What is the difference between "tao-toolkit-api-app" and "tao-toolkit-api-workflow"?

Hello. After the Tao Toolkit API service was deployed sucessfully, I found that there are two pods and containers which are named tao-toolkit-api-app and tao-toolkit-api-workflow.

What are the functions between the two services ?

The picture is the pods deployed on k8s

The picture is the pod created on docker

Thank you for your help in advance.

Please share the steps what you have done. Thanks.

After I unloaded the driver, those pods related with nividia gpu operator were running successfully.

Could you please explain the functions of tao-toolkit-api-app pod and tao-toolkit-api-workflow pod ?

TAO Toolkit API is a Kubernetes service that enables building end-to-end AI models .
Tao-toolkit-api-workflow will create datasets, models, obtaining pretrained models from NGC, obtaining default specs, training, evaluating, optimizing and exporting models for deployment on edge.

1 Like

Thank you very much, @Morganh Is Tao-toolkit-api-app pod a Kubernetes service that enables building end-to-end AI models ?

The api app pod is the interface where the API action (non DNN framework actions) take place: eg retrieving specs, uploading dataset etc - all the endpoints are accessed via this app pod.
The worklow pod is where the DNN container will be invoked, so when you run train action on notebook, the workflow pod creates a container according to the model and sends the command to the created container.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.