https://github.com/ashnkumar/theta-agents
https://github.com/ashnkumar/theta-agents
Last synced: 8 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/ashnkumar/theta-agents
- Owner: ashnkumar
- Created: 2024-07-30T16:01:11.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2024-08-01T21:10:30.000Z (almost 2 years ago)
- Last Synced: 2025-10-14T14:16:33.512Z (8 months ago)
- Language: Python
- Size: 15.6 KB
- Stars: 3
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Theta Agents SDK
The **Theta Agents SDK** is a Python library designed to help developers quickly build AI agent frameworks on the Theta EdgeCloud network. The SDK provides a range of prebuilt capabilities based on EdgeCloud's off-the-shelf containers, such as generating images and videos (Stable Diffusion), using EdgeCloud-hosted LLMs for underlying agent reasoning (e.g. Llama, Mistral) and writing code (CodeLlama), and Theta-specific capabilities like deploying to EdgeStore and the Theta Video API.
## Features
- **Create Images from Prompts**: Use EdgeCloud-hosted (or custom containers on EdgeCloud) image models like Stable Diffusion.
- **Create Videos from Images**: Generate videos using Stable Diffusion (Video) from EdgeCloud.
- **Generate Smart Contracts**: Easily create smart contracts based on user-defined prompts through CodeLlama hosted on EdgeCloud.
- **Deploy Smart Contracts**: Use web3-based agent tools to deploy to the Theta Testnet.
- **Upload files to Theta network's services**: Agents can upload files to EdgeCloud and videos to Theta's video services network.
- _**Coming soon:**_ Support for multi-agent workflows (2+ agents collaborating autonomously to solve tasks).
## Installation
To install the Theta Agents SDK, clone the repository and install the dependencies:
```bash
git clone https://github.com/yourusername/theta-agents-sdk.git
cd theta-agents-sdk
# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate
# Install dependencies
pip install -e .
```
## Setup
### 1. Environment Variables
Create a `.env` file in the root directory of your project. This file should contain all sensitive information such as API keys and private keys. Here's an example `.env` and you can look at `.env.example` for more:
```ini
# .env
HUGGING_FACE_TOKEN=your-hugging-face-token
GRADIO_API_KEY=your-gradio-api-key
BLOCKCHAIN_PRIVATE_KEY=your-blockchain-private-key # if deploying to Theta's chain.
```
### 2. Configuration File
Copy `config_template.yaml` to `config.yaml` and edit the values to match your specific setup:
```bash
cp config_template.yaml config.yaml
```
Edit `config.yaml` to include your specific configuration details, such as endpoints, model names, and environment variable names.
```yaml
llm_endpoint: "https://gemma2b....onthetaedgecloud.com/v1/chat/completions"
llm_model_name: "google/gemma-2b"
capabilities:
image_tools:
create_image_from_prompt:
edgecloud_endpoint: "https://stablediffi...onthetaedgecloud.com"
edgecloud_endpoint_type: "gradio"
video_tools:
create_video_from_image:
edgecloud_endpoint: "https://stablediffi...onthetaedgecloud.com"
edgecloud_endpoint_type: "gradio"
smart_contract_tools:
generate_smart_contract:
edgecloud_endpoint: "..."
edgecloud_endpoint_type: "openai"
model_name: "..."
```
## Usage
### Running the Chat Interface
To interact with an AI agent using the terminal chat interface, you can use the following code snippet in your script:
```python
from theta_agents import ThetaAgent, create_image_from_prompt, create_video_from_image
# Load environment variables
load_dotenv()
# Initialize the agent
agent = ThetaAgent(
capabilities=[create_image_from_prompt, create_video_from_image],
show_planning=True,
persona="You are a social media marketing expert."
)
def chat(agent: ThetaAgent):
"""
Function to interact with the ThetaAgent in a chat-like interface in the terminal.
"""
while True:
user_input = input("\nYou: ")
if user_input.lower() in ["quit", "exit"]:
print("Exiting the chat.")
break
# Get the response from the agent
response_data = agent(user_input)
# Extract and display the user-facing message and internal planning message
human_facing_text = response_data.get('user_facing_text', '') # The text meant for the user
planning_text = response_data.get('planning_text', '') # Internal planning details
error_message = response_data.get('error', '') # Any error messages
if human_facing_text:
print(f"\033[96mAI:\033[0m {human_facing_text}") # Cyan text for AI messages
if planning_text:
print(f"\033[93mInternal Plan:\033[0m {planning_text}") # Yellow text for internal planning messages
if error_message:
print(f"\033[91mError:\033[0m {error_message}") # Red text for errors
# Start the chat
chat(agent)
```
### Additional Capabilities
To add more capabilities (tools), update the `config.yaml` with the new tool configurations and ensure that the necessary environment variables are set.
## Contributing
We welcome contributions! Please follow these steps to contribute:
1. Fork the repository.
2. Create a new branch (`git checkout -b feature-branch-name`).
3. Make your changes and commit them (`git commit -m 'Add new feature'`).
4. Push to the branch (`git push origin feature-branch-name`).
5. Open a pull request.
Please ensure that your code adheres to our coding standards and passes all tests.
## License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
## Acknowledgements
Special thanks to the open-source community for providing the tools and libraries that make this project possible.