languru 0.21.0

The general purpose LLM app stacks.

Languru

The general-purpose LLM app stacks deploy AI services quickly and (stupidly) simply.

 _
| |    __ _ _ __   __ _ _   _ _ __ _   _
| |   / _` | '_ \ / _` | | | | '__| | | |
| |__| (_| | | | | (_| | |_| | |  | |_| |
|_____\__,_|_| |_|\__, |\__,_|_|   \__,_|
                  |___/

Documentation: Github Pages

Install Languru

pip install languru

# Install For LLM deployment.
pip install languru[all]

# Install development dependencies.
poetry install -E <extras> --with dev

# Or just install all dependencies.
poetry install -E all --with dev --with docs

OpenAI Clients

Supported OpenAI clients:

• openai.OpenAI
• openai.AzureOpenAI
• languru.openai_plugins.clients.anthropic.AnthropicOpenAI
• languru.openai_plugins.clients.google.GoogleOpenAI
• languru.openai_plugins.clients.groq.GroqOpenAI
• languru.openai_plugins.clients.pplx.PerplexityOpenAI
• languru.openai_plugins.clients.voyage.VoyageOpenAI

OpenAI Server

languru server run  # Remember set all needed `api-key` for OpenAI clients.

Query LLM service, which is fully compatible with OpenAI APIs.

from openai import OpenAI

client = OpenAI(base_url="http://localhost:8682/v1")
res = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Hello!"},
    ],
)
for choice in res.choices:
    print(f"{choice.message.role}: {choice.message.content}")
# assistant: Hello! How can I assist you today?

Chat streaming:

client = OpenAI(base_url="http://localhost:8682/v1")
res = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Hello!"},
    ],
    stream=True,
)
for chunk in res:
    for choice in chunk.choices:
        if choice.delta.content:
            print(choice.delta.content, end="", flush=True)
            # Hello! How can I assist you today?

OpenAI plugins clients:

client = OpenAI(base_url="http://localhost:8682/v1")
res = client.chat.completions.create(
    model="google/gemini-1.5-flash",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Hello!"},
    ],
    stream=True,
)
for choice in res.choices:
    print(f"{choice.message.role}: {choice.message.content}")