False,
|
||||
do_pad=True,
|
||||
do_normalize: bool = True,
|
||||
image_mean=[0.5, 0.5, 0.5],
|
||||
image_std=[0.5, 0.5, 0.5],
|
||||
):
|
||||
size = size if size is not None else {"height": 20, "width": 20}
|
||||
self.parent = parent
|
||||
self.batch_size = batch_size
|
||||
self.num_channels = num_channels
|
||||
self.image_size = image_size
|
||||
self.min_resolution = min_resolution
|
||||
self.max_resolution = max_resolution
|
||||
self.do_crop_margin = do_crop_margin
|
||||
self.do_resize = do_resize
|
||||
self.size = size
|
||||
self.do_thumbnail = do_thumbnail
|
||||
self.do_align_long_axis = do_align_long_axis
|
||||
self.do_pad = do_pad
|
||||
self.do_normalize = do_normalize
|
||||
self.image_mean = image_mean
|
||||
self.image_std = image_std
|
||||
|
||||
def prepare_image_processor_dict(self):
|
||||
return {
|
||||
"do_crop_margin": self.do_crop_margin,
|
||||
"do_resize": self.do_resize,
|
||||
"size": self.size,
|
||||
"do_thumbnail": self.do_thumbnail,
|
||||
"do_align_long_axis": self.do_align_long_axis,
|
||||
"do_pad": self.do_pad,
|
||||
"do_normalize": self.do_normalize,
|
||||
"image_mean": self.image_mean,
|
||||
"image_std": self.image_std,
|
||||
}
|
||||
|
||||
def expected_output_image_shape(self, images):
|
||||
return self.num_channels, self.size["height"], self.size["width"]
|
||||
|
||||
def prepare_dummy_image(self):
|
||||
filepath = hf_hub_download(
|
||||
repo_id="hf-internal-testing/fixtures_docvqa", filename="nougat_pdf.png", repo_type="dataset"
|
||||
)
|
||||
image = Image.open(filepath).convert("RGB")
|
||||
return image
|
||||
|
||||
def prepare_image_inputs(self, equal_resolution=False, numpify=False, torchify=False):
|
||||
return prepare_image_inputs(
|
||||
batch_size=self.batch_size,
|
||||
num_channels=self.num_channels,
|
||||
min_resolution=self.min_resolution,
|
||||
max_resolution=self.max_resolution,
|
||||
equal_resolution=equal_resolution,
|
||||
numpify=numpify,
|
||||
torchify=torchify,
|
||||
)
|
||||
|
||||
|
||||
@require_torch
|
||||
@require_vision
|
||||
class NougatImageProcessingTest(ImageProcessingTestMixin, unittest.TestCase):
|
||||
image_processing_class = NougatImageProcessor if is_vision_available() else None
|
||||
|
||||
def setUp(self):
|
||||
self.image_processor_tester = NougatImageProcessingTester(self)
|
||||
|
||||
@property
|
||||
def image_processor_dict(self):
|
||||
return self.image_processor_tester.prepare_image_processor_dict()
|
||||
|
||||
@cached_property
|
||||
def image_processor(self):
|
||||
return self.image_processing_class(**self.image_processor_dict)
|
||||
|
||||
def test_image_processor_properties(self):
|
||||
image_processing = self.image_processing_class(**self.image_processor_dict)
|
||||
self.assertTrue(hasattr(image_processing, "do_resize"))
|
||||
self.assertTrue(hasattr(image_processing, "size"))
|
||||
self.assertTrue(hasattr(image_processing, "do_normalize"))
|
||||
self.assertTrue(hasattr(image_processing, "image_mean"))
|
||||
self.assertTrue(hasattr(image_processing, "image_std"))
|
||||
|
||||
def test_image_processor_from_dict_with_kwargs(self):
|
||||
image_processor = self.image_processing_class.from_dict(self.image_processor_dict)
|
||||
self.assertEqual(image_processor.size, {"height": 20, "width": 20})
|
||||
|
||||
image_processor = self.image_processing_class.from_dict(self.image_processor_dict, size=42)
|
||||
self.assertEqual(image_processor.size, {"height": 42, "width": 42})
