Client Factory

The ClientFactory provides a convenient way to create LLM clients for different providers without having to import and instantiate each client type individually.

datapizza.clients.factory.ClientFactory

Factory for creating LLM clients

create staticmethod

create(
    provider,
    api_key,
    model,
    system_prompt="",
    temperature=0.7,
    **kwargs,
)

Create a client instance based on the specified provider.

Parameters:

Name	Type	Description	Default
provider	str | Provider	The LLM provider to use (openai, google, or anthropic)	required
api_key	str	API key for the provider	required
model	str	Model name to use (provider-specific)	required
system_prompt	str	System prompt to use	''
temperature	float	Temperature for generation (0-2)	0.7
**kwargs		Additional provider-specific arguments	{}

Returns:

Type	Description
Client	An instance of the appropriate client

Raises:

Type	Description
ValueError	If the provider is not supported

Example Usage

from datapizza.clients.factory import ClientFactory, Provider

# Create an OpenAI client
openai_client = ClientFactory.create(
    provider=Provider.OPENAI,
    api_key="OPENAI_API_KEY",
    model="gpt-4",
    system_prompt="You are a helpful assistant.",
    temperature=0.7
)

# Create a Google client using string provider
google_client = ClientFactory.create(
    provider="google",
    api_key="GOOGLE_API_KEY",
    model="gemini-pro",
    system_prompt="You are a helpful assistant.",
    temperature=0.5
)

# Create an Anthropic client with custom parameters
anthropic_client = ClientFactory.create(
    provider=Provider.ANTHROPIC,
    api_key="ANTHROPIC_API_KEY",
    model="claude-3-sonnet-20240229",
    system_prompt="You are a helpful assistant.",
    temperature=0.3,
)

# Use the client
response = openai_client.invoke("What is the capital of France?")
print(response.content)

Supported Providers

• openai - OpenAI GPT models
• google - Google Gemini models
• anthropic - Anthropic Claude models
• mistral - Mistral AI models