Midori AI Subsystem Manager

Midori AI photo Midori AI photo

How Docker Works

Docker is a containerization platform that allows you to package and run applications in isolated and portable environments called containers. Containers share the host operating system kernel but have their own dedicated file system, processes, and resources. This isolation allows applications to run independently of the host environment and each other, ensuring consistent and predictable behavior.

Midori AI Subsystem - Github Link

The Midori AI Subsystem extends Docker’s capabilities by providing a modular and extensible platform for managing AI workloads. Each AI system is encapsulated within its own dedicated Docker image, which contains the necessary software and dependencies. This approach provides several benefits:

  • Simplified Deployment: The Midori AI Subsystem provides a streamlined and efficient way to deploy AI systems using Docker container technology.
  • Eliminates Guesswork: Standardized configurations and settings reduce complexities, enabling seamless setup and management of AI programs.
Notice

Warnings / Heads up

  • This program is in beta! By using it you take on risk, please see the disclaimer in the footnotes
  • The Webserver should be back up, sorry for the outage

Known Issues

  • AnythingLLM Backend seems to be not upserting files, please report if you are having this bug to the github
  • Report Issuses -> Github Issue

Windows Users

Install Midori AI Subsystem Manager

Notice
  • As we are in beta, we have implemented telemetry to enhance bug discovery and resolution. This data is anonymized and will be configurable when out of beta.

Prerequisites

Docker Desktop Windows

Please make a folder for the Manager program with nothing in it, do not use the user folder.

Quick install

  1. Download - https://tea-cup.midori-ai.xyz/download/model_installer_windows.zip
  2. Unzip into LocalAI folder
  3. Run subsystem_manager.exe

Quick install with script

Open a Command Prompt or PowerShell terminal and run:

curl -sSL https://raw.githubusercontent.com/lunamidori5/Midori-AI/master/other_files/model_installer/shell_files/model_installer.bat -o subsystem_manager.bat && subsystem_manager.bat

Manual download and installation

Open a Command Prompt or PowerShell terminal and run:

curl -sSL https://tea-cup.midori-ai.xyz/download/model_installer_windows.zip -o subsystem_manager.zip
powershell Expand-Archive subsystem_manager.zip -DestinationPath .
subsystem_manager.exe

Prerequisites

Docker Desktop Linux

or

Docker Engine and Docker Compose

Quick install with script

curl -sSL https://raw.githubusercontent.com/lunamidori5/Midori-AI/master/other_files/model_installer/shell_files/model_installer.sh | sh

Manual download and installation

Open a terminal and run:

curl -sSL https://tea-cup.midori-ai.xyz/download/model_installer_linux.tar.gz -o subsystem_manager.tar.gz
tar -xzf subsystem_manager.tar.gz
chmod +x subsystem_manager
./subsystem_manager

Prerequisites

Download and set up Docker Compose Plugin

Manual download and installation

Copy and paste this into the Docker Compose Manager plugin

services:
  midori_ai_unraid:
    image: lunamidori5/deb11_subsystem_manager:master
    ports:
    - 39090:9090
    privileged: true
    restart: always
    tty: true
    volumes:
    - /var/lib/docker/volumes/midoriai_midori-ai-models/_data:/var/lib/docker/volumes/midoriai_midori-ai-models/_data
    - /var/lib/docker/volumes/midoriai_midori-ai-images/_data:/var/lib/docker/volumes/midoriai_midori-ai-images/_data
    - /var/lib/docker/volumes/midoriai_midori-ai-audio/_data:/var/lib/docker/volumes/midoriai_midori-ai-audio/_data
    - /var/run/docker.sock:/var/run/docker.sock
volumes:
  midori-ai:
    external: false
  midori-ai-audio:
    external: false
  midori-ai-images:
    external: false
  midori-ai-models:
    external: false

Start up that docker then run the following in it by clicking console

curl -sSL https://raw.githubusercontent.com/lunamidori5/Midori-AI/master/other_files/midori_ai_manager/subsystem_python_runner.py > subsystem_python_runner.py && python3 subsystem_python_runner.py

Running the program

Start up that docker then run the following in it by clicking console

python3 subsystem_python_runner.py

Prerequisites

Do not use on windows

Please make a folder for the Manager program with nothing in it, do not use the user folder.

