Datasets:
Tasks:
Text Classification
Sub-tasks:
natural-language-inference
Languages:
English
Size:
10K - 100K
License:
Filippo B
commited on
Commit
•
950e962
1
Parent(s):
bd382a2
Update loader script to use 16 labels
Browse files- README.md +43 -0
- osdg_cd.py +5 -5
README.md
CHANGED
@@ -16,6 +16,49 @@ task_categories:
|
|
16 |
task_ids:
|
17 |
- natural-language-inference
|
18 |
pretty_name: OSDG Community Dataset (OSDG-CD)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
---
|
20 |
# Dataset Card for OSDG-CD
|
21 |
|
|
|
16 |
task_ids:
|
17 |
- natural-language-inference
|
18 |
pretty_name: OSDG Community Dataset (OSDG-CD)
|
19 |
+
dataset_info:
|
20 |
+
config_name: main_config
|
21 |
+
features:
|
22 |
+
- name: doi
|
23 |
+
dtype: string
|
24 |
+
- name: text_id
|
25 |
+
dtype: string
|
26 |
+
- name: text
|
27 |
+
dtype: string
|
28 |
+
- name: sdg
|
29 |
+
dtype: uint16
|
30 |
+
- name: label
|
31 |
+
dtype:
|
32 |
+
class_label:
|
33 |
+
names:
|
34 |
+
'0': SDG 1
|
35 |
+
'1': SDG 2
|
36 |
+
'2': SDG 3
|
37 |
+
'3': SDG 4
|
38 |
+
'4': SDG 5
|
39 |
+
'5': SDG 6
|
40 |
+
'6': SDG 7
|
41 |
+
'7': SDG 8
|
42 |
+
'8': SDG 9
|
43 |
+
'9': SDG 10
|
44 |
+
'10': SDG 11
|
45 |
+
'11': SDG 12
|
46 |
+
'12': SDG 13
|
47 |
+
'13': SDG 14
|
48 |
+
'14': SDG 15
|
49 |
+
'15': SDG 16
|
50 |
+
- name: labels_negative
|
51 |
+
dtype: uint16
|
52 |
+
- name: labels_positive
|
53 |
+
dtype: uint16
|
54 |
+
- name: agreement
|
55 |
+
dtype: float32
|
56 |
+
splits:
|
57 |
+
- name: train
|
58 |
+
num_bytes: 29955216
|
59 |
+
num_examples: 42065
|
60 |
+
download_size: 29574964
|
61 |
+
dataset_size: 29955216
|
62 |
---
|
63 |
# Dataset Card for OSDG-CD
|
64 |
|
osdg_cd.py
CHANGED
@@ -120,7 +120,7 @@ class OSDGCD(datasets.GeneratorBasedBuilder):
|
|
120 |
"text_id": datasets.Value("string"),
|
121 |
"text": datasets.Value("string"),
|
122 |
"sdg": datasets.Value("uint16"),
|
123 |
-
"label": datasets.ClassLabel(num_classes=
|
124 |
"labels_negative": datasets.Value("uint16"),
|
125 |
"labels_positive": datasets.Value("uint16"),
|
126 |
"agreement": datasets.Value("float"),
|
@@ -145,10 +145,10 @@ class OSDGCD(datasets.GeneratorBasedBuilder):
|
|
145 |
def _generate_examples(self, filepath):
|
146 |
"""This function returns the examples in the raw (text) form."""
|
147 |
logger.info("generating examples from = %s", filepath)
|
148 |
-
key = 0
|
149 |
with open(filepath, encoding="utf-8") as f:
|
150 |
osdg = csv.DictReader(f, delimiter="\t")
|
151 |
for row in osdg:
|
152 |
-
|
153 |
-
|
154 |
-
|
|
|
|
120 |
"text_id": datasets.Value("string"),
|
121 |
"text": datasets.Value("string"),
|
122 |
"sdg": datasets.Value("uint16"),
|
123 |
+
"label": datasets.ClassLabel(num_classes=16, names=[f"SDG {sdg}" for sdg in range(1, 17)]),
|
124 |
"labels_negative": datasets.Value("uint16"),
|
125 |
"labels_positive": datasets.Value("uint16"),
|
126 |
"agreement": datasets.Value("float"),
|
|
|
145 |
def _generate_examples(self, filepath):
|
146 |
"""This function returns the examples in the raw (text) form."""
|
147 |
logger.info("generating examples from = %s", filepath)
|
|
|
148 |
with open(filepath, encoding="utf-8") as f:
|
149 |
osdg = csv.DictReader(f, delimiter="\t")
|
150 |
for row in osdg:
|
151 |
+
id_ = row["text_id"]
|
152 |
+
sdg = int(row["sdg"])
|
153 |
+
row["label"] = f"SDG {sdg}"
|
154 |
+
yield id_, row
|