Note
If you would like to schedule an evaluation of LEIP, please contact us at support@latentai.com
Step 1: Access our Container Repository
In order to install packages or pull any of our Docker images, you’ll need to create a personal access token. To do so, follow these steps:
- Login to Latent AI Artifact repository.
- Click the
Sign-in
link in the upper right. - Select
Sign-in with SSO
. - Enter your access credentials.
- Click the
- Create your Personal Access Token.
- Click on your profile in the upper right.
- Select
User Token
on the left navigation. - Select the
Access User Token
button. - View your user token name and user token passcode.
- Export your user token name and passcode.
Step 2: Install LEIP SDK
Tip
You can make
LICENSE_KEY
andLEIP_WORKSPACE
persist for every newbash
session by placing the variables in~/.bashrc
.
There are two components to the LEIP SDK, LEIP Design and LEIP Optimize.
There is also two methods of installing the SDK. Depending on your path (Bring Your Own Data (BYOD) or Bring Your Own Model (BYOM)), you will either just need
Prerequisites
Regardless of your installation method, make sure you configure the following:
- Clone
leip-tutorials
and go into theenvironment
directory. export
the following system variables:LICENSE_KEY
- You should have received this via email.LEIP_WORKSPACE
- The default path where all artifacts are placed as you use the SDK.
- Login to our Docker repository. Then use the user credentials obtained from Step 1 Access our Container Repository. You’ll have to manually copy and paste the variables listed below.
Method 1: Entirely Docker
The easiest way to install the LEIP SDK is by using docker compose
to initialize both LEIP Design and LEIP Optimize.
Method 2: Local Environment with LEIP Optimize
WIP
Step 3: Use LEIP
If you opted in for Method 1: Entirely Docker, a Jupyter Notebook instance will automatically start.
You can access this notebook in your browser via http://127.0.0.1:8888. The Jupyter password is leip
.
To get started with using LEIP, view ourtutorials .
Once you have optimized your model and you’re ready for deployment, refer to Installing LRE.
Additional Configurations
These are additional configurations in the even the default instructions above need to be modified.
Use LEIP via SSH
If you have deployed LEIP on a remote instance or server, you can tunnel the Jupyter instance back to your local machine with either method below.
Inline Terminal Command
Stored as an SSH Bookmark
You can store these SSH settings in ~/.ssh/config
:
Then, simply type in ssh bookmark_name
to connect.
Using Different Ports for LEIP
NOTE
In the event you need to use different ports for LEIP, make the following changes to the YAML file in leip-tutorials/environment
.
docker-compose.cpu.yml
networks:
...
volumes:
...
services:
leip-cf:
...
...
ports:
- CHANGE_LOCAL_PORT_TO_ACCESS_LEIP_CF:8888
...
leip-af:
...
...
ports:
- CHANGE_LOCAL_PORT_TO_JUPYTER_NOTEBOOK:8888
...
Changing the Docker Virtual Network Name for LEIP
Change the network name in the following YAML file in leip-tutorials/environment
.
docker-compose.yml
networks:
default:
name: CHANGE_NETWORK_NAME_HERE
volumes:
...
Spawning Multiple LEIP Instances
In the event you need to have multiple LEIP instances on a single server, make sure to:
- Change the ports mentioned in Using Different Ports for LEIP to avoid port conflict with other LEIP instances during initialization.
- Change the virtual network name in Changing the Docker Virtual Network Name for LEIP to avoid network name conflicts during initialization.
Once those steps are done, start an instance with the -p
flag to add a prefix to the beginning of the container names. This will create create the containers needed with the prefix in front to avoid container name conflicts with other LEIP instances.
The command above will spawn containers named new_user-leip-af
and new_user-leip-cf
.