Dataset Viewer
Full Screen Viewer
Full Screen
The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider
removing the
loading script
and relying on
automated data support
(you can use
convert_to_parquet
from the datasets
library). If this is not possible, please
open a discussion
for direct help.
TyDiQA-GoldP-Th
This dataset contains a removed Thai TyDiQA dataset obtained from Khalidalt's TyDiQA Dataset. This dataset version does the following additional preprocessing to the dataset
- Convert byte-level index into character-level index
- Fix any mismatch text between answer span and actual text
- Re-split train/development set such that there's no leakage in context passage
- Deduplicate questions from the same context passage
Dataset Format
The dataset is formatted to make it compatible to XTREME benchmark format. The data is formatted as the following pattern:
{
"version": "TyDiQA-GoldP-1.1-for-SQuAD-1.1",
"data": [
{
"paragrahs": [{
"context": [PASSAGE CONTEXT HERE],
"qas": [{
"answers": [{
"answer_start": [CONTEXT START CHAR INDEX OF ANSWER],
"text": [TEXT SPAN FROM CONTEXT],
}],
"question": [QUESTION],
"id": [ID]
}]
}],
},
...
]
}
Author
Chompakorn Chaksangchaichot
- Downloads last month
- 63