___ _ _
/ _ \/\_/\__ _ ___ _ __ | |_(_) ___
/ /_)/\_ _/ _` |/ _ \ '_ \| __| |/ __|
/ ___/ / \ (_| | __/ | | | |_| | (__
\/ \_/\__, |\___|_| |_|\__|_|\___|
|___/
Created by rUv
Pygentic is an innovative system designed to enhance the capabilities of AI assistants by providing a flexible and standardized API.
Based on the robust architecture of the OpenAI Assistants API, Pygentic abstracts the complexities of integrating with different Large Language Models (LLMs), both local and remote. This abstraction ensures a consistent and seamless interaction regardless of the underlying model.
The Pygentic library acts as a bridge between your application and various LLMs. It simplifies the process of connecting to different AI models by offering a uniform API, eliminating the need to manage the specifics of each model. Whether you're using OpenAI, a local model, or another remote service, Pygentic provides a reliable and consistent interface.
Pygentic introduces the concept of agents—intelligent assistants that can perform a range of tasks. These agents are configured to understand and respond to user inputs, utilizing different LLMs for processing. You can create and manage multiple agents, each tailored to specific functions or tasks, enhancing the versatility and efficiency of your AI-driven applications.
Concurrency in Pygentic ensures that multiple tasks can be handled simultaneously without compromising performance. This is particularly important for applications requiring real-time responses or handling numerous requests at once. Pygentic's design leverages modern asynchronous programming techniques to manage these tasks efficiently, providing a smooth and responsive experience.
Pygentic's plugin connector system allows you to easily extend the core functionality by adding custom connectors without modifying the main codebase. This system uses a flexible configuration approach with TOML files, enabling you to define new connectors in separate Python files and dynamically load them into the application. This ensures seamless integration and easy updates while maintaining stability and performance.
Pygentic supports serverless architectures, enabling you to deploy your AI applications without the need to manage infrastructure. This approach reduces operational complexity and costs, as the serverless platform handles scaling, monitoring, and resource allocation automatically. You can focus on developing and refining your AI capabilities, while the serverless environment ensures robust and scalable
The open-interpreter feature of Pygentic allows for multi-programming language support, enabling your agents to understand and generate responses in various programming languages. This expands the reach of your applications, making them compatible with a variety of different software projects. The flexibility to interpret and interact in multiple programming languages enhances the usability and effectiveness of your AI solutions.
Pygentic can be applied across numerous practical scenarios:
- Customer Support: Deploy intelligent agents to handle customer queries, providing instant and accurate responses.
- Content Generation: Use AI to create engaging content in multiple languages, tailored to your audience.
- Data Analysis: Leverage AI for interpreting and analyzing large datasets, extracting meaningful insights.
- Personal Assistants: Develop personalized AI assistants to help with daily tasks, scheduling, and information retrieval.
Pygentic empowers developers to harness the power of AI with ease, offering a flexible and scalable solution for integrating advanced language models into various applications. Its consistent API, multi-language support, and serverless capabilities make it a valuable tool for modern AI-driven projects.
Here's a detailed breakdown of the directory structure and code blocks for each file.
pygentic/
├── app/
│ ├── __init__.py
│ ├── main.py
│ ├── models.py
│ ├── routes/
│ │ ├── __init__.py
│ │ ├── assistants.py
│ │ ├── threads.py
│ │ ├── messages.py
│ │ ├── runs.py
│ └── services/
│ ├── __init__.py
│ ├── llm_service.py
│ ├── serverless_service.py
│ ├── open_interpreter_service.py
│ ├── database.py
│ ├── plugin_loader.py
│ ├── interfaces.py
├── tests/
│ ├── __init__.py
│ ├── test_assistants.py
│ ├── test_threads.py
│ ├── test_messages.py
│ ├── test_runs.py
├── connectors.toml
├── requirements.txt
└── README.md
You can install the Pygentic library using pip:
pip install pygentic
After installing the library, you can run the application using the following command:
pygentic
POST /v1/assistants
: Create a new assistant.GET /v1/assistants/{assistant_id}
: Retrieve an assistant by ID.- Note: The FastAPI UI now includes example JSON with test values for easier endpoint testing.
