Installation in AWS#

1. Initialize EC2 instance#

  • In the AWS console, go to the service EC2 and select one of the following zones: N. Virginia (us-east-1), Paris (eu-west-3), or Singapore (ap-southeast-1)

  • Launch an EC2 instance

2. Configure your EC2 instance#

  • Application and OS image: Select the default Ubuntu server 22.04 LTS 64-bit (x86)

  • Instance type: We recommend you to choose at least a t2.large instance type (2vCPU, 8GB memory)

  • Key pair: Choose your usual key pair. If you donโ€™t have one, go to the Amazon document to create the right one

  • Network settings: You need to **open the port 19000 to access the Giskard frontend and upload your model. To do so, click on Edit and add the following security groups:

  • Storage: Choose a minimum of 30 Gigs of SSD (this will mainly depend on the size of your datasets)

3. Launch the instance and install Giskard#

  • Click on Launch instance to create the instance

  • Connect in SSH to your instance. You can for example use the EC2 Instance connect to open a terminal directly in your AWS platform

  • Installation of the Giskard requirements (docker)

 curl -fsSL -o
 sudo sh
  • Installation of Giskard

giskard server start

4. Connect to your instance and start uploading an ML model#

  • Get your IP address: Click on the ID of the instance you just created and copy its Public IPv4 address (or Public IPv4 DNS)

  • Go to http://<your IP address>:19000 in your web browser

  • The user id is admin and the password is admin

Thatโ€™s it, you are now ready to use Giskard in AWS! Now you can start uploading an artifact!


You can stop the instance and restart it when you need to save AWS compute costs. However, note that the IP address will not necessarily be the same. So make sure you copy it again when itโ€™s launched