Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
sentiment-classification
Languages:
Turkish
Size:
< 1K
License:
Update data.py
Browse files
data.py
CHANGED
@@ -46,8 +46,11 @@ _TRAIN_DOWNLOAD_URL = (
|
|
46 |
_TEST_DOWNLOAD_URL = "https://github.com/PnrSvc/dataset/blob/main/turkish/test.csv"
|
47 |
|
48 |
|
49 |
-
|
50 |
-
"""
|
|
|
|
|
|
|
51 |
|
52 |
VERSION = datasets.Version("1.1.0")
|
53 |
|
|
|
46 |
_TEST_DOWNLOAD_URL = "https://github.com/PnrSvc/dataset/blob/main/turkish/test.csv"
|
47 |
|
48 |
|
49 |
+
BUILDER_CONFIGS = [
|
50 |
+
datasets.BuilderConfig(name="plain", version=VERSION, description="This loads the dataset in its plain format without any additional data transformations. In case of applying the dataset to a task (e.g. paraphrase classification or generation), some additional data transformations are suggested depending on the task (see'classification', 'classification-context', 'plain-context' and 'generation' for ready made transformations for paraphrase classification and paraphrase generation)."),
|
51 |
+
datasets.BuilderConfig(name="plain-context", version=VERSION, description="This loads the dataset in its plain format without any additional data transformations. In case of applying the dataset to a task (e.g. paraphrase classification or generation), some additional data transformations are suggested depending on the task (see 'classification', 'classification-context', 'plain' and 'generation' for ready made transformations for paraphrase classification and paraphrase generation). Unlike 'plain', this dataset includes the document contexts, which eats memory, but otherwise it is the same as 'plain'. This is useful for context-based modelling.")
|
52 |
+
|
53 |
+
]
|
54 |
|
55 |
VERSION = datasets.Version("1.1.0")
|
56 |
|