ArabicMMLU: Measuring massive multitask language understanding in Arabic
This dataset has been translated from the original MMLU with the help of GPT-4.
Original Title: `COPA`
The original data [MMLU](https://arxiv.org/pdf/2009.03300v3.pdf)
The translation has been done with AceGPT researchers [AceGPT](https://arxiv.org/abs/2309.12053)
ArabicMMLU is a comprehensive evaluation benchmark specifically designed to evaluate the knowledge and reasoning abilities of LLMs within the context of Arabic language and culture.
ArabicMMLU covers a wide range of subjects, comprising 57 topics that span from elementary to advanced professional levels.
The Choice Of Plausible Alternatives (COPA) evaluation provides researchers with a tool for assessing progress in open-domain commonsense causal reasoning.
AlGhafa has translated this dataset to Arabic[AlGafa](https://aclanthology.org/2023.arabicnlp-1.21.pdf)
The link to the Arabic version of the dataset [PICA](https://gitlab.com/tiiuae/alghafa/-/tree/main/arabic-eval/copa_ar)
### Citation
### Groups and Tasks
#### Groups
-`ammlu`: All 57 subjects of the ArabicMMLU dataset, evaluated following the methodology in MMLU's original implementation.
* Not part of a group yet.
#### Tasks
The following tasks evaluate subjects in the ArabicMMLU dataset using loglikelihood-based multiple-choice scoring:
-`ammlu_{subject_english}`
*`copa_ar`
### Checklist
For adding novel benchmarks/datasets to the library:
* [x] Is the task an existing benchmark in the literature?
* [x] Have you referenced the original paper that introduced the task?
* [x] If yes, does the original paper provide a reference implementation?
* [x] Yes, original implementation contributed by author of the benchmark
* [x] If yes, does the original paper provide a reference implementation? If so, have you checked against the reference implementation and documented how to run such a test?
If other tasks on this dataset are already supported:
* [x] Is the "Main" variant of this task clearly denoted?
Original Title: `PIQA: Reasoning about Physical Commonsense in Natural Language`
Original paper: [PICA](https://arxiv.org/abs/1911.11641)
Physical Interaction: Question Answering (PIQA) is a physical commonsense
reasoning and a corresponding benchmark dataset. PIQA was designed to investigate
the physical knowledge of existing models. To what extent are current approaches
actually learning about the world?
[Homepage](https://yonatanbisk.com/piqa)
AlGhafa has translated this dataset to Arabic[AlGafa](https://aclanthology.org/2023.arabicnlp-1.21.pdf)
The link to the Arabic version of the dataset [PICA](https://gitlab.com/tiiuae/alghafa/-/tree/main/arabic-eval/pica_ar)
### Citation
### Groups and Tasks
#### Groups
* Not part of a group yet.
#### Tasks
*`piqa_ar`
### Checklist
For adding novel benchmarks/datasets to the library:
* [x] Is the task an existing benchmark in the literature?
* [x] Have you referenced the original paper that introduced the task?
* [x] If yes, does the original paper provide a reference implementation? If so, have you checked against the reference implementation and documented how to run such a test?
If other tasks on this dataset are already supported:
* [x] Is the "Main" variant of this task clearly denoted?
* [x] Have you provided a short sentence in a README on what each new variant adds / evaluates?
* [x] Have you noted which, if any, published evaluation setups are matched by this variant?
