---
task_categories:
- image-classification

---
# AutoTrain Dataset for project: fine-tune

## Dataset Description

This dataset has been automatically processed by AutoTrain for project fine-tune.

### 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": "<382x256 RGB PIL image>",
    "target": 17
  },
  {
    "image": "<341x256 RGB PIL image>",
    "target": 7
  }
]
```

### Dataset Fields

The dataset has the following fields (also called "features"):

```json
{
  "image": "Image(decode=True, id=None)",
  "target": "ClassLabel(names=['apple', 'banana', 'cake', 'candy', 'carrot', 'cookie', 'doughnut', 'grape', 'hot dog', 'ice cream', 'juice', 'muffin', 'orange', 'pineapple', 'popcorn', 'pretzel', 'salad', 'strawberry', 'waffle', 'watermelon'], 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        | 5394 |
| valid        | 1351 |