Datasets:
All tasks working
Browse files- sd-nlp-non-tokenized.py +28 -19
sd-nlp-non-tokenized.py
CHANGED
@@ -19,9 +19,7 @@
|
|
19 |
from __future__ import absolute_import, division, print_function
|
20 |
|
21 |
import json
|
22 |
-
import pdb
|
23 |
import datasets
|
24 |
-
import os
|
25 |
|
26 |
_BASE_URL = "https://huggingface.co/datasets/EMBO/sd-nlp-non-tokenized/resolve/main/"
|
27 |
|
@@ -71,8 +69,6 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
71 |
|
72 |
_URLS = {
|
73 |
"NER": f"{_BASE_URL}sd_panels_general_tokenization.zip",
|
74 |
-
"ROLES": f"{_BASE_URL}sd_panels_general_tokenization.zip",
|
75 |
-
"BORING": f"{_BASE_URL}sd_panels_general_tokenization.zip",
|
76 |
"PANELIZATION": f"{_BASE_URL}sd_fig_general_tokenization.zip",
|
77 |
}
|
78 |
BUILDER_CONFIGS = [
|
@@ -95,49 +91,57 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
95 |
{
|
96 |
"words": datasets.Sequence(feature=datasets.Value("string")),
|
97 |
"labels": datasets.Sequence(
|
98 |
-
feature=datasets.ClassLabel(num_classes=len(self._NER_LABEL_NAMES),
|
|
|
99 |
),
|
|
|
100 |
}
|
101 |
)
|
102 |
elif self.config.name == "GENEPROD_ROLES":
|
103 |
features = datasets.Features(
|
104 |
{
|
105 |
-
"
|
106 |
"labels": datasets.Sequence(
|
107 |
feature=datasets.ClassLabel(
|
108 |
-
num_classes=len(self._SEMANTIC_GENEPROD_ROLES_LABEL_NAMES),
|
|
|
109 |
)
|
110 |
),
|
111 |
-
|
112 |
}
|
113 |
)
|
114 |
elif self.config.name == "SMALL_MOL_ROLES":
|
115 |
features = datasets.Features(
|
116 |
{
|
117 |
-
"
|
118 |
"labels": datasets.Sequence(
|
119 |
feature=datasets.ClassLabel(
|
120 |
-
num_classes=len(self._SEMANTIC_SMALL_MOL_ROLES_LABEL_NAMES),
|
|
|
121 |
)
|
122 |
),
|
|
|
123 |
}
|
124 |
)
|
125 |
elif self.config.name == "BORING":
|
126 |
features = datasets.Features(
|
127 |
{
|
128 |
-
"
|
129 |
"labels": datasets.Sequence(
|
130 |
-
feature=datasets.ClassLabel(num_classes=len(self._BORING_LABEL_NAMES),
|
|
|
131 |
),
|
132 |
}
|
133 |
)
|
134 |
elif self.config.name == "PANELIZATION":
|
135 |
features = datasets.Features(
|
136 |
{
|
137 |
-
"
|
138 |
"labels": datasets.Sequence(
|
139 |
-
feature=datasets.ClassLabel(num_classes=len(self._PANEL_START_NAMES),
|
|
|
140 |
),
|
|
|
141 |
}
|
142 |
)
|
143 |
|
@@ -153,11 +157,14 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
153 |
def _split_generators(self, dl_manager: datasets.DownloadManager):
|
154 |
"""Returns SplitGenerators.
|
155 |
Uses local files if a data_dir is specified. Otherwise downloads the files from their official url."""
|
156 |
-
|
157 |
-
data_dir = dl_manager.download_and_extract(url)
|
158 |
if self.config.name in ["NER", "GENEPROD_ROLES", "SMALL_MOL_ROLES", "BORING"]:
|
|
|
|
|
159 |
data_dir += "/sd_panels_general_tokenization"
|
160 |
elif self.config.name == "PANELIZATION":
|
|
|
|
|
161 |
data_dir += "/sd_fig_general_tokenization"
|
162 |
else:
|
163 |
raise ValueError(f"unkonwn config name: {self.config.name}")
|
@@ -185,6 +192,7 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
185 |
"""Yields examples. This method will receive as arguments the `gen_kwargs` defined in the previous `_split_generators` method.