POST /v1/threads
: Create a new thread.GET /v1/threads/{thread_id}
: Retrieve a thread by ID.
POST /v1/threads/{thread_id}/messages
: Add a message to a thread.GET /v1/threads/{thread_id}/messages
: Retrieve messages in a thread.
POST /v1/threads/{thread_id}/runs
: Run a thread with an assistant.GET /v1/threads/{thread_id}/runs/{run_id}
: Retrieve a run by ID.
Pygentic uses a TOML configuration file to define and manage connectors for various services. This approach enhances flexibility, speed, concurrency, and stability by centralizing the configuration in an easily readable and editable format.
Create a connectors.toml
file in your project root with the following structure:
[connectors]
[connectors.database]
url = "sqlite:///./test.db"
[connectors.llm_service]
model = "gpt-3.5-turbo"
[connectors.serverless_service]
endpoint = "https://api.example.com/endpoint"
The Pygentic library automatically parses the connectors.toml file at startup. Ensure your configuration file is correctly formatted and placed in the root directory of your project.
Pygentic's connectors provide a flexible and modular way to integrate various services into your AI applications. They abstract the complexities of interacting with different APIs, including the OpenAI Assistants API, allowing you to connect endpoints to various services without altering the core code. This section explains how connectors work and how you can create custom connectors.
Connectors in Pygentic are designed to manage connections to various services such as databases, large language models (LLMs), and serverless endpoints. The default implementation includes an OpenAIAssistantConnector
that interacts with the OpenAI Assistants API. However, you can easily create custom connectors to integrate other services.
Pygentic includes several built-in connector classes:
- DatabaseConnector: Manages the connection to a database.
- LLMServiceConnector: Manages the connection to a large language model service.
- ServerlessServiceConnector: Manages the connection to a serverless endpoint.
- OpenAIAssistantConnector: Connects to the OpenAI Assistants API to handle various assistant operations.
These connectors are initialized with configurations from the connectors.toml
file and used within FastAPI middleware to make services available to your endpoints.
Pygentic's modular structure allows users to extend its functionality by creating custom connectors without modifying the core code. Follow these steps to create and use your custom connectors:
Create a new file in your custom connectors directory (e.g., custom_connectors/
) and define a class inheriting from BaseConnector
.
Example:
from app.services.interfaces import BaseConnector
class MyCustomConnector(BaseConnector):
def initialize(self, config: dict):
self.config = config
self.connected = False
async def connect(self):
self.connected = True
async def disconnect(self):
self.connected = False
async def generate_text(self, prompt: str) -> str:
if not self.connected:
await self.connect()
return "Generated text based on prompt: " + prompt
Add your custom connector configuration to the connectors.toml
file.
Example:
[connectors]
[connectors.database]
url = "postgres://user:password@localhost/dbname"
[connectors.llm_service]
model = "gpt-4"
[connectors.serverless_service]
endpoint = "https://example.com/serverless"
[connectors.mycustomconnector]
custom_setting = "value"
Set the environment variable PYGENTIC_PLUGIN_DIR
to point to your custom connectors directory, and then run the Pygentic application.
export PYGENTIC_PLUGIN_DIR=custom_connectors
pygentic
Your custom connector methods can be accessed within your FastAPI endpoints via request.state
.
@app.get("/custom_action")
async def custom_action(request: Request):
result = await request.state.mycustomconnector.generate_text("Hello, Pygentic!")
return {"result": result}
By following these steps, you can extend Pygentic's functionality with custom connectors tailored to your specific needs without modifying the core library code. This approach ensures that your integrations remain robust and adaptable to various services, providing a seamless and consistent API for your AI applications.
To switch the SQLite database to another persistence storage using the connectors system, modify the connectors.toml
file in your project root. Below is an example configuration for a PostgreSQL setup:
[connectors]
[connectors.database]
url = "postgresql://user:password@localhost/dbname"
[connectors.llm_service]
model = "gpt-3.5-turbo"
[connectors.serverless_service]
endpoint = "https://api.example.com/endpoint"
This configuration demonstrates how to change the database connector from SQLite to PostgreSQL. Ensure that your PostgreSQL server is running and accessible at the specified URL.