Quick install with script

Download this file

curl -sSL https://raw.githubusercontent.com/lunamidori5/Midori-AI/master/other_files/midori_ai_manager/subsystem_python_runner.py > subsystem_python_runner.py && python3 subsystem_python_runner.py

Running the program

Open a terminal and run:

python3 subsystem_python_runner.py

Auto Lint, Test, and Build. Auto Lint, Test, and Build.

Notice

Reminder to always use your computers IP address not localhost when using the Midori AI Subsystem!

Support and Assistance

If you encounter any issues or require further assistance, please feel free to reach out through the following channels:

—– Install Backends —–

  • LocalAI - For LLM inference, Embedding, and more
  • AnythingLLM - For chating with your docs using LocalAI or other LLM hosts
  • Big-AGI - For chating with your docs using LocalAI or other LLM hosts
  • InvokeAI - For making photos using AI models
  • Axolotl - For training your own fine tuned LLMs

Check out our Model Repository for info about the models used and supported!

—– FAQs about the subsystem —–

  1. What is the purpose of the Midori AI Subsystem?

    • The Midori AI Subsystem is a modular and extensible platform for managing AI workloads, providing simplified deployment, standardized configurations, and isolation for AI systems.
  2. How does the Midori AI Subsystem simplify AI deployment?

    • The Midori AI Subsystem simplifies AI deployment by providing a streamlined and efficient way to deploy AI systems using Docker container technology, reducing complexities and ensuring consistent and predictable behavior.
  3. What are the benefits of using the Midori AI Subsystem?

    • The benefits of using the Midori AI Subsystem include simplified deployment, standardized configurations, isolation for AI systems, and a growing library of supported backends and tools.
  4. What are the limitations of the Midori AI Subsystem?

    • The limitations of the Midori AI Subsystem include its current beta status, potential for bugs, and reliance on Docker container technology.
  5. What are the recommended prerequisites for using the Midori AI Subsystem?

    • The recommended prerequisites for using the Midori AI Subsystem include Docker Desktop Windows or Docker installed on other operating systems, and a dedicated folder for the Manager program.
  6. How do I install the Midori AI Subsystem Manager?

  7. Where can I find more information about the Midori AI Subsystem?

    • You can find more information about the Midori AI Subsystem on the Midori AI Subsystem website, which includes documentation, tutorials, and a community Discord.
  8. What is the difference between the Midori AI Subsystem and other AI frameworks?

    • The Midori AI Subsystem differs from other AI frameworks by providing a modular and extensible platform specifically designed for managing AI workloads, offering features such as simplified deployment, standardized configurations, and isolation for AI systems.
  9. How does the Midori AI Subsystem handle security?

    • The Midori AI Subsystem does not handle security directly, but it relies on the security features provided by the underlying Docker container technology and the specific AI backends and tools being used.
  10. What are the plans for future development of the Midori AI Subsystem?

  • The plans for future development of the Midori AI Subsystem include adding new features, improving stability and performance, and expanding the library of supported backends and tools.

Questions from Carly Kay

—– Disclaimer —–

The functionality of this product is subject to a variety of factors that are beyond our control, and we cannot guarantee that it will work flawlessly in all situations. We have taken every possible measure to ensure that the product functions as intended, but there may be instances where it does not perform as expected. Please be aware that we cannot be held responsible for any issues that arise due to the product’s functionality not meeting your expectations. By using this product, you acknowledge and accept the inherent risks associated with its use, and you agree to hold us harmless for any damages or losses that may result from its functionality not being guaranteed.

—– Footnotes —–

*For your safety we have posted the code of this program onto github, please check it out! - Github

**If you would like to give to help us get better servers - Give Support

***If you or someone you know would like a new backend supported by Midori AI Subsystem please reach out to us at [email protected]