             █▀█                  
          ▄▄▄▀█▀            
          █▄█ █    █▀█        
       █▀█ █  █ ▄▄▄▀█▀      
    ▄▄▄▀█▀ █  █ █▄█ █ ▄▄▄       
    █▄█ █  █  █  █  █ █▄█        ███╗   ██╗ █████╗ ██████╗ ████████╗██╗  ██╗ █████╗ 
 ▄▄▄ █  █  █  █  █  █  █ ▄▄▄     ████╗  ██║██╔══██╗██╔══██╗╚══██╔══╝██║  ██║██╔══██╗
 █▄█ █  █  █  █▄█▀  █  █ █▄█     ██╔██╗ ██║███████║██████╔╝   ██║   ███████║███████║
  █  █   ▀█▀  █▀▀  ▄█  █  █      ██║╚██╗██║██╔══██║██╔═══╝    ██║   ██╔══██║██╔══██║
  █  ▀█▄  ▀█▄ █ ▄█▀▀ ▄█▀  █      ██║ ╚████║██║  ██║██║        ██║   ██║  ██║██║  ██║
   ▀█▄ ▀▀█  █ █ █ ▄██▀ ▄█▀       ╚═╝  ╚═══╝╚═╝  ╚═╝╚═╝        ╚═╝   ╚═╝  ╚═╝╚═╝  ╚═╝
     ▀█▄ █  █ █ █ █  ▄█▀                         Decentralized Multi-Agent Workflows
        ▀█  █ █ █ █ ▌▀                                                 www.naptha.ai
          ▀▀█ █ ██▀▀

Naptha Python SDK

Naptha enables users to build decentralized multi-agent workflows. Decentralized workflows can run on one or more nodes (rather than on one central server), with different LLMs, and with many local data sources, opening up new use cases for AI devs.

Here's Yohei (creator of BabyAGI) admitting that BabyAGI isn't a true multi-agent system since the agents use the same LLM and code base. You can watch a demo video where we run BabyAGI as a true multi-agent system here.

Install

From source:

git clone https://github.com/NapthaAI/naptha-sdk.git
cd naptha-sdk
poetry install
poetry shell

Create a copy of the .env file:

cp .env.example .env

You will need to add a value to PRIVATE_KEY in .env. You can generate and output one to the commandline using naptha user (just copy and paste the value into the .env file).

Get Started

Nodes

Explore Nodes on the Network

You can use the CLI to see a list of available nodes:

naptha nodes

Make note of a Node ID for running a workflow below.

Check out available Modules

Modules can be workflows, agents or multi-agent systems. Modules typically involve one or more LLMs, along with humans in the loop. You can also use the CLI to explore available modules that you can run on a node:

naptha modules

For each module, you will see a url where you can check out the code.

Run a Module

Now you've found a node and a module you'd like to run, so let's run it locally! You can use the commandline tool to connect with the node and run the workflow.

# usage: naptha run <module_name> <module args>
naptha run hello_world -p "param1=world param2=naptha"

Try a module that uses the local LLM running on your node:

naptha run chat -p "prompt='tell me a joke'"

Try a module that makes predictions about future events using:

naptha run olas_prediction -p "prompt='Will there be an initial public offering on either the Shanghai Stock Exchange or the Shenzhen Stock Exchange before 1 January 2016?'"

You can also try a module that generates images (make sure that the .env file in node has a valid Stability platform API key):

naptha run generate_image -p "prompt='Beautiful green mountains and clear blue skies. Sun shining and birds chirping. A perfect day for a hike. You are walking through the forest, enjoying the scenery, when you come across a fork in the road. Do you go left or right?'"

Now let's run an image-to-image model on this image:

naptha run image_to_image -p "prompt='Cyberpunk with a wolf' input_dir=<module_run_id_1>"

You can also run modules from yaml files using:

naptha run create_profile_description -f ./example_yamls/create_profile_description.yml

Or docker images:

naptha run docker_hello_world -f ./example_yamls/docker_hello_world.yml

Interact with Node Storage

After the module runs finish, you can download the file from the node using:

naptha read_storage -id <module_run_id>

You can write to the node using:

naptha write_storage -i files/<filename>.jpg

Interact with IPFS thorugh Node

naptha write_storage -i files/<filename>.jpg --ipfs

Run Multi-Node Workflows

naptha run multiplayer_chat -p "prompt='i would like to count up to ten, one number at a time. ill start. one.'" --worker_nodes "http://node.naptha.ai:7001,http://node1.naptha.ai:7001"

naptha run babyagi -p "objective='Research the history of football'" --worker_nodes "http://node.naptha.ai:7001,http://node1.naptha.ai:7001"

Using the SDK non-interactively

To use the SDK as part of a script, start with importing the hub and service subcomponents.

import asyncio
from naptha_sdk.client.naptha import Naptha
from naptha_sdk.client.node import Node
from naptha_sdk.task import Task
from naptha_sdk.flow import Flow
from naptha_sdk.user import generate_user
import os

async def main():
  naptha = await Naptha(
      user=generate_user()[0],
      hub_username=os.getenv("HUB_USER"), 
      hub_password=os.getenv("HUB_PASS"), 
      hub_url="ws://node.naptha.ai:3001/rpc",
      node_url="http://node.naptha.ai:7001",
  )

  flow_inputs = {"prompt": 'i would like to count up to ten, one number at a time. ill start. one.'}
  worker_nodes = [Node("http://node.naptha.ai:7001"), Node("http://node1.naptha.ai:7001")]

  flow = Flow(name="multiplayer_chat", user_id=naptha.user["id"], worker_nodes=worker_nodes, module_params=flow_inputs)

  task1 = Task(name="chat_initiator", fn="chat", worker_node=worker_nodes[0], orchestrator_node=naptha.node, flow_run=flow.flow_run)
  task2 = Task(name="chat_receiver", fn="chat", worker_node=worker_nodes[1], orchestrator_node=naptha.node, flow_run=flow.flow_run)

  response = await task1(prompt=flow_inputs["prompt"])

  for i in range(10):
      response = await task2(prompt=response)
      response = await task1(prompt=response)

asyncio.run(await main())

Run a Node

You can run your own Naptha node, and earn rewards for running workflows. Follow the instructions at https://github.com/NapthaAI/node

Community

Bounties and Microgrants

Have an idea for a cool use case to build with our SDK? Get in touch at [email protected].

richardblythman
/

mas_multiplayer_chat