AxiomAi.PythonInterface (AxiomAI v0.1.11)

View Source

Python Interface - Interface module for embedded python functionality.

This module provides a clean API for interacting with the embedded python library within the axiom project. It handles Python environment management, code execution, and provides utilities for AI model inference.

Summary

Functions

cleanup_environment(category \\ :default)

Clean up Python environment for a specific category.

execute_inference(model_path, message, python_code, config, category \\ :default)

Execute AI model inference using Python.

execute_python(code, globals \\ %{}, category \\ :default)

Execute Python code with python_interface.

execute_streaming_inference(model_path, message, python_code, config, category \\ :default)

Execute streaming inference for AI models.

get_installed_packages(category \\ :default)

Get available Python packages in the environment.

init_environment(python_deps, category \\ :default)

Initialize Python environment with specified dependencies.

Functions

cleanup_environment(category \\ :default)

Clean up Python environment for a specific category.

Parameters

  • category: Category to clean up

Returns

  • :ok

execute_inference(model_path, message, python_code, config, category \\ :default)

Execute AI model inference using Python.

Parameters

  • model_path: Path to the AI model
  • message: Input message for the model
  • python_code: Python code template for inference
  • config: Configuration map with model parameters
  • category: Category for environment isolation

Returns

  • {:ok, response} with model response
  • {:error, reason} on failure

execute_python(code, globals \\ %{}, category \\ :default)

Execute Python code with python_interface.

Parameters

  • code: Python code string to execute
  • globals: Global variables dictionary (optional)
  • category: Category for environment isolation

Returns

  • {:ok, result} with decoded Python result
  • {:error, reason} on failure

execute_streaming_inference(model_path, message, python_code, config, category \\ :default)

Execute streaming inference for AI models.

Parameters

  • model_path: Path to the AI model
  • message: Input message for the model
  • python_code: Python code template for streaming inference
  • config: Configuration map with model parameters
  • category: Category for environment isolation

Returns

  • Stream of responses or error

get_installed_packages(category \\ :default)

Get available Python packages in the environment.

Parameters

  • category: Category for environment isolation

Returns

  • {:ok, packages} list of installed packages
  • {:error, reason} on failure

init_environment(python_deps, category \\ :default)

Initialize Python environment with specified dependencies.

Parameters

  • python_deps: List of Python dependencies to install
  • category: Category identifier for environment isolation

Returns

  • :ok on success
  • {:error, reason} on failure