anam-python-sdk 0.3.0

An unofficial package to interact with Anam's Avatar engine.

Getting Started

This client uses Poetry for Python package & dependency management.

Backlog

• Core API
  • Establishing a P2P Websocket to the Anam engine
  • Configuring the API with a Pythonic interface
  • Plug and play with
    • Gradio, Streamlit or FastHTML
    • Fast API
• Character Dev
  • Create 3 to 4 personas
• Brains
  • Creating a RAG tool call & establishing a talk() command
  • Langchain integration
• Persona API
  • Configuring a new persona using the client (i.e. what you do in Anam Lab)
  • Selecting a persona

Installation (Package)

    pip install anam-python-sdk

    from anamai.core import AnamClient
    from anamai.config import DEFAULT_FILTER_PHRASES_EN
    from dotenv import dotenv_values

    ENV = dotenv_values(".env.example")

    anc = AnamClient(ENV)

    eva = anc.Persona(
        personality = """
        You are role-playing as a text chatbot hotel receptionist at The Sunset Hotel. 
        Your name is Eva.
        """, 
        system_prompt="""
        You are role-playing as a text chatbot hotel receptionist at The Sunset Hotel. 
        Your name is Eva. Start with 'Hi, this is The Sunset Hotel reception, how may I help you?' 
        Be mindful of the hotel's availability: single and double rooms 
        are still available all next week, except Tuesday. 
        Dogs are allowed. There's a restaurant and bar in the lobby. 
        If communication breaks down, ask if they want to speak to a human. 
        Finish every response with a question to drive the conversation. 
        Do not repeat yourself.
        """, 
        filter_phrases=DEFAULT_FILTER_PHRASES_EN
    )

Installation (Local)

Using Conda

1. Create a python environment (^3.10) in your top-level directory.
   • .conda/bin/python
   • Ensure that its activated; i.e. (.conda) shows.
2. Configure poetry to use .conda/bin/python:

    (.conda) poetry config virtualenvs.path $CONDA_ENV_PATH
    (.conda) poetry config virtualenvs.create false
    (.conda) poetry env use .conda/bin/python

3. Install the dependencies

    (.conda) poetry install