rag-assistant 0.2.5.dev1

This repository includes rag_assistant

Swarmauri Rag Assistant

Overview

The Swarmauri Rag Assistant is your go-to tool for managing different configurations and settings for your application. With a simple command-line interface, you can customize and deploy the Assistant using various arguments according to your needs.

Installation

Getting started with Swarmauri’s RAG Assistant is straightforward. You can install it via pip by running the following command:

pip install rag_assistant==0.2.0 --user

Usage

Below is a comprehensive explanation of each argument you can provide to the Assistant and how to use them effectively:

Required Arguments

Generating config files

• For json config files

rag_assistant generate -o config.json

• For yaml config files

rag_assistant generate -o config.yaml

Launching app

• To launch the app:

rag_assistant launch --api_key $OPENAI_API_KEY --provider-llm openai

• N/B: to see the currently supported LLMs do: rag_assistant launch --help under --provider-llm

For more help on rag_assistant:

rag_assistant --help

Example results

Swarmauri Developer Assistant

positional arguments:
  {generate,launch}
    generate         Generate a configuration file
    launch           Launch the Gradio UI application

options:
  -h, --help         show this help message and exit

For more help on rag_assistant generate

rag_assistant generate --help

Example results

options:
  -h, --help            show this help message and exit
  -o OUTPUT, --output OUTPUT
                        Output file path

For more help on rag_assistant launch

rag_assistant launch --help

Example result

options:
  -h, --help            show this help message and exit
  -api_key API_KEY, --api_key API_KEY
                        Your API key
  -provider_llm PROVIDER_LLM, --provider_llm PROVIDER_LLM
                        Your provider LLM: openai | groq | mistral | gemini | anthropic
  -provider_model PROVIDER_MODEL, --provider_model PROVIDER_MODEL
                        Your provider model
  -config_file CONFIG_FILE, --config_file CONFIG_FILE
                        Path to config file