Skip to content

Latest commit

 

History

History
131 lines (93 loc) · 4.94 KB

README.md

File metadata and controls

131 lines (93 loc) · 4.94 KB
Raw

Installation on Azure VM

Requirements

For this installation, we recommand to host User interface (UI), Backend server and Ollama (LLM server) on 3 seperate VMs. The Ollama VM should have a GPU with CUDA installed on the server/VM.

For the demo purpose, we have hosted MyGPT on following Azure VMs. You can use the same configuration or modify it as per your requirements.:

MyGPT Component Azure resource name vRAM vCPU Disk Space OS Additional Hardware/ Software
Frontend (UI) server Premium_v3_P1V3 8GB 2 250GB Windows Node 14 LTS
Backend server Standard_E4ads_v5 32GB 4 150GB Ubuntu 24.04 LTS Docker, Docker Compose
Ollama (LLM server) Standard_NC8as_T4_v3 32GB 8 56GB Ubuntu 24.04 LTS NVIDIA GPU - Tesla T4

Installation

Frontend (UI) server

  1. Build the frontend on your local machine

    git clone https://github.com/stjude-c3d/MyGPT.git
cd MyGPT/frontend
npm install
npm run build

  2. Copy the build files to the server Create an azure website using the official documentation: Create an Azure Web App

    And copy the build files from your local machine to the Azure Web App root folder - site/wwwroot.

  3. Access the UI on the browser

    http://<vm-ip/vm-url>/

Backend server

  1. Install Docker and Docker Compose

    Check if you have docker installed on your system by running following command.

    docker --version

    If you get an error that docker not found, Go to the official Docker installation page for Linux and install the appropriate Docker on your system: https://docs.docker.com/desktop/install/linux-install/

  2. Clone the backend repository If you have zip file with MyGPT installation instructions, you can unzip it and copy it on your desktop or your desired location. You can skip remaining instructions from below and go to Step 2.

    Or you can get installation instructions by running following command from Terminal. It will create MyGPT folder on your Desktop or desired folder. To get this from GitHub, we will need GitHub username and GitHub access token as the GitHub repo is private. When you run the following command, it will ask for this credentials.

    git clone https://github.com/stjude-c3d/MyGPT.git

  3. Edit the .env file Edit the .env file in the backend folder and set the following

    cd MyGPT/backend
vi env_example
    SECRET_KEY = 'add django key'
ZOTERO_API_KEY = 'Add zotero key here'
DJANGO_SUPERUSER_PASSWORD = 'add your django superuser password'
HUGGINGFACE_API_KEY = 'add your hugging face api key'
OLLAMA_SERVER = 'add your ollama server URL'

    Add necessary keys values in the env_example file and save it. The docker compose will copy and rename this file to .env file and save it in the backend folder.

  4. Edit the settings.py file in Django app Edit the settings.py file in the backend folder and change the following fields.

    • CORS_ALLOWED_ORIGINS (line 37): CORS_ALLOWED_ORIGINS = ['https://<frontend-url>']
    • CSRF_TRUSTED_ORIGINS (line 41): CSRF_TRUSTED_ORIGINS = ['https://<backend-url>']

  5. Build the backend server To build the backend server, go to the backend folder and run the following command.

    cd MyGPT
docker-compose build backend

  6. Edit nginx configuration Edit the nginx.conf file in the nginx folder and change dafault.conf file for following fields.

    • server_name (line 9): server_name <server-url>;
    • return 301 (line 23): return 301 $request_uri;`
    • server_name (line 44): server_name <server-url>;
    • ssl_certificate (line 46): ssl_certificate <path-to-.pem-file>;
    • ssl_certificate_key (line 47): ssl_certificate_key <path-to-.key-file>;

  7. Build nginx server To build the nginx server, go to the backend folder and run the following command.

    cd MyGPT
docker-compose build nginx

  8. Run the backend server To run the backend server, go to the backend folder and run the following command.

    cd MyGPT
docker-compose up -d db
docker-compose up -d backend
docker-compose up -d nginx

    You can access the backend server on the browser using the following URL.

    http://<server-url>

Ollama (LLM server)

To install Ollama on the server, follow the instructions in the Ollama installation guide.

We also recommand to change envrironment variables in the ollama folder and set the following by following guide from Ollama

How do I configure Ollama server?

We also securely hosted Ollama server using Nginx. You can follow the instructions in the Nginx configuration guide for Ollama