task_categories: | |
- image-classification | |
# AutoTrain Dataset for project: image-classification | |
## Dataset Description | |
This dataset has been automatically processed by AutoTrain for project 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: | |
```json | |
[ | |
{ | |
"image": "<500x333 RGB PIL image>", | |
"target": 0 | |
}, | |
{ | |
"image": "<320x240 RGB PIL image>", | |
"target": 4 | |
} | |
] | |
``` | |
### Dataset Fields | |
The dataset has the following fields (also called "features"): | |
```json | |
{ | |
"image": "Image(decode=True, id=None)", | |
"target": "ClassLabel(num_classes=5, names=['daisy', 'dandelion', 'roses', 'sunflowers', 'tulips'], 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 | 160 | | |
| valid | 40 | | |