Zalo AI

license

not-clearly-stated-on-card(README 'Licenses: TBU'

access

public(zip downloads offered on vmlu.ai

dataset_card

present(README gives subject table, format and examples

answers

held-out(test predictions are submitted as an id,answer CSV to vmlu.ai/submit after login

Questions can be downloaded from the project site, and test scoring runs through a submission page that requires login. No license for the data itself is published: the repository's MIT file covers the code, and its license section is unfinished.

• https://raw.githubusercontent.com/ZaloAI-Jaist/VMLU/main/LICENSE.txt recorded 2026-09-24

  MIT License, Copyright (c) 2023 Zalo AI and JAIST; grants rights to 'the Software'

• https://raw.githubusercontent.com/ZaloAI-Jaist/VMLU/main/README.md recorded 2026-09-24

  "## Licenses TBU"; "Download the zip file: Please visit our website"

• https://raw.githubusercontent.com/ZaloAI-Jaist/VMLU/main/README.md recorded 2026-09-24

  "prepare a UTF-8 encoded CSV file ... id,answer"; "submit the prepared csv file here (https://vmlu.ai/submit), note that you need to first log in"

• https://vmlu.ai recorded 2026-09-24

  "The VMLU datasets can be downloaded as shown below: Vi-MQA v1.5, Vi-SQuAD v1.0, Vi-Drop v1.0, Vi-Dialog v1.0"

GitHub stars on the VMLU repository; the benchmark is also served from vmlu.ai, which publishes no download count, and a star is attention rather than use.

• https://ungh.cc/repos/ZaloAI-Jaist/VMLU recorded 2026-09-24

  GitHub repository record for ZaloAI-Jaist/VMLU (via the ungh.cc mirror of the GitHub API): 83 stargazers, last push 2024-05-04.

VMLU gathers Vietnamese exam questions in 58 subjects and runs a public leaderboard that scores open and API models. It covers one language and one kind of task, where SEA-HELM spans many task types across eleven languages, and its documentation is a README and website rather than a paper.

• https://raw.githubusercontent.com/ZaloAI-Jaist/VMLU/main/README.md recorded 2026-09-24

  "VMLU dataset covers 58 subjects including 10,880 multiple-choice questions and answers in Vietnamese language"; zero-shot leaderboard table

• https://vmlu.ai recorded 2026-09-24

  "The dataset primarily originates from examinations administered by a diverse array of esteemed educational institutions"