metadata
task_categories:
- image-classification
AutoTrain Dataset for project: map_no_map_twitter_demo
Dataset Description
This dataset has been automatically processed by AutoTrain for project map_no_map_twitter_demo.
Languages
The BCP-47 code for the dataset's language is unk.
Dataset Structure
Data Instances
A sample from this dataset looks as follows:
[
{
"image": "<260x365 RGB PIL image>",
"target": 1,
"feat_created_at": "2023-03-08T13:56:12.283125Z",
"feat_lead_time": 0.768,
"feat_annotation_id": 23,
"feat_id": 24,
"feat_annotator": 1,
"feat_updated_at": "2023-03-08T13:56:12.283151Z"
},
{
"image": "<639x1298 RGB PIL image>",
"target": 0,
"feat_created_at": "2023-03-08T13:59:10.554628Z",
"feat_lead_time": 0.533,
"feat_annotation_id": 87,
"feat_id": 88,
"feat_annotator": 1,
"feat_updated_at": "2023-03-08T13:59:10.554650Z"
}
]
Dataset Fields
The dataset has the following fields (also called "features"):
{
"image": "Image(decode=True, id=None)",
"target": "ClassLabel(names=['map', 'no_map'], id=None)",
"feat_created_at": "Value(dtype='string', id=None)",
"feat_lead_time": "Value(dtype='float64', id=None)",
"feat_annotation_id": "Value(dtype='int64', id=None)",
"feat_id": "Value(dtype='int64', id=None)",
"feat_annotator": "Value(dtype='int64', id=None)",
"feat_updated_at": "Value(dtype='string', id=None)"
}
Dataset Splits
This dataset is split into a train and validation split. The split sizes are as follow:
| Split name | Num samples |
|---|---|
| train | 72 |
| valid | 19 |