|
||||
|
||||
def test_expected_output(self):
|
||||
dummy_image = self.image_processor_tester.prepare_dummy_image()
|
||||
image_processor = self.image_processor
|
||||
inputs = image_processor(dummy_image, return_tensors="pt")
|
||||
self.assertTrue(torch.allclose(inputs["pixel_values"].mean(), torch.tensor(0.4906), atol=1e-3, rtol=1e-3))
|
||||
|
||||
def test_crop_margin_all_white(self):
|
||||
image = np.uint8(np.ones((100, 100, 3)) * 255)
|
||||
image_processor = self.image_processor
|
||||
cropped_image = image_processor.crop_margin(image)
|
||||
self.assertTrue(np.array_equal(image, cropped_image))
|
||||
|
||||
def test_crop_margin_centered_black_square(self):
|
||||
image = np.ones((100, 100, 3), dtype=np.uint8) * 255
|
||||
image[45:55, 45:55, :] = 0
|
||||
image_processor = self.image_processor
|
||||
cropped_image = image_processor.crop_margin(image)
|
||||
expected_cropped = image[45:55, 45:55, :]
|
||||
self.assertTrue(np.array_equal(expected_cropped, cropped_image))
|
||||
|
||||
def test_align_long_axis_no_rotation(self):
|
||||
image = np.uint8(np.ones((100, 200, 3)) * 255)
|
||||
image_processor = self.image_processor
|
||||
size = {"height": 200, "width": 300}
|
||||
aligned_image = image_processor.align_long_axis(image, size)
|
||||
self.assertEqual(image.shape, aligned_image.shape)
|
||||
|
||||
def test_align_long_axis_with_rotation(self):
|
||||
image = np.uint8(np.ones((200, 100, 3)) * 255)
|
||||
image_processor = self.image_processor
|
||||
size = {"height": 300, "width": 200}
|
||||
aligned_image = image_processor.align_long_axis(image, size)
|
||||
self.assertEqual((200, 100, 3), aligned_image.shape)
|
||||
|
||||
def test_align_long_axis_data_format(self):
|
||||
image = np.uint8(np.ones((100, 200, 3)) * 255)
|
||||
data_format = "channels_first"
|
||||
size = {"height": 200, "width": 300}
|
||||
image_processor = self.image_processor
|
||||
aligned_image = image_processor.align_long_axis(image, size, data_format=data_format)
|
||||
self.assertEqual((3, 100, 200), aligned_image.shape)
|
||||
|
||||
def prepare_dummy_np_image(self):
|
||||
filepath = hf_hub_download(
|
||||
repo_id="hf-internal-testing/fixtures_docvqa", filename="nougat_pdf.png", repo_type="dataset"
|
||||
)
|
||||
image = Image.open(filepath).convert("RGB")
|
||||
return np.array(image)
|
||||
|
||||
def test_crop_margin_equality_cv2_python(self):
|
||||
image = self.prepare_dummy_np_image()
|
||||
image_processor = self.image_processor
|
||||
image_cropped_python = image_processor.crop_margin(image)
|
||||
|
||||
self.assertEqual(image_cropped_python.shape, (850, 685, 3))
|
||||
self.assertEqual(image_cropped_python.mean(), 237.43881150708458)
|
||||
196
tests/models/nougat/test_tokenization_nougat.py
Normal file
196
tests/models/nougat/test_tokenization_nougat.py
Normal file
@@ -0,0 +1,196 @@
|
||||
# coding=utf-8
|
||||
# Copyright 2023 The HuggingFace Team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
import unittest
|
||||
|
||||
from transformers import NougatTokenizerFast
|
||||
from transformers.models.nougat.tokenization_nougat_fast import markdown_compatible, normalize_list_like_lines
|
||||
from transformers.testing_utils import require_levenshtein, require_nltk, require_tokenizers
|
||||
|
||||
from ...test_tokenization_common import TokenizerTesterMixin
|
||||
|
||||
|
||||
@require_tokenizers
|
||||
class NougatTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
