Skip to content

Deploy AI Models

1. Deploy AI Models

There are 3 ways to distribute artificial intelligence to inference servers.

1. Deploying AUTOML of CIick AI

2. Loading and distributing external artificial intelligence models to inference servers

3. Create and distribute artificial intelligence through custom learning by renting a learning server from Click AI

Please check [CLICK AI → Use custom learning] to rent a learning server from Click AI and generate and distribute artificial intelligence through custom learning.

On this page, we will find out how to distribute developed artificial intelligence through inference servers.

1-1. Utilize AI using CLICK AI's AutoML

You can load an artificial intelligence model developed by Click AI's AutoML into the inference server and use SKYHUBAI.


이미지이름

1) Click AutoML button under "SKYHUB AI|DEPLOY".

2) Find the project name to deploy model.


이미지이름

3) Verify that the AI generation of the project you want to distribute is completed.

4) Check the various models created within the project, and choose the one you want to deploy via SKYHUB AI. Select the model's 'Deploy'.


이미지이름

5) Click Create Server by checking cloud providers, regions, and server specifications selections and usage fees. - With native support for AWS cloud servers, Goggle Cloud and Azure can contact the sales team separately before renting the server.

6) With a new project generation notification pop-up, you can instantly view dashboards for your AI deployment.

1-2. Load external AI models

You can load and distribut external artificial intelligence models to inference servers


이미지이름

1) Click "SKYHUB AI|DEPLOY".


이미지이름

2) To bring up an external artificial intelligence model, select 'new start' on the first screen of SKYHUB AI.


이미지이름

3) After comparing the characteristics of CLICK AI and SKYHUB AI, select SKYHUB AI's 'Upload Model'.

4) SKYHUBAI's model load supports PyTorch and TensorFlow2. Drag and upload files of the appropriate format for pth and zip, or click the box to upload them.

5) Once the file upload is confirmed, press OK to move on.


이미지이름

6) Click Create Server by checking cloud providers, regions, and server specifications selections and usage fees.
- With native support for AWS cloud servers today, Google Cloud and Azure can contact the sales team separately before renting the server.

7) With a new project generation notification pop-up, you can instantly view dashboards for your AI deployment.