|
186 |
It is in charge of opening the given file and yielding (key, example) tuples from the dataset
|
187 |
The key is not important, it's more here for legacy reason (legacy from tfds)"""
|
|
|
188 |
|
189 |
with open(filepath, encoding="utf-8") as f:
|
190 |
# logger.info("⏳ Generating examples from = %s", filepath)
|
@@ -208,7 +216,7 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
208 |
"tag_mask": tag_mask,
|
209 |
}
|
210 |
elif self.config.name == "SMALL_MOL_ROLES":
|
211 |
-
labels = data["
|
212 |
small_mol = ["B-SMALL_MOLECULE", "I-SMALL_MOLECULE"]
|
213 |
tag_mask = [1 if t in small_mol else 0 for t in labels]
|
214 |
yield id_, {
|
@@ -217,12 +225,13 @@ class SourceDataNLP(datasets.GeneratorBasedBuilder):
|
|
217 |
"tag_mask": tag_mask,
|
218 |
}
|
219 |
elif self.config.name == "BORING":
|
220 |
-
yield id_, {"words": data["words"],
|
|
|
221 |
elif self.config.name == "PANELIZATION":
|
222 |
labels = data["label_ids"]["panel_start"]
|
223 |
tag_mask = [1 if t == "B-PANEL_START" else 0 for t in labels]
|
224 |
yield id_, {
|
225 |
-
"
|
226 |
"labels": data["label_ids"]["panel_start"],
|
227 |
"tag_mask": tag_mask,
|
228 |
}
|
|
|
19 |
from __future__ import absolute_import, division, print_function
|
20 |
|
21 |
import json
|
|
|
22 |
import datasets
|
|
|
23 |
|
24 |
_BASE_URL = "https://huggingface.co/datasets/EMBO/sd-nlp-non-tokenized/resolve/main/"
|
25 |
|
|
|
69 |
|
70 |
_URLS = {
|
71 |
"NER": f"{_BASE_URL}sd_panels_general_tokenization.zip",
|
|
|
|
|
72 |
"PANELIZATION": f"{_BASE_URL}sd_fig_general_tokenization.zip",
|
73 |
}
|
74 |
BUILDER_CONFIGS = [
|
|
|
91 |
{
|
92 |
"words": datasets.Sequence(feature=datasets.Value("string")),
|
93 |
"labels": datasets.Sequence(
|
94 |
+
feature=datasets.ClassLabel(num_classes=len(self._NER_LABEL_NAMES),
|
95 |
+
names=self._NER_LABEL_NAMES)
|
96 |
),
|
97 |
+
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
|
98 |
}
|
99 |
)
|
100 |
elif self.config.name == "GENEPROD_ROLES":
|
101 |
features = datasets.Features(
|
102 |
{
|
103 |
+
"words": datasets.Sequence(feature=datasets.Value("string")),
|
104 |
"labels": datasets.Sequence(
|
105 |
feature=datasets.ClassLabel(
|
106 |
+
num_classes=len(self._SEMANTIC_GENEPROD_ROLES_LABEL_NAMES),
|
107 |
+
names=self._SEMANTIC_GENEPROD_ROLES_LABEL_NAMES
|
108 |
)
|
109 |
),
|
110 |
+
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
|
111 |
}
|
112 |
)
|
113 |
elif self.config.name == "SMALL_MOL_ROLES":
|
114 |
features = datasets.Features(
|
115 |
{
|
116 |
+
"words": datasets.Sequence(feature=datasets.Value("string")),
|
117 |
"labels": datasets.Sequence(
|
118 |
feature=datasets.ClassLabel(
|
119 |
+
num_classes=len(self._SEMANTIC_SMALL_MOL_ROLES_LABEL_NAMES),
|
120 |
+
names=self._SEMANTIC_SMALL_MOL_ROLES_LABEL_NAMES
|
121 |
)
|
122 |
),
|
123 |
+
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
