Filmmaker's best friend - Error in deployment to endpoint

[misterwhisper]

Hi all, I'm a filmmaker/creator with a passion for AI-assisted VFX scripting. On VertexAI, I am trying to deploy successful trained models to a functioning endpoint so I can use them on my film and its future AR/VR campaigns.  However, even the act of deploying a trained google model to an endpoint has been surprisingly hard.   Any advice on this would be greatly appreciated and I'd hope to return the favor. 

Also, anyone here into Nvidia Omniverse or OpenUSD?!  The future is bright my friends. 🙂 

[granik3]

In the Google Cloud Console, navigate to Vertex AI and create a new model resource. This is where you'll associate your model with a name and version. Make sure your model is trained and saved in a format supported by Vertex AI, such as TensorFlow SavedModel or AutoML. Associate your trained model with the endpoint you created in the previous step. You can do this by specifying the model version you want to deploy. After deployment, you can send prediction requests to the endpoint using the Vertex AI SDK or REST API. This is how you'll use your model to make predictions for your VFX or other applications. 

My BK Experience