Made kwarg dependent builder configs for loading DUDE, in next step could make nonbinary_to_imdb mode
Browse files- DUDE_loader.py +39 -26
- test_loader.py +1 -1
DUDE_loader.py
CHANGED
@@ -17,11 +17,10 @@
|
|
17 |
import os
|
18 |
import copy
|
19 |
import json
|
20 |
-
from typing import List
|
21 |
import pdf2image
|
22 |
from tqdm import tqdm
|
23 |
|
24 |
-
|
25 |
import datasets
|
26 |
|
27 |
|
@@ -113,7 +112,9 @@ class DUDEConfig(datasets.BuilderConfig):
|
|
113 |
|
114 |
def __init__(
|
115 |
self,
|
116 |
-
binary_mode: bool,
|
|
|
|
|
117 |
**kwargs,
|
118 |
):
|
119 |
"""BuilderConfig for DUDE.
|
@@ -121,9 +122,28 @@ class DUDEConfig(datasets.BuilderConfig):
|
|
121 |
binary_mode: `boolean`, load binary PDFs/OCR or pass along paths on local file system
|
122 |
**kwargs: keyword arguments forwarded to super.
|
123 |
"""
|
124 |
-
BINARY_MODE = False
|
125 |
super(DUDEConfig, self).__init__(description=_DESCRIPTION, **kwargs)
|
126 |
-
self.binary_mode = binary_mode
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
127 |
|
128 |
|
129 |
class DUDE(datasets.GeneratorBasedBuilder):
|
@@ -131,13 +151,11 @@ class DUDE(datasets.GeneratorBasedBuilder):
|
|
131 |
|
132 |
VERSION = datasets.Version("0.0.1")
|
133 |
|
134 |
-
BUILDER_CONFIGS =
|
135 |
-
DUDEConfig(name='DUDE', version=VERSION, binary_mode=False),
|
136 |
-
DUDEConfig(name='DUDE-binary', version=VERSION, binary_mode=True)
|
137 |
-
]
|
138 |
-
|
139 |
-
DEFAULT_CONFIG_NAME = "DUDE" #for some reason not working
|
140 |
|
|
|
|
|
|
|
141 |
|
142 |
def _info(self):
|
143 |
features = datasets.Features(
|
@@ -158,8 +176,12 @@ class DUDE(datasets.GeneratorBasedBuilder):
|
|
158 |
"answers_variants": datasets.Sequence(datasets.Value("string")),
|
159 |
"answer_type": datasets.Value("string"),
|
160 |
"data_split": datasets.Value("string"),
|
161 |
-
"document": datasets.Value("binary")
|
162 |
-
|
|
|
|
|
|
|
|
|
163 |
}
|
164 |
)
|
165 |
|
@@ -179,19 +201,14 @@ class DUDE(datasets.GeneratorBasedBuilder):
|
|
179 |
annotations = json.load(open(_URLS[f"annotations"], "r"))
|
180 |
# binaries_archive = dl_manager.iter_archive(binaries_path)
|
181 |
|
182 |
-
if self.config.data_dir:
|
183 |
binary_extraction_path = self.config.data_dir
|
184 |
else:
|
185 |
-
binaries_path = dl_manager.download(_URLS[
|
186 |
binary_extraction_path = dl_manager.extract(binaries_path)
|
187 |
-
binary_extraction_path += "/home/jordy/Downloads/" + _URLS["binaries"].split("/")[
|
188 |
-
-1
|
189 |
-
].replace(
|
190 |
-
".tar.gz", ""
|
191 |
-
) # weird unpacking behaviour
|
192 |
|
193 |
splits = []
|
194 |
-
for split in _SPLITS:
|
195 |
splits.append(
|
196 |
datasets.SplitGenerator(
|
197 |
name=split,
|
@@ -209,16 +226,12 @@ class DUDE(datasets.GeneratorBasedBuilder):
|
|
209 |
extracted_path = os.path.join(binary_extraction_path, "PDF", split, docid + ".pdf")
|
210 |
return extracted_path
|
211 |
|
212 |
-
|
213 |
def retrieve_OCR(docid, ocr_engine="Amazon", format="original"):
|
214 |
extracted_path = os.path.join(
|
215 |
binary_extraction_path, "OCR", ocr_engine, docid + f"_{format}.json"
|
216 |
)
|
217 |
return extracted_path
|
218 |
|
219 |
-
question = self.info.features["question"]
|
220 |
-
answers = self.info.features["answers"]
|
221 |
-
|
222 |
annotations = [x for x in annotations if x["data_split"] == split]
|
223 |
|
224 |
for i, a in enumerate(annotations):
|
@@ -235,5 +248,5 @@ class DUDE(datasets.GeneratorBasedBuilder):
|
|
235 |
a["OCR"] = g.read()
|
236 |
else:
|
237 |
a["document"] = docpath
|
238 |
-
a["OCR"] = ocrpath
|
239 |
yield i, a
|
|
|
17 |
import os
|
18 |
import copy
|
19 |
import json
|
20 |
+
from typing import List, Literal
|
21 |
import pdf2image
|
22 |
from tqdm import tqdm
|
23 |
|
|
|
24 |
import datasets
|
25 |
|
26 |
|
|
|
112 |
|
113 |
def __init__(
|
114 |
self,
|
115 |
+
binary_mode: bool = False,
|
116 |
+
ocr_engine: Literal["Azure", "Amazon", "Tesseract"] = "Amazon",
|
117 |
+
format: Literal["original", "due"] = "original",
|
118 |
**kwargs,
|
119 |
):
|
120 |
"""BuilderConfig for DUDE.