|
124 |
}
|
125 |
)
|
126 |
elif self.config.name == "BORING":
|
127 |
features = datasets.Features(
|
128 |
{
|
129 |
+
"words": datasets.Sequence(feature=datasets.Value("string")),
|
130 |
"labels": datasets.Sequence(
|
131 |
+
feature=datasets.ClassLabel(num_classes=len(self._BORING_LABEL_NAMES),
|
132 |
+
names=self._BORING_LABEL_NAMES)
|
133 |
),
|
134 |
}
|
135 |
)
|
136 |
elif self.config.name == "PANELIZATION":
|
137 |
features = datasets.Features(
|
138 |
{
|
139 |
+
"words": datasets.Sequence(feature=datasets.Value("string")),
|
140 |
"labels": datasets.Sequence(
|
141 |
+
feature=datasets.ClassLabel(num_classes=len(self._PANEL_START_NAMES),
|
142 |
+
names=self._PANEL_START_NAMES)
|
143 |
),
|
144 |
+
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
|
145 |
}
|
146 |
)
|
147 |
|
|
|
157 |
def _split_generators(self, dl_manager: datasets.DownloadManager):
|
158 |
"""Returns SplitGenerators.
|
159 |
Uses local files if a data_dir is specified. Otherwise downloads the files from their official url."""
|
160 |
+
|
|
|
161 |
if self.config.name in ["NER", "GENEPROD_ROLES", "SMALL_MOL_ROLES", "BORING"]:
|
162 |
+
url = self._URLS["NER"]
|
163 |
+
data_dir = dl_manager.download_and_extract(url)
|
164 |
data_dir += "/sd_panels_general_tokenization"
|
165 |
elif self.config.name == "PANELIZATION":
|
166 |
+
url = self._URLS[self.config.name]
|
167 |
+
data_dir = dl_manager.download_and_extract(url)
|
168 |
data_dir += "/sd_fig_general_tokenization"
|
169 |
else:
|
170 |
raise ValueError(f"unkonwn config name: {self.config.name}")
|
|
|
192 |
"""Yields examples. This method will receive as arguments the `gen_kwargs` defined in the previous `_split_generators` method.
|
193 |
It is in charge of opening the given file and yielding (key, example) tuples from the dataset
|
194 |
The key is not important, it's more here for legacy reason (legacy from tfds)"""
|
195 |
+
print(" This line is taking place")
|
196 |
|
197 |
with open(filepath, encoding="utf-8") as f:
|
198 |
# logger.info("⏳ Generating examples from = %s", filepath)
|
|
|
216 |
"tag_mask": tag_mask,
|
217 |
}
|
218 |
elif self.config.name == "SMALL_MOL_ROLES":
|
219 |
+
labels = data["label_ids"]["small_mol_roles"]
|
220 |
small_mol = ["B-SMALL_MOLECULE", "I-SMALL_MOLECULE"]
|
221 |
tag_mask = [1 if t in small_mol else 0 for t in labels]
|
222 |
yield id_, {
|
|
|
225 |
"tag_mask": tag_mask,
|
226 |
}
|
227 |
elif self.config.name == "BORING":
|
228 |
+
yield id_, {"words": data["words"],
|
229 |
+
"labels": data["label_ids"]["boring"]}
|
230 |
elif self.config.name == "PANELIZATION":
|
231 |
labels = data["label_ids"]["panel_start"]
|
232 |
tag_mask = [1 if t == "B-PANEL_START" else 0 for t in labels]
|
233 |
yield id_, {
|
234 |
+
"words": data["words"],
|
235 |
"labels": data["label_ids"]["panel_start"],
|
236 |
"tag_mask": tag_mask,
|
237 |
}
|