Datasets:
Add task category and link to paper
#2
by nielsr HF Staff - opened
README.md
CHANGED
|
@@ -1,5 +1,16 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: cc-by-nc-4.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
dataset_info:
|
| 4 |
features:
|
| 5 |
- name: audio
|
|
@@ -24,45 +35,38 @@ tags:
|
|
| 24 |
- audio
|
| 25 |
- in-the-wild
|
| 26 |
- compound-task
|
| 27 |
-
language:
|
| 28 |
-
- en
|
| 29 |
-
- zh
|
| 30 |
-
- ru
|
| 31 |
-
- th
|
| 32 |
-
- bn
|
| 33 |
-
multilinguality:
|
| 34 |
-
- multilingual
|
| 35 |
-
pretty_name: GlobeAudio
|
| 36 |
configs:
|
| 37 |
- config_name: en
|
| 38 |
data_files:
|
| 39 |
-
|
| 40 |
-
|
| 41 |
default: true
|
| 42 |
- config_name: zh
|
| 43 |
data_files:
|
| 44 |
-
|
| 45 |
-
|
| 46 |
- config_name: sg
|
| 47 |
data_files:
|
| 48 |
-
|
| 49 |
-
|
| 50 |
- config_name: ru
|
| 51 |
data_files:
|
| 52 |
-
|
| 53 |
-
|
| 54 |
- config_name: th
|
| 55 |
data_files:
|
| 56 |
-
|
| 57 |
-
|
| 58 |
- config_name: bn
|
| 59 |
data_files:
|
| 60 |
-
|
| 61 |
-
|
| 62 |
---
|
| 63 |
|
| 64 |
# GlobeAudio Dataset
|
| 65 |
|
|
|
|
|
|
|
| 66 |
## Table of Contents
|
| 67 |
- [Dataset Description](#dataset-description)
|
| 68 |
- [Language Selection](#language-selection)
|
|
@@ -100,9 +104,9 @@ Each audio clip is approximately 20-40 seconds long, and contains naturally occu
|
|
| 100 |
The `datasets` library lets you load and preprocess datasets in Python at scale. You may load the dataset to your local drive with `load_dataset`.
|
| 101 |
|
| 102 |
For example, to download the English config, specify the corresponding language config name (i.e., "en" for English):
|
| 103 |
-
```
|
| 104 |
from datasets import load_dataset
|
| 105 |
|
| 106 |
# load a specific language eg. "en", "ru", "zh", "sg", "th", "bn"
|
| 107 |
-
dataset = load_dataset("iNLP-Lab/GlobeAudio",
|
| 108 |
```
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
- zh
|
| 5 |
+
- ru
|
| 6 |
+
- th
|
| 7 |
+
- bn
|
| 8 |
license: cc-by-nc-4.0
|
| 9 |
+
multilinguality:
|
| 10 |
+
- multilingual
|
| 11 |
+
pretty_name: GlobeAudio
|
| 12 |
+
task_categories:
|
| 13 |
+
- audio-text-to-text
|
| 14 |
dataset_info:
|
| 15 |
features:
|
| 16 |
- name: audio
|
|
|
|
| 35 |
- audio
|
| 36 |
- in-the-wild
|
| 37 |
- compound-task
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
configs:
|
| 39 |
- config_name: en
|
| 40 |
data_files:
|
| 41 |
+
- split: train
|
| 42 |
+
path: data/en/en-*
|
| 43 |
default: true
|
| 44 |
- config_name: zh
|
| 45 |
data_files:
|
| 46 |
+
- split: train
|
| 47 |
+
path: data/zh/zh-*
|
| 48 |
- config_name: sg
|
| 49 |
data_files:
|
| 50 |
+
- split: train
|
| 51 |
+
path: data/sg/sg-*
|
| 52 |
- config_name: ru
|
| 53 |
data_files:
|
| 54 |
+
- split: train
|
| 55 |
+
path: data/ru/ru-*
|
| 56 |
- config_name: th
|
| 57 |
data_files:
|
| 58 |
+
- split: train
|
| 59 |
+
path: data/th/th-*
|
| 60 |
- config_name: bn
|
| 61 |
data_files:
|
| 62 |
+
- split: train
|
| 63 |
+
path: data/bn/bn-*
|
| 64 |
---
|
| 65 |
|
| 66 |
# GlobeAudio Dataset
|
| 67 |
|
| 68 |
+
This is the dataset for the paper [GlobeAudio: A Multilingual Multicultural Benchmark for Naturalistic Evaluation of Large Audio-Language Models](https://huggingface.co/papers/2606.08194).
|
| 69 |
+
|
| 70 |
## Table of Contents
|
| 71 |
- [Dataset Description](#dataset-description)
|
| 72 |
- [Language Selection](#language-selection)
|
|
|
|
| 104 |
The `datasets` library lets you load and preprocess datasets in Python at scale. You may load the dataset to your local drive with `load_dataset`.
|
| 105 |
|
| 106 |
For example, to download the English config, specify the corresponding language config name (i.e., "en" for English):
|
| 107 |
+
```python
|
| 108 |
from datasets import load_dataset
|
| 109 |
|
| 110 |
# load a specific language eg. "en", "ru", "zh", "sg", "th", "bn"
|
| 111 |
+
dataset = load_dataset("iNLP-Lab/GlobeAudio", "en")
|
| 112 |
```
|