Configuration

Configuration

Global Configuration

For Chat API (Recommended)

ChatGPT.configure do |config|
  config.api_key = ENV['OPENAI_API_KEY']
  config.api_version = 'v1'
  config.default_engine = 'gpt-3.5-turbo'  # For chat
  config.request_timeout = 30
  config.max_retries = 3
  config.default_parameters = {
    max_tokens: 16,
    temperature: 0.5,
    top_p: 1.0,
    n: 1
  }
end

For Completions API

ChatGPT.configure do |config|
  config.api_key = ENV['OPENAI_API_KEY']
  config.api_version = 'v1'
  config.default_engine = 'gpt-3.5-turbo-instruct'  # For completions
  config.request_timeout = 30
  config.max_retries = 3
  config.default_parameters = {
    max_tokens: 16,
    temperature: 0.5,
    top_p: 1.0,
    n: 1
  }
end

Configuration Options

Available Settings

Option	Type	Default	Description
api_key	String	nil	Your OpenAI API key
api_version	String	'v1'	API version to use
default_engine	String	'gpt-3.5-turbo'	Default model to use
request_timeout	Integer	30	Request timeout in seconds
max_retries	Integer	3	Maximum retry attempts
default_parameters	Hash	See below	Default parameters for requests

Default Parameters

{
  max_tokens: 16,
  temperature: 0.5,
  top_p: 1.0,
  n: 1
}

Model Compatibility

API Method	Compatible Models	Recommended Configuration
Chat	gpt-4, gpt-3.5-turbo	default_engine = 'gpt-3.5-turbo'
Completions	gpt-3.5-turbo-instruct	default_engine = 'gpt-3.5-turbo-instruct'

Per-Client Configuration

client = ChatGPT::Client.new(
  api_key: 'different-key',
  timeout: 60,
  max_retries: 5
)

Environment Variables

OPENAI_API_KEY=your-api-key
OPENAI_ORGANIZATION=your-org-id # Optional

Rails Configuration

# config/initializers/chatgpt.rb
ChatGPT.configure do |config|
  config.api_key = Rails.application.credentials.openai[:api_key]
  config.request_timeout = Rails.env.development? ? 60 : 30
  
  # For Chat API (default)
  config.default_engine = 'gpt-3.5-turbo'
  
  # Or for Completions API
  # config.default_engine = 'gpt-3.5-turbo-instruct'
end

Best Practices

• Use environment variables for sensitive data
• Configure timeouts based on your use case
• Set appropriate retry limits
• Use organization ID if applicable
• Choose the appropriate model based on your use case:
  • Use Chat API (gpt-3.5-turbo) for conversational interactions
  • Use Completions API (gpt-3.5-turbo-instruct) for simple text completions
• Configure different settings for development and production environments
• Use Rails credentials for secure API key storage in production

Parameter Guidelines

Temperature

• 0.0: More focused and deterministic
• 0.5: Balanced creativity and consistency (default)
• 1.0: More creative and varied

Max Tokens

• Consider your use case requirements
• Longer responses need higher values
• Default (16) is suitable for short responses
• Adjust based on expected response length

Top P

• Works alongside temperature
• Lower values (0.1) make output more focused
• Higher values (1.0) allow more diversity