Scribd Logo
Upload

Welcome to Scribd!

  • 13 views

    Ubuntu 22.04 PrivateGPT Instructions

    Uploaded by

    d28900835

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Ubuntu 22.04 PrivateGPT Instructions

    Uploaded by

    d28900835
    13 views4 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    13 views4 pages

    Ubuntu 22.04 PrivateGPT Instructions

    Uploaded by

    d28900835

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 4

    ##Create VM with at least 50GB of space

    ##Download/Install Ubuntu LTS 22.04 (Minimal Install)

    sudo apt update


    sudo apt upgrade

    ##Nvidia Driver/Cuda Installation

    ##Blacklist nouveau driver


    sudo nano /etc/modprobe.d/blacklist-nouveau.conf

    ##add to end of file


    blacklist nouveau
    options nouveau modeset=0

    #Update
    sudo update-initramfs -u

    ##Reboot VM
    reboot

    ##Grab the current driver for your card (ie 470 for my Tesla P4)
    sudo ubuntu-drivers devices

    sudo apt install nvidia-driver-470


    reboot

    ##Verify drivers
    nvidia-smi

    ##Verify gcc installed


    gcc —version

    ##Install CUDA Toolkit


    wget
    https://developer.download.nvidia.com/compute/cuda/repos/ub
    untu2204/x86_64/cuda-keyring_1.1-1_all.deb
    sudo dpkg -i cuda-keyring_1.1-1_all.deb
    sudo apt-get update
    sudo apt-get -y install cuda
    reboot
    ##Add Environment Variable
    sudo nano ~/.bashrc

    ##add at the bottom


    export PATH=/usr/local/cuda/bin${PATH:+:${PATH}}
    export LD_LIBRARY_PATH=/usr/local/cuda-
    12.3/lib64\${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
    Source ~/.bashrc

    ##Verify cuda install


    nvcc -V

    ##PrivateGPT installation

    ##Install Python3.11

    sudo add-apt-repository ppa:deadsnakes/ppa


    sudo apt install python3.11 python3.11-venv git pip -y
    sudo update-alternatives --install /usr/bin/python3 python3
    /usr/bin/python3.10 110
    sudo update-alternatives --install /usr/bin/python3 python3
    /usr/bin/python3.11 100
    sudo update-alternatives --config python3
    select option2 (3.11)

    ##Verify version 3.11


    python3 --version

    ##verify pip
    python3 -m pip install pip

    ##Install poetry
    pip install poetry

    ##Add Environment Variable


    sudo nano ~/.bashrc

    ##add to end of file


    export PATH=“/home/{insert-username-here}/.local/bin:$PATH”
    ##update source
    source ~/.bashrc

    ##Verify Poetry Version (1.7)


    poetry --version

    ##Setup privateGPT
    git clone https://github.com/imartinez/privateGPT

    ##Set permissions on privateGPT


    sudo chown -R {insert-username-here} privateGPT

    ##Change Directory
    cd privateGPT

    ##Create/enter python environment


    python3 -m venv venv
    source venv/bin/activate
    poetry install --with ui,local
    poetry run python scripts/setup
    CMAKE_ARGS='-DLLAMA_CUBLAS=on' poetry run pip install --
    force-reinstall --no-cache-dir llama-cpp-python

    ##Optional

    ##Make BASH script to run on boot (There is probably a better way to do this
    and if someone wants to provide it ill update this)

    sudo nano /privateGPT.sh

    #!/bin/bash
    cd /home/{put-username-here}/privateGPT
    python3 -m venv venv
    source venv/bin/activate
    PGPT_PROFILE=local make run

    ##Change ownership and permissions on file


    sudo chmod +x /privateGPT.sh
    sudo chown root /privateGPT.sh
    ##Add bash script to cron to run on startup
    Crontab -e
    ##Choose nano and add to the end
    @reboot /privateGPT.sh

    ##To stop the service if its running in the background


    ps -ef | grep python
    ##take note of the process number
    Kill -9 {insert-process-number}

    ##If you run into memory issues when you are trying to ingest files into GPT
    ##Kill the process from the previous step

    cd ~/privateGPT
    nano settings.yaml

    ##Edit the “embedding” mode from local to mock


    ##Ingest the files you want
    ##Then edit the file again back to local

    ##If you ingest files and need to remove them you can erase them all with
    Make wipe

    You might also like