| task_categories: | |
| - image-classification | |
| # AutoTrain Dataset for project: resnet50_test | |
| ## Dataset Description | |
| This dataset has been automatically processed by AutoTrain for project resnet50_test. | |
| ### 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": "<1920x1920 RGB PIL image>", | |
| "target": 2 | |
| }, | |
| { | |
| "image": "<1080x721 RGB PIL image>", | |
| "target": 2 | |
| } | |
| ] | |
| ``` | |
| ### Dataset Fields | |
| The dataset has the following fields (also called "features"): | |
| ```json | |
| { | |
| "image": "Image(decode=True, id=None)", | |
| "target": "ClassLabel(names=['000', '005', '033'], 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 | 2244 | | |
| | valid | 564 | | |