|
||||
slow_tokenizer_class = None
|
||||
rust_tokenizer_class = NougatTokenizerFast
|
||||
tokenizer_class = NougatTokenizerFast
|
||||
test_rust_tokenizer = True
|
||||
test_slow_tokenizer = False
|
||||
from_pretrained_vocab_key = "tokenizer_file"
|
||||
special_tokens_map = {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
|
||||
|
||||
def setUp(self):
|
||||
super().setUp()
|
||||
tokenizer = NougatTokenizerFast.from_pretrained("facebook/nougat-base")
|
||||
tokenizer.save_pretrained(self.tmpdirname)
|
||||
|
||||
def get_rust_tokenizer(self, **kwargs):
|
||||
kwargs.update(self.special_tokens_map)
|
||||
return NougatTokenizerFast.from_pretrained(self.tmpdirname, **kwargs)
|
||||
|
||||
def test_padding(self, max_length=6):
|
||||
for tokenizer, pretrained_name, kwargs in self.tokenizers_list:
|
||||
with self.subTest(f"{tokenizer.__class__.__name__} ({pretrained_name})"):
|
||||
tokenizer_r = self.rust_tokenizer_class.from_pretrained(pretrained_name, **kwargs)
|
||||
# Simple input
|
||||
sentence1 = "This is a simple input"
|
||||
sentence2 = ["This is a simple input 1", "This is a simple input 2"]
|
||||
pair1 = ("This is a simple input", "This is a pair")
|
||||
pair2 = [
|
||||
("This is a simple input 1", "This is a simple input 2"),
|
||||
("This is a simple pair 1", "This is a simple pair 2"),
|
||||
]
|
||||
|
||||
# Simple input tests
|
||||
try:
|
||||
tokenizer_r.encode(sentence1, max_length=max_length)
|
||||
tokenizer_r.encode_plus(sentence1, max_length=max_length)
|
||||
|
||||
tokenizer_r.batch_encode_plus(sentence2, max_length=max_length)
|
||||
tokenizer_r.encode(pair1, max_length=max_length)
|
||||
tokenizer_r.batch_encode_plus(pair2, max_length=max_length)
|
||||
except ValueError:
|
||||
self.fail("Nougat Tokenizer should be able to deal with padding")
|
||||
|
||||
tokenizer_r.pad_token = None # Hotfixing padding = None
|
||||
self.assertRaises(
|
||||
ValueError, tokenizer_r.encode, sentence1, max_length=max_length, padding="max_length"
|
||||
)
|
||||
|
||||
# Simple input
|
||||
self.assertRaises(
|
||||
ValueError, tokenizer_r.encode_plus, sentence1, max_length=max_length, padding="max_length"
|
||||
)
|
||||
|
||||
# Simple input
|
||||
self.assertRaises(
|
||||
ValueError,
|
||||
tokenizer_r.batch_encode_plus,
|
||||
sentence2,
|
||||
max_length=max_length,
|
||||
padding="max_length",
|
||||
)
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(ValueError, tokenizer_r.encode, pair1, max_length=max_length, padding="max_length")
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(
|
||||
ValueError, tokenizer_r.encode_plus, pair1, max_length=max_length, padding="max_length"
|
||||
)
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(
|
||||
ValueError,
|
||||
tokenizer_r.batch_encode_plus,
|
||||
pair2,
|
||||
max_length=max_length,
|
||||
padding="max_length",
|
||||
)
|
||||
|
||||
@unittest.skip("NougatTokenizerFast does not have tokenizer_file in its signature")
|
||||
def test_rust_tokenizer_signature(self):
|
||||
pass
|
||||
|
||||
@unittest.skip("NougatTokenizerFast does not support pretokenized inputs")
|
||||
def test_pretokenized_inputs(self):
|
||||
pass
|
||||
|
||||
@unittest.skip("NougatTokenizerFast directly inherits from PreTrainedTokenizerFast")
|
||||
def test_prepare_for_model(self):
|
||||
pass
|
||||
|
||||
@unittest.skip("This needs a slow tokenizer. Nougat does not have one!")
