https://github.com/cosmaadrian/rocode

Official repository for "RoCode: A Dataset for Measuring Code Intelligence from Romanian Problem Definitions"

https://github.com/cosmaadrian/rocode

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
  • .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: arxiv.org, scholar.google
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (9.4%) to scientific vocabulary

Keywords

benchmark code-intelligence llm romanian
Last synced: 7 months ago · JSON representation

Repository

Official repository for "RoCode: A Dataset for Measuring Code Intelligence from Romanian Problem Definitions"

Basic Info
Statistics
  • Stars: 6
  • Watchers: 1
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Topics
benchmark code-intelligence llm romanian
Created almost 3 years ago · Last pushed about 2 years ago
Metadata Files
Readme License

README.md

RoCode: A Dataset for Measuring Code Intelligence from Problem Definitions in 🇷🇴 Romanian 🇷🇴

Accepted at the 2024 International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024).

[Adrian Cosma](https://scholar.google.com/citations?user=cdYk_RUAAAAJ&hl=en), [Bogdan Iordache](https://scholar.google.com/citations?user=wOvradYAAAAJ&hl=ro), [Paolo Rosso](https://scholar.google.es/citations?user=HFKXPH8AAAAJ&hl=en)
[📘 Abstract](#intro) | [📖 Citation](#citation) | [📝 License](#license)

In this work, we present RoCode, a competitive programming dataset, consisting of 2,642 problems written in Romanian, 11k solutions in C, C++ and Python and comprehensive testing suites for each problem.

TL;DR

[📜 Arxiv Link](https://arxiv.org/abs/2402.13222) | [🤗 Huggingface Dataset](https://huggingface.co/datasets/cosmadrian/rocode)

RoCode is a dataset of programming puzzles with problem statements in Romanian for evaluation of LLM code intelligence. Leetcode, but for Romanian.

📘 Abstract

Recently, large language models (LLMs) have become increasingly powerful and have become capable of solving a plethora of tasks through proper instructions in natural language. However, the vast majority of testing suites assume that the instructions are written in English, the de facto prompting language. Code intelligence and problem solving still remain a difficult task, even for the most advanced LLMs. Currently, there are no datasets to measure the generalization power for code-generation models in a language other than English. In this work, we present RoCode, a competitive programming dataset, consisting of 2,642 problems written in Romanian, 11k solutions in C, C++ and Python and comprehensive testing suites for each problem. The purpose of RoCode is to provide a benchmark for evaluating the code intelligence of language models trained on Romanian / multilingual text as well as a fine-tuning set for pretrained Romanian models. Through our results and review of related works, we argue for the need to develop code models for languages other than English.

📖 Citation

If you found our work useful, please cite our paper:

RoCode: A Dataset for Measuring Code Intelligence from Problem Definitions in 🇷🇴 Romanian 🇷🇴

@InProceedings{cosma24rocode, author="Cosma, Adrian and Iordache, Ioan-Bogdan and Rosso, Paolo", title="RoCode: A Dataset for Measuring Code Intelligence from Problem Definitions in Romanian", booktitle="The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", year="2024", }

📝 License

This work is protected by Attribution-NonCommercial 4.0 International

Owner

  • Name: Adrian Cosma
  • Login: cosmaadrian
  • Kind: user
  • Location: Bucharest, Romania
  • Company: University Politehnica of Bucharest

Mercenary Researcher

GitHub Events

Total
  • Watch event: 4
Last Year
  • Watch event: 4