Datasets:
vilm
/

Modalities:
Text
Formats:
json
ArXiv:
Libraries:
Datasets
pandas
License:
File size: 1,284 Bytes
4371fbc
 
 
927c320
d7482e0
 
91f1597
97dfcc4
 
 
 
a00aec7
97dfcc4
 
 
 
 
 
 
a00aec7
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
---
license: apache-2.0
---
## 🇻🇳 Vietnamese OpenOrca is here 🐋
<img src="https://i.ibb.co/kgmJG96/orca-viet.png" alt="drawing" width="512"/>

Dive into the Vietnamese linguistic landscape with OpenOrca, a cutting-edge dataset crafted through a pioneering partnership between **Virtual Interactive** and **Alignment Lab AI**. Drawing inspiration and methodology from the renowned [Orca paper](https://arxiv.org/abs/2306.02707), we've expanded our horizons to distill knowledge from a more eclectic mix of leading LLMs including GPT-4, PaLM-2, and Claude. Our vision with this dataset is to fuel research and development that will catapult the performance of Vietnamese Language Models into uncharted territories. Join us on this exhilarating journey to redefine AI's linguistic prowess.

The main original source of tasks/questions is a translated version of *FLAN*, **vi-FLAN**. We further augmented **vi-FLAN** on better state-of-the-art LLMs.

## Citation
```
@misc{OpenOrcaViet,
  title = {OpenOrca-Viet: GPT Augmented FLAN Reasoning for Vietnamese},
  author = {Virtual Interactive and Alignment Lab AI},
  year = {2023},
  publisher = {HuggingFace},
  journal = {HuggingFace repository},
  howpublished = {\url{https://https://huggingface.co/vilm/OpenOrca-Viet}},
}
```