|
||||
def test_encode_decode_with_spaces(self):
|
||||
pass
|
||||
|
||||
|
||||
class MarkdownCompatibleTest(unittest.TestCase):
|
||||
def test_equation_tag(self):
|
||||
input_text = "(3.2) \\[Equation Text\\]"
|
||||
excepted_output = "\\[Equation Text \\tag{3.2}\\]"
|
||||
self.assertEqual(markdown_compatible(input_text), excepted_output)
|
||||
|
||||
def test_equation_tag_letters(self):
|
||||
input_text = "(18a) \\[Equation Text\\]"
|
||||
excepted_output = "\\[Equation Text \\tag{18a}\\]"
|
||||
self.assertEqual(markdown_compatible(input_text), excepted_output)
|
||||
|
||||
def test_bold_formatting(self):
|
||||
input_text = r"This is \bm{bold} text."
|
||||
expected_output = r"This is \mathbf{bold} text."
|
||||
self.assertEqual(markdown_compatible(input_text), expected_output)
|
||||
|
||||
def test_url_conversion(self):
|
||||
input_text = "Visit my website at https://www.example.com"
|
||||
expected_output = "Visit my website at [https://www.example.com](https://www.example.com)"
|
||||
self.assertEqual(markdown_compatible(input_text), expected_output)
|
||||
|
||||
def test_algorithm_code_block(self):
|
||||
input_text = "```python\nprint('Hello, world!')\n```"
|
||||
expected_output = "```\npython\nprint('Hello, world!')\n```"
|
||||
self.assertEqual(markdown_compatible(input_text), expected_output)
|
||||
|
||||
def test_escape_characters(self):
|
||||
input_text = r"Escaped characters like \n should not be \\[affected\\]"
|
||||
expected_output = r"Escaped characters like \n should not be \\[affected\\]"
|
||||
self.assertEqual(markdown_compatible(input_text), expected_output)
|
||||
|
||||
def test_nested_tags(self):
|
||||
input_text = r"This is a super nested \bm{\bm{\bm{\bm{\bm{bold}}}}} tag."
|
||||
expected_output = r"This is a super nested \mathbf{\mathbf{\mathbf{\mathbf{\mathbf{bold}}}}} tag."
|
||||
self.assertEqual(markdown_compatible(input_text), expected_output)
|
||||
|
||||
|
||||
class TestNormalizeListLikeLines(unittest.TestCase):
|
||||
def test_two_level_lines(self):
|
||||
input_str = "* Item 1 * Item 2"
|
||||
expected_output = "* Item 1\n* Item 2\n"
|
||||
self.assertEqual(normalize_list_like_lines(input_str), expected_output)
|
||||
|
||||
def test_three_level_lines(self):
|
||||
input_str = "- I. Item 1 - II. Item 2 - III. Item 3"
|
||||
expected_output = "- I. Item 1\n- II. Item 2\n- III. Item 3\n"
|
||||
self.assertEqual(normalize_list_like_lines(input_str), expected_output)
|
||||
|
||||
def test_nested_lines(self):
|
||||
input_str = "- I. Item 1 - I.1 Sub-item 1 - I.1.1 Sub-sub-item 1 - II. Item 2"
|
||||
expected_output = "- I. Item 1\n\t- I.1 Sub-item 1\n\t\t- I.1.1 Sub-sub-item 1\n- II. Item 2\n"
|
||||
self.assertEqual(normalize_list_like_lines(input_str), expected_output)
|
||||
|
||||
|
||||
@require_tokenizers
|
||||
class NougatPostProcessingTest(unittest.TestCase):
|
||||
def setUp(self):
|
||||
super().setUp()
|
||||
self.tokenizer = NougatTokenizerFast.from_pretrained("facebook/nougat-base")
|
||||
|
||||
def test_correct_tables_basic(self):
|
||||
input_str = "\\begin{table} \\begin{tabular}{l l} & \\ \\end{tabular} \\end{table}"
|
||||
expected_output = "\\begin{table}\n\\begin{tabular}{l l} & \\ \\end{tabular}\n\\end{table}"
|
||||
self.assertEqual(self.tokenizer.correct_tables(input_str), expected_output)
|
||||
|
||||
def test_correct_tables_high_count(self):
|
||||
input_str = "\\begin{tabular}" * 20
|
||||
expected_output = ""
|
||||
self.assertEqual(self.tokenizer.correct_tables(input_str), expected_output)
|
||||
|
||||
@require_levenshtein
|
||||
@require_nltk
|
||||
def test_postprocess_as_nougat_no_markdown(self):
|
||||
