metadata

task_categories:
  - image-classification

AutoTrain Dataset for project: pr_final_covid-19

Dataset Description

This dataset has been automatically processed by AutoTrain for project pr_final_covid-19.

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": "<299x299 L PIL image>",
    "target": 0
  },
  {
    "image": "<299x299 L PIL image>",
    "target": 0
  }
]

Dataset Fields

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

{
  "image": "Image(decode=True, id=None)",
  "target": "ClassLabel(names=['Covid', 'Covid_test', 'Lung_Opacity', 'Lung_Opacity_test', 'Normal', 'Normal_test'], 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	399
valid	99