README

swapinvidya/laravel-huggingface-client is a Laravel package that simplifies integration with the Hugging Face API. This package provides an easy way to interact with models for text generation, image creation, and other AI-powered features.

Features

Text Generation: Generate text using models like gpt2 and gpt-neo.
Image Generation: Create images based on text prompts using models such as stable-diffusion.
Custom Model Completions: Use any available Hugging Face model for specialized tasks.
Seamless Integration: Built for Laravel, with easy dependency injection.

Installation

Step 1: Install via Composer

Run the following command in your Laravel project directory:

composer require swapinvidya/laravel-huggingface-client

Step 2: Publish Configuration

Publish the configuration file to customize your API key and other options:

php artisan vendor:publish --provider="Swapinvidya\HuggingFaceClient\HuggingFaceServiceProvider"

Step 3: Set Up API Key

Add your Hugging Face API key to the .env file:

HUGGINGFACE_API_KEY=your_huggingface_api_key

Configuration

The configuration file will be published to config/huggingface.php. You can customize your API key and other settings there.

return [
    'api_key' => env('HUGGINGFACE_API_KEY'),
    'base_uri' => 'https://api-inference.huggingface.co/',
];

Usage

Inject the HuggingFaceClient in your controllers or services to use its capabilities.

Example: Generate Text

use Swapinvidya\HuggingFaceClient\HuggingFaceClient;

public function generateText(HuggingFaceClient $huggingFaceClient)
{
    $response = $huggingFaceClient->generateText('gpt2', 'Write a short story about a hero.');
    return response()->json($response);
}

Example: Generate Image

public function generateImage(HuggingFaceClient $huggingFaceClient)
{
    $response = $huggingFaceClient->generateImage('stable-diffusion-v1', 'A futuristic city with flying cars.');
    return response()->json($response);
}

Example: Custom Model Completion

public function generateCompletion(HuggingFaceClient $huggingFaceClient)
{
    $params = [
        'inputs' => 'Explain the theory of relativity in simple terms.',
        'parameters' => [
            'temperature' => 0.7,
            'max_new_tokens' => 200,
        ],
    ];
    $response = $huggingFaceClient->generateCompletion('gpt-neo', $params);
    return response()->json($response);
}

Available Methods

`generateText($model, $input)`

Generates text output using the specified model.

$model: The model to use (e.g., gpt2, gpt-neo).
$input: The text input/prompt for the model.

`generateImage($model, $input)`

Creates an image based on a text prompt.

$model: The image generation model (e.g., stable-diffusion).
$input: The text prompt describing the image.

`generateCompletion($model, $params)`

Generates a completion response with custom parameters.

$model: The model to use (e.g., gpt-neo, opt).
$params: Array of parameters including inputs and optional model settings like temperature and max_new_tokens.

Error Handling

The client will return error information in the response if a request fails:

$response = $huggingFaceClient->generateText('gpt2', 'Hello world');

if (isset($response['error'])) {
    // Handle the error
    return response()->json(['error' => $response['error']], 500);
}

return response()->json($response);

Testing

To test your package locally, you can link it to a Laravel project using:

composer config repositories.local '{"type": "path", "url": "../path/to/laravel-huggingface-client"}'
composer require swapinvidya/laravel-huggingface-client:@dev

Contributing

Feel free to open issues or submit pull requests for improvements and new features.

License

This package is open-source software licensed under the MIT license.

Developed by swapinvidya

swapinvidya / laravel-huggingface-client

Maintainers

Details