input_str = "# Nougat: Neural Optical Understanding for Academic Documents\n\n Lukas Blecher\n\nCorrespondence to: lblecher@meta.com\n\nGuillem Cucurull\n\nThomas Scialom\n\nRobert Stojnic\n\nMeta AI\n\nThe paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\n###### Abstract\n\nScientific knowledge is predominantly stored in books and scientific journals, often in the form of PDFs. However, the PDF format leads to a loss of semantic information, particularly for mathematical expressions. We propose Nougat (**N**eural **O**ptical **U**nderstanding for **A**cademic Documents), a Visual Transformer model that performs an _Optical Character Recognition_ (OCR) task for processing scientific documents into a markup language, and demonstrate the effectiveness of our model on a new dataset of scientific documents. The proposed approach offers a promising solution to enhance the accessibility of scientific knowledge in the digital age, by bridging the gap between human-readable documents and machine-readable text. We release the models and code to accelerate future work on scientific text recognition.\n\n## 1 Introduction\n\nThe majority of scientific knowledge is stored in books or published in scientific journals, most commonly in the Portable Document Format (PDF). Next to HTML, PDFs are the second most prominent data format on the internet, making up 2.4% of common crawl [1]. However, the information stored in these files is very difficult to extract into any other formats. This is especially true for highly specialized documents, such as scientific research papers, where the semantic information of mathematical expressions is lost.\n\nExisting Optical Character Recognition (OCR) engines, such as Tesseract OCR [2], excel at detecting and classifying individual characters and words in an image, but fail to understand the relationship between them due to their line-by-line approach. This means that they treat superscripts and subscripts in the same way as the surrounding text, which is a significant drawback for mathematical expressions. In mathematical notations like fractions, exponents, and matrices, relative positions of characters are crucial.\n\nConverting academic research papers into machine-readable text also enables accessibility and searchability of science as a whole. The information of millions of academic papers can not be fully accessed because they are locked behind an unreadable format. Existing corpora, such as the S2ORC dataset [3], capture the text of 12M2 papers using GROBID [4], but are missing meaningful representations of the mathematical equations.\n\nFootnote 2: The paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\nTo this end, we introduce Nougat, a transformer based model that can convert images of document pages to formatted markup text.\n\nThe primary contributions in this paper are\n\n* Release of a pre-trained model capable of converting a PDF to a lightweight markup language. We release the code and the model on GitHub3 Footnote 3: https://github.com/facebookresearch/nougat\n* We introduce a pipeline to create dataset for pairing PDFs to source code\n* Our method is only dependent on the image of a page, allowing access to scanned papers and books" # noqa: E231
|
||||
expected_output = "\n\n# Nougat: Neural Optical Understanding for Academic Documents\n\n Lukas Blecher\n\nCorrespondence to: lblecher@meta.com\n\nGuillem Cucurull\n\nThomas Scialom\n\nRobert Stojnic\n\nMeta AI\n\nThe paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\n###### Abstract\n\nScientific knowledge is predominantly stored in books and scientific journals, often in the form of PDFs. However, the PDF format leads to a loss of semantic information, particularly for mathematical expressions. We propose Nougat (**N**eural **O**ptical **U**nderstanding for **A**cademic Documents), a Visual Transformer model that performs an _Optical Character Recognition_ (OCR) task for processing scientific documents into a markup language, and demonstrate the effectiveness of our model on a new dataset of scientific documents. The proposed approach offers a promising solution to enhance the accessibility of scientific knowledge in the digital age, by bridging the gap between human-readable documents and machine-readable text. We release the models and code to accelerate future work on scientific text recognition.