metadata
task_categories:
- image-classification
AutoTrain Dataset for project: dog-classifiers
Dataset Descritpion
This dataset has been automatically processed by AutoTrain for project dog-classifiers.
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": "<474x592 RGB PIL image>",
"target": 1
},
{
"image": "<474x296 RGB PIL image>",
"target": 1
}
]
Dataset Fields
The dataset has the following fields (also called "features"):
{
"image": "Image(decode=True, id=None)",
"target": "ClassLabel(num_classes=5, names=['akita inu', 'corgi', 'leonberger', 'samoyed', 'shiba inu'], 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 | 598 |
valid | 150 |