|
|
|
122 |
binary_mode: `boolean`, load binary PDFs/OCR or pass along paths on local file system
|
123 |
**kwargs: keyword arguments forwarded to super.
|
124 |
"""
|
|
|
125 |
super(DUDEConfig, self).__init__(description=_DESCRIPTION, **kwargs)
|
126 |
+
self.binary_mode = binary_mode
|
127 |
+
self.ocr_engine = ocr_engine
|
128 |
+
self.format = format
|
129 |
+
|
130 |
+
|
131 |
+
def builder_configs(version):
|
132 |
+
configurations = []
|
133 |
+
for binary_mode in [True, False]:
|
134 |
+
for ocr_engine in ["Azure", "Amazon", "Tesseract"]:
|
135 |
+
for format in ["original", "due"]:
|
136 |
+
binary_name = "bin_" if binary_mode else ""
|
137 |
+
configurations.append(
|
138 |
+
DUDEConfig(
|
139 |
+
name=f"{binary_name}{ocr_engine}_{format}",
|
140 |
+
version=version,
|
141 |
+
binary_mode=binary_mode,
|
142 |
+
ocr_engine=ocr_engine,
|
143 |
+
format=format,
|
144 |
+
)
|
145 |
+
)
|
146 |
+
return configurations
|
147 |
|
148 |
|
149 |
class DUDE(datasets.GeneratorBasedBuilder):
|
|
|
151 |
|
152 |
VERSION = datasets.Version("0.0.1")
|
153 |
|
154 |
+
BUILDER_CONFIGS = builder_configs(VERSION)
|
|
|
|
|
|
|
|
|
|
|
155 |
|
156 |
+
DEFAULT_CONFIG_NAME = (
|
157 |
+
"Amazon_original" # for some reason not working, need to pass a config anyway
|
158 |
+
)
|
159 |
|
160 |
def _info(self):
|
161 |
features = datasets.Features(
|
|
|
176 |
"answers_variants": datasets.Sequence(datasets.Value("string")),
|
177 |
"answer_type": datasets.Value("string"),
|
178 |
"data_split": datasets.Value("string"),
|
179 |
+
"document": datasets.Value("binary")
|
180 |
+
if self.config.binary_mode
|
181 |
+
else datasets.Value("string"),
|
182 |
+
"OCR": datasets.Value("binary")
|
183 |
+
if self.config.binary_mode
|
184 |
+
else datasets.Value("string"),
|
185 |
}
|
186 |
)
|
187 |
|
|
|
201 |
annotations = json.load(open(_URLS[f"annotations"], "r"))
|
202 |
# binaries_archive = dl_manager.iter_archive(binaries_path)
|
203 |
|
204 |
+
if self.config.data_dir: # unpacked it to a custom directory
|
205 |
binary_extraction_path = self.config.data_dir
|
206 |
else:
|
207 |
+
binaries_path = dl_manager.download(_URLS["binaries"])
|
208 |
binary_extraction_path = dl_manager.extract(binaries_path)
|
|
|
|
|
|
|
|
|
|
|
209 |
|
210 |
splits = []
|
211 |
+
for split in _SPLITS:
|
212 |
splits.append(
|
213 |
datasets.SplitGenerator(
|
214 |
name=split,
|
|
|
226 |
extracted_path = os.path.join(binary_extraction_path, "PDF", split, docid + ".pdf")
|
227 |
return extracted_path
|
228 |
|
|
|
229 |
def retrieve_OCR(docid, ocr_engine="Amazon", format="original"):
|
230 |
extracted_path = os.path.join(
|
231 |
binary_extraction_path, "OCR", ocr_engine, docid + f"_{format}.json"
|
232 |
)
|
233 |
return extracted_path
|
234 |
|
|
|
|
|
|
|
235 |
annotations = [x for x in annotations if x["data_split"] == split]
|
236 |
|
237 |
for i, a in enumerate(annotations):
|
|
|
248 |
a["OCR"] = g.read()
|
249 |
else:
|
250 |
a["document"] = docpath
|
251 |
+
a["OCR"] = ocrpath
|
252 |
yield i, a
|
test_loader.py
CHANGED
@@ -22,7 +22,7 @@ from codetiming import Timer
|
|
22 |
for binding in ["dict_annotations (new)"]: #"dict_PDF",
|
23 |
with Timer(name=f"{binding}", text=binding + " Elapsed time: {:.4f} seconds"):
|
24 |
if binding == "dict_annotations (new)":
|
25 |
-
ds = load_dataset("../DUDE_loader/DUDE_loader.py", '
|
26 |
else:
|
27 |
ds = load_dataset("jordyvl/DUDE_loader", revision='db20bbf751b14e14e8143170bc201948ef5ac83c')
|
28 |
|
|
|
22 |
for binding in ["dict_annotations (new)"]: #"dict_PDF",
|
23 |
with Timer(name=f"{binding}", text=binding + " Elapsed time: {:.4f} seconds"):
|
24 |
if binding == "dict_annotations (new)":
|
25 |
+
ds = load_dataset("../DUDE_loader/DUDE_loader.py", 'bin_Amazon_original', data_dir="/home/jordy/Downloads/DUDE_train-val-test_binaries", ocr_engine='Azure') #ignore_verifications=True, , writer_batch_size=10
|
26 |
else:
|
27 |
ds = load_dataset("jordyvl/DUDE_loader", revision='db20bbf751b14e14e8143170bc201948ef5ac83c')
|
28 |
|