Guti

A Utility Library for Golang Developer

Usage

go get github.com/shaharia-lab/guti

And start using like -

import (
    "github.com/shaharia-lab/guti"
)

guti.ContainsAll()

AI Operations

The ai package provides a flexible interface for interacting with various Language Learning Models (LLMs). Currently supports OpenAI's GPT models with an extensible interface for other providers.

Basic Usage

import (
    "github.com/shaharia-lab/guti/ai"
)

// Create an OpenAI provider
provider := ai.NewOpenAILLMProvider(ai.OpenAIProviderConfig{
    APIKey: "your-api-key",
    Model:  "gpt-3.5-turbo", // Optional, defaults to gpt-3.5-turbo
})

// Create a request with default configuration
request := ai.NewLLMRequest(ai.NewRequestConfig())

// Generate a response
response, err := request.Generate([]LLMMessage{{Role: "user", Text: "What is the capital of France?"}}, provider)
if err != nil {
    log.Fatal(err)
}

fmt.Printf("Response: %s\n", response.Text)
fmt.Printf("Input tokens: %d\n", response.TotalInputToken)
fmt.Printf("Output tokens: %d\n", response.TotalOutputToken)
fmt.Printf("Completion time: %.2f seconds\n", response.CompletionTime)

Custom Configuration

You can customize the LLM request configuration using the functional options pattern:

// Use specific configuration options
config := ai.NewRequestConfig(
    ai.WithMaxToken(2000),
    ai.WithTemperature(0.8),
    ai.WithTopP(0.95),
    ai.WithTopK(100),
)

request := ai.NewLLMRequest(config)

Using Templates

The package also supports templated prompts:

template := &ai.LLMPromptTemplate{
    Template: "Hello {{.Name}}! Please tell me about {{.Topic}}.",
    Data: map[string]interface{}{
        "Name":  "Alice",
        "Topic": "artificial intelligence",
    },
}

prompt, err := template.Parse()
if err != nil {
    log.Fatal(err)
}

response, err := request.Generate(prompt, provider)

Configuration Options

Option	Default	Description
MaxToken	1000	Maximum number of tokens to generate
TopP	0.9	Nucleus sampling parameter (0-1)
Temperature	0.7	Randomness in output (0-2)
TopK	50	Top-k sampling parameter

Error Handling

The package provides structured error handling:

response, err := request.Generate(prompt, provider)
if err != nil {
    if llmErr, ok := err.(*ai.LLMError); ok {
        fmt.Printf("LLM Error %d: %s\n", llmErr.Code, llmErr.Message)
    } else {
        fmt.Printf("Error: %v\n", err)
    }
}

Custom Providers

You can implement the LLMProvider interface to add support for additional LLM providers:

type LLMProvider interface {
    GetResponse(messages []LLMMessage, config LLMRequestConfig) (LLMResponse, error)
}

Documentation

Full documentation is available on pkg.go.dev/github.com/shaharia-lab/guti

📝 License

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