Given a partial description like "she opened the hood of the car," humans can reason about the situation and anticipate what might come next ("then, she examined the engine").
Dataset Card
Given a partial description like "she opened the hood of the car," humans can reason about the situation and anticipate what might come next ("then, she examined the engine"). SWAG (Situations With Adversarial Generations) is a large-scale dataset for this task of grounded commonsense inference, unifying natural language inference and physically grounded reasoning. The dataset consists of 113k multiple choice questions about grounded situations (73k training, 20k validation, 20k test). Each question is a video caption from LSMDC or ActivityNet Captions, with four answer choices about what might happen next in the scene. The correct answer is the (real) video caption for the next event in…
Excerpt from the card by Ai2, licensed unknown.
Structure
regular 113,557 rows
| Split | Rows | Size |
|---|---|---|
| train | 73,546 | 30.2 MB |
| validation | 20,006 | 8.5 MB |
| test | 20,005 | 8.4 MB |
full 93,552 rows
| Split | Rows | Size |
|---|---|---|
| train | 73,546 | 35.0 MB |
| validation | 20,006 | 9.9 MB |
Details
- Repository
- allenai/swag
- Publisher
- Ai2
- Task category
- Text classification
- Tags
- Not stated by the source
- Size category
- 100K<n<1M
- Languages
- en
- Revision
- dc48148372b3853a9c7bae7bb06c161b46d8364a
- Last updated
- 2024-06-14
Files
7 files, 44.9 MB in total.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| full/train-00000-of-00001.parquet | Data | 15.2 MB | 0e98fa0f1805 |
| full/validation-00000-of-00001.parquet | Data | 5.3 MB | 029f32dd7fbb |
| regular/test-00000-of-00001.parquet | Data | 4.8 MB | fd899621f712 |
| regular/train-00000-of-00001.parquet | Data | 14.8 MB | 8f698d37e7fc |
| regular/validation-00000-of-00001.parquet | Data | 4.8 MB | 647853ce2222 |
| README.md | Documentation | 9.2 KB | — |
| .gitattributes | Repository | 1.2 KB | — |
License and Download
- License
- unknown
- Access
- No access gate
Released by Ai2 through its official repository on Hugging Face.