参照:
次のコマンドを使用して、このデータセットをTFDSにロードします。
ds = tfds.load('huggingface:code_x_glue_tc_text_to_code')
- 説明:
CodeXGLUE text-to-code dataset, available at https://github.com/microsoft/CodeXGLUE/tree/main/Text-Code/text-to-code
We use concode dataset which is a widely used code generation dataset from Iyer's EMNLP 2018 paper Mapping Language to Code in Programmatic Context. See paper for details.
- ライセンス:既知のライセンスはありません
- バージョン:0.0.0
- 分割:
スプリット | 例 |
---|---|
'test' | 2000 |
'train' | 100000 |
'validation' | 2000 |
- 特徴:
{
"id": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"nl": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"code": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}