\n\n## 1 Introduction\n\nThe majority of scientific knowledge is stored in books or published in scientific journals, most commonly in the Portable Document Format (PDF). Next to HTML, PDFs are the second most prominent data format on the internet, making up 2.4% of common crawl [1]. However, the information stored in these files is very difficult to extract into any other formats. This is especially true for highly specialized documents, such as scientific research papers, where the semantic information of mathematical expressions is lost.\n\nExisting Optical Character Recognition (OCR) engines, such as Tesseract OCR [2], excel at detecting and classifying individual characters and words in an image, but fail to understand the relationship between them due to their line-by-line approach. This means that they treat superscripts and subscripts in the same way as the surrounding text, which is a significant drawback for mathematical expressions. In mathematical notations like fractions, exponents, and matrices, relative positions of characters are crucial.\n\nConverting academic research papers into machine-readable text also enables accessibility and searchability of science as a whole. The information of millions of academic papers can not be fully accessed because they are locked behind an unreadable format. Existing corpora, such as the S2ORC dataset [3], capture the text of 12M2 papers using GROBID [4], but are missing meaningful representations of the mathematical equations.\n\nFootnote 2: The paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\nTo this end, we introduce Nougat, a transformer based model that can convert images of document pages to formatted markup text.\n\nThe primary contributions in this paper are\n\n* Release of a pre-trained model capable of converting a PDF to a lightweight markup language. We release the code and the model on GitHub3 Footnote 3: https://github.com/facebookresearch/nougat\n* We introduce a pipeline to create dataset for pairing PDFs to source code\n* Our method is only dependent on the image of a page, allowing access to scanned papers and books" # noqa: E231
|
||||
self.assertEqual(self.tokenizer.post_process_single(input_str, fix_markdown=False), expected_output)
|
||||
@@ -18,10 +18,13 @@ import tempfile
|
||||
import unittest
|
||||
|
||||
from datasets import load_dataset
|
||||
from huggingface_hub import hf_hub_download
|
||||
from packaging import version
|
||||
|
||||
from transformers import DonutProcessor, TrOCRProcessor
|
||||
from transformers import DonutProcessor, NougatProcessor, TrOCRProcessor
|
||||
from transformers.testing_utils import (
|
||||
require_levenshtein,
|
||||
require_nltk,
|
||||
require_sentencepiece,
|
||||
require_torch,
|
||||
require_vision,
|
||||
@@ -998,3 +1001,79 @@ class DonutModelIntegrationTest(unittest.TestCase):
|
||||
outputs.scores[0][0, :3], torch.tensor([-17.6490, -4.8381, -15.7577], device=torch_device), atol=1e-4
|
||||
)
|
||||
)
|
||||
|
||||
|
||||
@require_levenshtein
|
||||
@require_nltk
|
||||
@require_torch
|
||||
@require_vision
|
||||
@slow
|
||||
class NougatModelIntegrationTest(unittest.TestCase):
|
||||
@cached_property
|
||||
def default_processor(self):
|
||||
return NougatProcessor.from_pretrained("facebook/nougat-base") if is_vision_available() else None
|
||||
|
||||
@cached_property
|
||||
def default_model(self):
|
||||
return VisionEncoderDecoderModel.from_pretrained("facebook/nougat-base").to(torch_device)
|
||||
|
||||
@cached_property
|
||||
def default_image(self):
|
||||
filepath = hf_hub_download(
|
||||
repo_id="hf-internal-testing/fixtures_docvqa", filename="nougat_pdf.png", repo_type="dataset"
|
||||
)
|
||||
image = Image.open(filepath).convert("RGB")
|
||||
return image
|
||||
|
||||
def test_forward_pass(self):
|
||||
processor = self.default_processor
|
||||
model = self.default_model
|
||||
image = self.default_image
|
||||
pixel_values = processor(images=image, return_tensors="pt").pixel_values.to(torch_device)
|
||||
|
||||
decoder_input_ids = torch.tensor([[0]]).to(torch_device)
|
||||
outputs = model(pixel_values=pixel_values, decoder_input_ids=decoder_input_ids)
|
||||
logits = outputs.logits
|
||||
|
||||
