- Overview
- Getting Started Guide
- UserGuide
-
References
-
ABEJA Platform CLI
- CONFIG COMMAND
- DATALAKE COMMAND
- DATASET COMMAND
- TRAINING COMMAND
-
MODEL COMMAND
- check-endpoint-image
- check-endpoint-json
- create-deployment
- create-endpoint
- create-model
- create-service
- create-trigger
- create-version
- delete-deployment
- delete-endpoint
- delete-model
- delete-service
- delete-version
- describe-deployments
- describe-endpoints
- describe-models
- describe-service-logs
- describe-services
- describe-versions
- download-versions
- run-local
- run-local-server
- start-service
- stop-service
- submit-run
- update-endpoint
- startapp command
-
ABEJA Platform CLI
- FAQ
- Appendix
Create an inference API without coding
Introduction
This page explains how to use “ABEJA Template”. This feature can perform machine learning without coding. This page uses the task of Image Classification.
Prepare a training model
Firtst of all, Prepare a training model. How to create a training model here
Clicking on “Model” reflects the result of the previous learning job. Use this model to deploy the inference model API.
Create a deployment
First, create a frame (deployment) for creating the inference API. The deployment manages the model created by training and the code for performing inference, and provides API services.
Select “Deployment” from the left menu. Then select “Create Deployment”.
When creating a deployment, you can enter the Deployment Name and Details. After entering, select “Create Deployment”.
You have a frame for managing your deployment now.
Create inference code
Next, manage the code to perform the inference. Select “Code” from the left menu.
You can bring in your inference code or use the template provided by ABEJA. This time, we will create an inference code using a template. Select “Create Version”.
Select “Template”, enter the version to be managed, and select “Create”.
Now that you have created the “code” that will be used for deployment, you are ready to deploy the API service.
Create HTTP service
Next, This section explains how to create an HTTP service. Deploy the API service using the created “Code” and “Model”.
Select “Service” from the left menu, and select “Create HTTP Service”.
Next, select the version, model, version, instance type, number of instances of the created code, and select “Create HTTP service”.
After a few minutes, the status will be “Available” and the model will be deployed as a Web API. After training was complete, the model could be deployed in just a few minutes.
Operation check
Click “✔︎confirmation” Click the URL to put a link to the Wikipedia photo.
When you click submit, the inference results will be returned. With this learning model, the probability is 98.99%, and it is predicted to be DANDELION (dandelion).
