Datasets:
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -33,4 +33,29 @@ configs:
|
|
| 33 |
path: data/train-*
|
| 34 |
- split: validation
|
| 35 |
path: data/validation-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
path: data/train-*
|
| 34 |
- split: validation
|
| 35 |
path: data/validation-*
|
| 36 |
+
language:
|
| 37 |
+
- en
|
| 38 |
+
license: other
|
| 39 |
+
source_datasets: []
|
| 40 |
+
task_categories:
|
| 41 |
+
- multiple-choice
|
| 42 |
+
- visual-question-answering
|
| 43 |
+
tags:
|
| 44 |
+
- decision-model
|
| 45 |
+
- procedural
|
| 46 |
+
license_name: tron-research-use
|
| 47 |
+
license_link: https://github.com/YangTianze009/TRON#license
|
| 48 |
---
|
| 49 |
+
|
| 50 |
+
# tron-mcq
|
| 51 |
+
|
| 52 |
+
TRON static MC: native text options or verifier-rejected same-environment answers.
|
| 53 |
+
|
| 54 |
+
Fresh, deterministic train/validation seeds; no model-derived labels or cross-task
|
| 55 |
+
distractors. Generation failures, opaque image-choice tasks and insufficient
|
| 56 |
+
distractor diversity are reported. Original generated PNGs are preserved in grouped
|
| 57 |
+
image rows. Source: https://github.com/YangTianze009/TRON (research-use terms).
|
| 58 |
+
|
| 59 |
+
Original data: [original distribution](https://github.com/YangTianze009/TRON). Repackaged as parquet for [tasksource](https://github.com/sileod/tasksource) by [scripts/repackage_dataset/procedural/tron.py](https://github.com/sileod/tasksource/blob/main/scripts/repackage_dataset/procedural/tron.py#L49).
|
| 60 |
+
|
| 61 |
+
[Conversion provenance](provenance.json): source revisions, hashes, configuration and exclusions.
|