metadata
task_categories:
- image-classification
AutoTrain Dataset for project: histopathological_image_classification
Dataset Description
This dataset has been automatically processed by AutoTrain for project histopathological_image_classification.
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": "<700x460 RGB PIL image>",
"target": 6
},
{
"image": "<700x460 RGB PIL image>",
"target": 2
}
]
Dataset Fields
The dataset has the following fields (also called "features"):
{
"image": "Image(decode=True, id=None)",
"target": "ClassLabel(names=['1', '2', '3', '4', '5', '6', '7', '8'], 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 | 333 |
| valid | 89 |