README.md

---
license: apache-2.0
task_categories:
- text-generation
- text2text-generation
- translation
language:
- en
tags:
- code
pretty_name: BabelCode Transcoder
size_categories:
- 1K<n<10K
---
# Dataset Card for BabelCode Transcoder

## Dataset Description

- **Repository:** [GitHub Repository](https://github.com/google-research/babelcode)
- **Paper:** [Measuring The Impact Of Programming Language Distribution](https://arxiv.org/abs/2302.01973)

### How To Use This Dataset

To quickly evaluate BC-Transcoder predictions, save the `qid` and `language` keys along with the postprocessed prediction code in a JSON lines file. Then follow the install instructions for [BabelCode](https://github.com/google-research/babelcode), and you can evaluate your predictions. 

### Dataset Summary

The [Transcoder](https://github.com/facebookresearch/CodeGen) dataset in BabelCode format. Currently supports translation from C++ and Python.

### Supported Tasks and Leaderboards

### Languages
BC-Transcoder supports:
* C++
* C#
* Dart
* Go
* Haskell
* Java
* Javascript
* Julia
* Kotlin
* Lua
* PHP
* Python
* R
* Rust
* Scala
* TypeScript

## Dataset Structure

```python
>>> from datasets import load_dataset
>>> load_dataset("gabeorlanski/bc-transcoder")
DatasetDict({
    test: Dataset({
        features: ['qid', 'title', 'language', 'signature', 'arguments', 'entry_fn_name', 'entry_cls_name', 'test_code', 'source_py', 'source_cpp'],
        num_rows: 8384
    })
})
```

### Data Fields

- `qid`: The question ID used for running tests.
- `title`: The title of the question.
- `language`: The programming language of the example.
- `signature`: The signature for the problem.
- `arguments`: The arguments of the problem.
- `entry_fn_name`: The function's name to use an entry point.
- `entry_cls_name`: The class name to use an entry point.
- `test_code`: The raw testing script used in the language. If you want to use this, replace `PLACEHOLDER_FN_NAME` (and `PLACEHOLDER_CLS_NAME` if needed) with the corresponding entry points. Next, replace `PLACEHOLDER_CODE_BODY` with the postprocessed prediction.  
- `source_py`: The source solution in Python.
- `source_cpp`: The source in C++.

## Dataset Creation
See section 2 of the [BabelCode Paper](https://arxiv.org/abs/2302.01973) to learn more about how the datasets are translated. 

For information on the original curation of the Transcoder Dataset, please see [Unsupervised Translation of Programming Languages](https://arxiv.org/pdf/2006.03511.pdf) by Roziere et. al.

### Dataset Curators
Google Research

### Licensing Information
CC-BY-4.0

### Citation Information
```
@article{orlanski2023measuring,
  title={Measuring The Impact Of Programming Language Distribution},
  author={Orlanski, Gabriel and Xiao, Kefan and Garcia, Xavier and Hui, Jeffrey and Howland, Joshua and Malmaud, Jonathan and Austin, Jacob and Singh, Rishah and Catasta, Michele},
  journal={arXiv preprint arXiv:2302.01973},
  year={2023}
}
@article{roziere2020unsupervised,
  title={Unsupervised translation of programming languages},
  author={Roziere, Baptiste and Lachaux, Marie-Anne and Chanussot, Lowik and Lample, Guillaume},
  journal={Advances in Neural Information Processing Systems},
  volume={33},
  year={2020}
}
```