Hugging Face datasets

¶Is Valid

The Is Valid API checks whether a Hugging Face dataset repository supports specific functionalities, such as:

• Preview: Whether the dataset can be previewed on the Hugging Face UI.
• Viewer: Whether the dataset has an interactive viewer (e.g., table or file view).
• Search: Whether the dataset supports search operations.
• Filter: Whether filtering capabilities are available.
• Statistics: Whether dataset statistics (e.g., column distributions) are available.

This API helps you verify compatibility and features availability for a given dataset.

Tip

Use this method before integrating with a dataset to ensure it supports the features your application or workflow requires.

¶Code

use Partitech\PhpMistral\Clients\HuggingFace\HuggingFaceDatasetClient;
use Partitech\PhpMistral\MistralClientException;

$apiKey = getenv('HF_TOKEN');  // Hugging Face API token

$client = new HuggingFaceDatasetClient(apiKey: (string) $apiKey);

try {
    // Validate the 'google/civil_comments' dataset
    $isValid = $client->isValid('google/civil_comments');

    print_r($isValid);  // Output dataset capabilities

} catch (MistralClientException $e) {
    print_r($e);
}

¶Result

Array
(
    [preview] => 1
    [viewer] => 1
    [search] => 1
    [filter] => 1
    [statistics] => 1
)

• preview: Whether the dataset can be previewed in the Hugging Face interface.
• viewer: Whether the dataset supports an interactive viewer.
• search: Whether search functionality is available.
• filter: Whether filtering options are enabled.
• statistics: Whether dataset statistics are computed and available.

A value of 1 means the feature is supported; 0 means it is not.

¶Use Cases

• Feature detection: Determine what functionalities are available for a dataset before integrating it into your application.
• Compatibility checks: Ensure a dataset supports specific operations (e.g., search, filtering).
• UI customization: Adjust your application behavior based on dataset capabilities.

¶Common Pitfalls

Warning

This API only validates dataset repositories. It does not work with models or spaces.

Note

Even if a dataset supports viewer or statistics, certain features (e.g., search) might depend on the dataset size, format, or structure.