AMALIA VL Eval
Collection
Datasets used to evaluate AMALIA-VL. Also included in AMALIA VL Eval. • 21 items • Updated
Datasets:
dataset stringclasses 25
values | question_type stringclasses 10
values | answer listlengths 1 4 | image dict | question stringlengths 12 164 |
|---|---|---|---|---|
IIIT5K | Regular Text Recognition | [
"CENTRE"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
104,
0,
0,
0,
27,
8,
2,
0,
0,
0,
31,
200,
93,
153,
0,
0,
16,
122,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"FRIEND"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
164,
0,
0,
0,
60,
8,
2,
0,
0,
0,
103,
226,
134,
56,
0,
0,
36,
79,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"CHAIN"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
114,
0,
0,
0,
54,
8,
2,
0,
0,
0,
148,
118,
90,
80,
0,
0,
40,
137,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"CLOSE"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
107,
0,
0,
0,
37,
8,
2,
0,
0,
0,
202,
143,
128,
71,
0,
0,
23,
44,
73,
68,
... | o que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"MARKET"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
1,
5,
0,
0,
0,
69,
8,
2,
0,
0,
0,
204,
219,
100,
30,
0,
0,
110,
142,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"EXTRA"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
186,
0,
0,
0,
56,
8,
2,
0,
0,
0,
197,
20,
117,
117,
0,
0,
84,
69,
73,
68,
... | o que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"MOBI"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
60,
0,
0,
0,
23,
8,
2,
0,
0,
0,
216,
245,
187,
208,
0,
0,
11,
211,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"COTTAGE"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
121,
0,
0,
0,
39,
8,
2,
0,
0,
0,
164,
28,
112,
153,
0,
0,
40,
31,
73,
68,
... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"AHEAD"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
72,
0,
0,
0,
24,
8,
2,
0,
0,
0,
214,
1,
44,
231,
0,
0,
13,
14,
73,
68,
65,... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"TIMES"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
50,
0,
0,
0,
30,
8,
2,
0,
0,
0,
225,
51,
218,
171,
0,
0,
12,
17,
73,
68,
6... | O que está escrito na imagem? |
IIIT5K | Regular Text Recognition | [
"AIRES"
] | {
"bytes": [
137,
80,
78,
71,
13,
10,
26,
10,
0,
0,
0,
13,
73,
72,
68,
82,
0,
0,
0,
248,
0,
0,
0,
120,
8,
2,
0,
0,
0,
80,
3,
168,
132,
0,
0,
170,
218,
73,
68,
... | O que está escrito na imagem? |
End of preview. Expand in Data Studio
OCRBench-PT
European Portuguese (pt-PT) machine translation of OCRBench, a benchmark evaluating the optical character recognition and text understanding abilities of multimodal models.
Translated from the original English test split using gemini-3.1-pro.
Original Dataset: https://huggingface.co/datasets/echo840/OCRBench
Note: This dataset is machine translated and may contain translation errors or artifacts.
This dataset is provided as part of the AMALIA project and is included in amalia-vl-eval, a comprehensive benchmark suite for evaluating large vision-language models on European Portuguese.
Licensing
This dataset is dual-licensed:
- Translated content — the questions, answers, and any other text translated to European Portuguese with gemini-3.1-pro is released under
cc-by-nc-4.0(Attribution–NonCommercial). It may be used for non-commercial purposes with attribution. - Images and untranslated columns — all images and any columns that were not translated retain the original OCRBench license (refer to the original dataset).
When using this dataset you must comply with both licenses. The NonCommercial terms of the translated content do not override any stricter conditions of the original benchmark's license, and vice versa.
Citation
If you use this dataset or AMALIA-VL in your work, please cite:
@article{gloria2026amalia,
title={AMALIA-VL: A Native European Portuguese Open-Source Vision and Language Model},
author={Gl{\'o}ria-Silva, Diogo and Cardeira, Jo{\~a}o and da Luz, Manuel Letras and Simpl{\'\i}cio, Afonso and Vinagre, Gon{\c{c}}alo and Tavares, Diogo and Ferreira, Rafael and Calvo, In{\^e}s and Vieira, In{\^e}s and Semedo, David and others},
journal={arXiv preprint arXiv:2606.19100},
year={2026}
}
- Downloads last month
- 3