In VDP 101 [1/7] Introduction to VDP, we briefly introduce the high-level concept of VDP. To benefit a broader community, we release VDP under the open-source Apache license 2.0. Check it out here. We've made it easy to get started with VDP on your local machines and Kubernetes clusters.
Before installing VDP, Make sure you have the prerequisites set up:
- macOS or Linux - VDP works on macOS or Linux. It does not currently support Windows, but it is coming soon!
- Git - You need Git to download VDP and manage your code.
- Docker and Docker Compose - VDP uses Docker Compose to run all services locally.
- (Optional) NVIDIA Container Toolkit - To enable GPU support in VDP, this must be installed for Docker to recognise the GPUs.
Git is a distributed version control system: tracking changes in any set of files, usually used for coordinating work among programmers when developing source code. There are many ways to Install Git. Git provides different installation options for all three OS. Use the one you like the most. You can also install Git via Conda.
#Install Docker and Docker Compose
Docker is a set of platform as a service (PaaS) products that use OS-level virtualisation to deliver software in packages called containers. Users won't need to set up their environment before running services.
VDP uses Docker Compose (specifically,
Compose V2 and
Compose specification) to run all services locally.
Install Docker Desktop:
- on Mac - Be aware of your Mac model. It could come with an Intel chip or Apple silicon.
- on Linux - Make sure your platform is supported.
- on Windows - VDP does not support Windows yet. Stay tuned!
To enable GPU support in VDP, please refer to the installation guides of the NVIDIA Container Toolkit.
#Download and launch VDP
VDP is currently an open-source project maintained on our Github repository. Don't worry if you are not familiar with it. We have set up everything for you.
Execute the following
bash commands to start pre-built images with all the dependencies:
# clone VDP repository to localgit clone https://github.com/instill-ai/vdp.git && cd vdp# start VDP servicemake all
The installation may take a while. Have a cup of tea 🍵. Once the services are up, Voilà, the Console UI is ready to go at http://localhost:3000.
⚠︎ Don't panic if you don't see the Console UI right after
Downloading the model-backend and Triton Inference Server image will take a
while, but it should be a one-time effort.
↓↓ You should see the onboarding page. Enter your email, and you are all set! ↓↓
You can also check if VDP is running on your machine by listing the running compose project
docker compose ls --filter name=instill-vdpNAME STATUS CONFIG FILESinstill-vdp running < VDP Docker Compose file paths >
or by checking whether the
instill-vdp Docker Compose is running correctly in your Docker Desktop.
We highly recommend you allocate at least 32GB memory and 256GB disk
space to Docker for better stability. Go to
Settings > Resources of Docker
Desktop to set it up.
#Shut down VDP
If you want to shut down VDP, please run the
bash command below:
# shut down VDP servicesmake down
We love to hear your feedback and comments on VDP. Join us in our community on Discord or create a discussion on Github.
Okay, now you have VDP running on your machine. Let's create the first pipeline following our next tutorial → VDP 101 [3/7] Create Your First Pipeline on VDP.
↓↓↓ VDP 101 - Get familiar with the basics ↓↓↓