| task_categories: | |
| - image-classification | |
| # AutoTrain Dataset for project: donut-vs-croissant | |
| ## Dataset Descritpion | |
| This dataset has been automatically processed by AutoTrain for project donut-vs-croissant. | |
| ### Languages | |
| The BCP-47 code for the dataset's language is unk. | |
| ## Dataset Structure | |
| ### Data Instances | |
| A sample from this dataset looks as follows: | |
| ```json | |
| [ | |
| { | |
| "image": "<512x512 RGB PIL image>", | |
| "target": 0 | |
| }, | |
| { | |
| "image": "<512x512 RGB PIL image>", | |
| "target": 0 | |
| } | |
| ] | |
| ``` | |
| ### Dataset Fields | |
| The dataset has the following fields (also called "features"): | |
| ```json | |
| { | |
| "image": "Image(decode=True, id=None)", | |
| "target": "ClassLabel(num_classes=2, names=['croissant', 'donut'], 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 | 133 | | |
| | valid | 362 | | |