File size: 2,974 Bytes
52983f2 f0eef0c 52983f2 a15f4f7 52983f2 a15f4f7 52983f2 a15f4f7 52983f2 a15f4f7 52983f2 a15f4f7 52983f2 a15f4f7 f0eef0c 9b49eec f0eef0c 52983f2 f0eef0c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 |
---
license: cc-by-4.0
task_categories:
- multiple-choice
language:
- es
tags:
- spanish
pretty_name: teleia
size_categories:
- n<1K
configs:
- config_name: cervantes_ave
data_files:
- split: train
path: data/cervantes_ave/teleia_cervantes_ave_train.jsonl
- split: val
path: data/cervantes_ave/teleia_cervantes_ave_val.jsonl
- split: test
path: data/cervantes_ave/teleia_cervantes_ave_test.jsonl
- config_name: pce
data_files:
- split: train
path: data/pce/teleia_pce_train.jsonl
- split: val
path: data/pce/teleia_pce_val.jsonl
- split: test
path: data/pce/teleia_pce_test.jsonl
- config_name: siele
data_files:
- split: train
path: data/siele/teleia_siele_train.jsonl
- split: val
path: data/siele/teleia_siele_val.jsonl
- split: test
path: data/siele/teleia_siele_test.jsonl
---
# Spanish Language Benchmark for Artificial Intelligence Models (TELEIA)
## Authors and Affiliations
Marina Mayor-Rocher<sup>1</sup> , Nina Melero<sup>2,3</sup> , Elena Merino-Gómez<sup>4</sup> , Miguel González<sup>2</sup> , Raquel Ferrando<sup>2</sup> , Javier Conde<sup>2</sup> and Pedro Reviriego<sup>2</sup>
1. Universidad Autónoma de Madrid
2. Universidad Politécnica de Madrid
3. New York University
4. Universidad de Valladolid
## Description
The dataset contains test questions to evaluate LLMs in Spanish
* **TELEIA_Cervantes_AVE:** test questions on vocabulary and grammatical structures, following the format of the Cervantes AVE exam. The column "question" contains a sentence with a gap that must be filled in. The columns "option_a", "option_b", "option_c", "option_d" and "correct_answer" (A,B,C or D) represent the different possibilities for the gap and the correct one.
* **TELEIA_PCE:** test on morphology and semantics resembling the style of the PCE exam, consisting of short questions or sentences to be completed. It is composed of the following columns: "question", "option_a", "option_b", "option_c" and "correct_answer" (A, B or C).
* **TELEIA_SIELE:** different texts with questions related to them, based on the reading comprehension task of the SIELE exam. It is composed of the following columns: "question", "option_a", "option_b", "option_c" and "correct_answer" (A, B or C).
## Citation
```
@dataset{spanish_benchmark_teleia,
author = {Mayor-Rocher, M. and
Melero, N. and
Merino-Gómez, E. and
González, M. and
Ferrando, R. and
Conde, J. and
Reviriego, P.},
title = {{Spanish Language Benchmark for Artificial Intelligence Models (TELEIA)}},
month = jan,
year = 2024,
publisher = {Zenodo},
version = {1.0},
doi = {10.5281/zenodo.12571763},
url = {https://doi.org/10.5281/zenodo.12571763}
}
```
---
license: cc-by-4.0
--- |