HuggingFace.Net 1.0.0

HuggingFace.Net

HuggingFace.Net is a .NET client library for interacting with HuggingFace's Inference Endpoints API. It provides a robust and easy-to-use interface for managing and calling HuggingFace machine learning models.

Features

• 🔐 Secure API Authentication
• 🚀 Inference Endpoint Management
  • Create, delete, and manage endpoints
  • Scale endpoints up/down
  • Monitor endpoint status
• 🤖 Model Inference
  • Call models with typed responses
  • Raw response handling
  • Automatic retries with exponential backoff
• ⚡ Performance
  • Connection pooling
  • Configurable timeouts
  • Rate limit handling
• 🛡️ Error Handling
  • Custom exceptions for different error scenarios
  • Detailed error messages
  • HTTP status code handling

Installation

You can install the package via NuGet:

dotnet add package HuggingFace.Net

Usage

Here's a quick example of how to use the library:

// Create client with API key
var client = HuggingFaceClientFactory.Create("your-api-key-here", "namespace");

try
{
    // Test connection
    var connected = await client.ConnectAsync();
    Console.WriteLine($"Connected: {connected}");

    // Check endpoint status
    var status = await client.GetEndpointStatusAsync("model-name");
    Console.WriteLine($"Endpoint Status: {status.Status}, Available: {status.IsAvailable}");

    // Call model for text generation
    var input = new Models.TextGenerationInput
    {
        Inputs = "Hello, how are you?",
        Parameters = new Models.TextGenerationParameters
        {
            MaxNewTokens = 50,
            Temperature = 0.7
        }
    };

    var response = await client.CallModelAsync<Models.TextGenerationResponse[]>(
        "model-name", input);
}
catch (HuggingFaceException ex)
{
    Console.WriteLine($"Error: {ex.Message}");
}

Configuration

You can configure the client with custom options:

var options = new HuggingFaceClientOptions
{
    ApiKey = "your-api-key",
    Namespace = "your-namespace",
    BaseUrl = "https://api.endpoints.huggingface.cloud/v2",
    Timeout = TimeSpan.FromSeconds(30),
    MaxRetries = 3,
    RetryDelay = TimeSpan.FromSeconds(1),
    UseExponentialBackoff = true
};

var client = HuggingFaceClientFactory.Create(options);

Error Handling

The library includes custom exceptions for different error scenarios:

• HuggingFaceException: Base exception for all library errors
• HuggingFaceAuthenticationException: Authentication errors
• HuggingFaceRateLimitException: Rate limit exceeded errors

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.