Because she used custom training instead of AutoML. When she trained it, she used custom containers. For the endpoint, she used a prebuilt tensorflow container.
vertex ai is supposed to be a virtually automated thing. here it is talking about having to write code for prepackaged containers. 0 out of 5 stars for this video. reported for spam and misleading.
Does vertexAI support multi model endpoints similar to AWS sagemaker.
Does Vertex AI allow to manage the batch prediction in an easy way?
Great! But unfortunately way too many steps...
When you made use of tensorflow framework to build the model, then why did you choose "custom" model in the options
Because she used custom training instead of AutoML. When she trained it, she used custom containers. For the endpoint, she used a prebuilt tensorflow container.
Nice explanation
Ah. Posted 3 days ago. I'm lucky vertex AI is getting deployed and documented as I'm getting into ml.
So "Custom Training" is only enabled for Tensorflow, scikit-learn and XG-Boost?
Not "that custom" after all...
Prebuilt containers only support those 3. You can use whatever you want if you create a custom container.
Awesome job
UI Inconsistent with reality - where is the Notebooks tab??? 👎
Wertex AI changed my life ..
Dear Sir please do not be racism. 🙏🏼
awesome
Good
this tutorial really blows... also, the container registry used in the video is now deprecated
esto ya se ha ido de madre
vertex ai is supposed to be a virtually automated thing. here it is talking about having to write code for prepackaged containers. 0 out of 5 stars for this video. reported for spam and misleading.
Good luck getting a GPU quota.
Spolier: you will not get it.