# verify the logits
|
||||
expected_shape = torch.Size((1, 1, model.decoder.config.vocab_size))
|
||||
self.assertEqual(outputs.logits.shape, expected_shape)
|
||||
|
||||
expected_slice = torch.tensor(
|
||||
[1.6253, -4.2179, 5.8532, -2.7911, -5.0609, -4.7397, -4.2890, -5.1073, -4.8908, -4.9729]
|
||||
).to(torch_device)
|
||||
|
||||
self.assertTrue(torch.allclose(logits[0, 0, :10], expected_slice, atol=1e-4))
|
||||
|
||||
def test_generation(self):
|
||||
processor = self.default_processor
|
||||
model = self.default_model
|
||||
image = self.default_image
|
||||
pixel_values = processor(images=image, return_tensors="pt").pixel_values.to(torch_device)
|
||||
|
||||
outputs = model.generate(
|
||||
pixel_values,
|
||||
min_length=1,
|
||||
max_length=3584,
|
||||
bad_words_ids=[[processor.tokenizer.unk_token_id]],
|
||||
return_dict_in_generate=True,
|
||||
output_scores=True,
|
||||
)
|
||||
|
||||
# verify generated sequence
|
||||
generated = processor.batch_decode(outputs.sequences, skip_special_tokens=True)[0]
|
||||
expected_raw_generation = "# Nougat: Neural Optical Understanding for Academic Documents\n\n Lukas Blecher\n\nCorrespondence to: lblecher@meta.com\n\nGuillem Cucurull\n\nThomas Scialom\n\nRobert Stojnic\n\nMeta AI\n\nThe paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\n###### Abstract\n\nScientific knowledge is predominantly stored in books and scientific journals, often in the form of PDFs. However, the PDF format leads to a loss of semantic information, particularly for mathematical expressions. We propose Nougat (**N**eural **O**ptical **U**nderstanding for **A**cademic Documents), a Visual Transformer model that performs an _Optical Character Recognition_ (OCR) task for processing scientific documents into a markup language, and demonstrate the effectiveness of our model on a new dataset of scientific documents. The proposed approach offers a promising solution to enhance the accessibility of scientific knowledge in the digital age, by bridging the gap between human-readable documents and machine-readable text. We release the models and code to accelerate future work on scientific text recognition.\n\n## 1 Introduction\n\nThe majority of scientific knowledge is stored in books or published in scientific journals, most commonly in the Portable Document Format (PDF). Next to HTML, PDFs are the second most prominent data format on the internet, making up 2.4% of common crawl [1]. However, the information stored in these files is very difficult to extract into any other formats. This is especially true for highly specialized documents, such as scientific research papers, where the semantic information of mathematical expressions is lost.\n\nExisting Optical Character Recognition (OCR) engines, such as Tesseract OCR [2], excel at detecting and classifying individual characters and words in an image, but fail to understand the relationship between them due to their line-by-line approach. This means that they treat superscripts and subscripts in the same way as the surrounding text, which is a significant drawback for mathematical expressions. In mathematical notations like fractions, exponents, and matrices, relative positions of characters are crucial.\n\nConverting academic research papers into machine-readable text also enables accessibility and searchability of science as a whole. The information of millions of academic papers can not be fully accessed because they are locked behind an unreadable format. Existing corpora, such as the S2ORC dataset [3], capture the text of 12M2 papers using GROBID [4], but are missing meaningful representations of the mathematical equations.\n\nFootnote 2: The paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\nTo this end, we introduce Nougat, a transformer based model that can convert images of document pages to formatted markup text.\n\nThe primary contributions in this paper are\n\n* Release of a pre-trained model capable of converting a PDF to a lightweight markup language. We release the code and the model on GitHub3 Footnote 3: https://github.com/facebookresearch/nougat\n* We introduce a pipeline to create dataset for pairing PDFs to source code\n* Our method is only dependent on the image of a page, allowing access to scanned papers and books"
