参考：
gutenberg
使用以下命令在 TFDS 中加载此数据集：
ds = tfds.load('huggingface:gutenberg_time/gutenberg')
- 说明：
A clean data resource containing all explicit time references in a dataset of 52,183 novels whose full text is available via Project Gutenberg.
- 许可：[需要更多信息]
- 版本：0.0.0
- 拆分：
| 拆分 | 样本 |
|---|---|
'train' |
120694 |
- 特征：
{
"guten_id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"hour_reference": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"time_phrase": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"is_ambiguous": {
"dtype": "bool_",
"id": null,
"_type": "Value"
},
"time_pos_start": {
"dtype": "int64",
"id": null,
"_type": "Value"
},
"time_pos_end": {
"dtype": "int64",
"id": null,
"_type": "Value"
},
"tok_context": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}