Datasets:
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -33,4 +33,23 @@ configs:
|
|
| 33 |
path: data/train-*
|
| 34 |
- split: test
|
| 35 |
path: data/test-*
|
|
|
|
|
|
|
|
|
|
| 36 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
path: data/train-*
|
| 34 |
- split: test
|
| 35 |
path: data/test-*
|
| 36 |
+
license: apache-2.0
|
| 37 |
+
source_datasets:
|
| 38 |
+
- prajjwal1/discosense
|
| 39 |
---
|
| 40 |
+
|
| 41 |
+
# discosense
|
| 42 |
+
|
| 43 |
+
DiscoSense: four-choice commonsense reasoning with discourse connectives.
|
| 44 |
+
|
| 45 |
+
Original distribution: https://github.com/prajjwal1/discosense.
|
| 46 |
+
Source paper: https://arxiv.org/abs/2210.12478.
|
| 47 |
+
All native fields, option order, integer gold indices and train/test partitions
|
| 48 |
+
are preserved verbatim. No validation partition or new distractors are created.
|
| 49 |
+
Use both `context` and `marker` as input; `label` indexes `option_0`–`option_3`.
|
| 50 |
+
The pinned source Hub README declares Apache-2.0; see provenance for evidence.
|
| 51 |
+
Rebuild with `python scripts/upload_repackaged.py discosense`.
|
| 52 |
+
|
| 53 |
+
Original data: [prajjwal1/discosense](https://huggingface.co/datasets/prajjwal1/discosense), [original distribution](https://github.com/prajjwal1/discosense). Repackaged as parquet for [tasksource](https://github.com/sileod/tasksource) by [scripts/repackage_dataset/text.py](https://github.com/sileod/tasksource/blob/main/scripts/repackage_dataset/text.py#L39).
|
| 54 |
+
|
| 55 |
+
[Conversion provenance](provenance.json): source revisions, hashes, configuration and exclusions.
|