|
||||
self.assertTrue(generated == expected_raw_generation)
|
||||
|
||||
# verify postprocessed sequence
|
||||
generated = processor.post_process_generation(generated, fix_markdown=False)
|
||||
expected_generation = "\n\n# Nougat: Neural Optical Understanding for Academic Documents\n\n Lukas Blecher\n\nCorrespondence to: lblecher@meta.com\n\nGuillem Cucurull\n\nThomas Scialom\n\nRobert Stojnic\n\nMeta AI\n\nThe paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\n###### Abstract\n\nScientific knowledge is predominantly stored in books and scientific journals, often in the form of PDFs. However, the PDF format leads to a loss of semantic information, particularly for mathematical expressions. We propose Nougat (**N**eural **O**ptical **U**nderstanding for **A**cademic Documents), a Visual Transformer model that performs an _Optical Character Recognition_ (OCR) task for processing scientific documents into a markup language, and demonstrate the effectiveness of our model on a new dataset of scientific documents. The proposed approach offers a promising solution to enhance the accessibility of scientific knowledge in the digital age, by bridging the gap between human-readable documents and machine-readable text. We release the models and code to accelerate future work on scientific text recognition.\n\n## 1 Introduction\n\nThe majority of scientific knowledge is stored in books or published in scientific journals, most commonly in the Portable Document Format (PDF). Next to HTML, PDFs are the second most prominent data format on the internet, making up 2.4% of common crawl [1]. However, the information stored in these files is very difficult to extract into any other formats. This is especially true for highly specialized documents, such as scientific research papers, where the semantic information of mathematical expressions is lost.\n\nExisting Optical Character Recognition (OCR) engines, such as Tesseract OCR [2], excel at detecting and classifying individual characters and words in an image, but fail to understand the relationship between them due to their line-by-line approach. This means that they treat superscripts and subscripts in the same way as the surrounding text, which is a significant drawback for mathematical expressions. In mathematical notations like fractions, exponents, and matrices, relative positions of characters are crucial.\n\nConverting academic research papers into machine-readable text also enables accessibility and searchability of science as a whole. The information of millions of academic papers can not be fully accessed because they are locked behind an unreadable format. Existing corpora, such as the S2ORC dataset [3], capture the text of 12M2 papers using GROBID [4], but are missing meaningful representations of the mathematical equations.\n\nFootnote 2: The paper reports 8.1M papers but the authors recently updated the numbers on the GitHub page https://github.com/allenai/s2orc\n\nTo this end, we introduce Nougat, a transformer based model that can convert images of document pages to formatted markup text.\n\nThe primary contributions in this paper are\n\n* Release of a pre-trained model capable of converting a PDF to a lightweight markup language. We release the code and the model on GitHub3 Footnote 3: https://github.com/facebookresearch/nougat\n* We introduce a pipeline to create dataset for pairing PDFs to source code\n* Our method is only dependent on the image of a page, allowing access to scanned papers and books"
|
||||
self.assertTrue(generated == expected_generation)
|
||||
|
||||
# verify scores
|
||||
self.assertEqual(len(outputs.scores), 741)
|
||||
self.assertTrue(
|
||||
torch.allclose(
|
||||
outputs.scores[0][0, :3], torch.tensor([1.6253, -4.2179, 5.8532], device=torch_device), atol=1e-4
|
||||
)
|
||||
)
|
||||
|
||||
Reference in New Issue
Block a user