LinkedIn
Twitter
Hey,
I'm running VertexAI to generate some data and can run it locally, yet when I try to run it on GCP, it hits an error.
The error is quite generic:
"Error: [VertexAI.GoogleGenerativeAIError]: exception posting request
at generateJob (...)
as shown in picture. Any clues how to debug?
Bump
I went through the documentation some more and tinkering with my settings. I was looking up the integration of my project with VertexAI but it seems to integrate and yet not show any available resources, as shown bellow.
But the documentation does not give any clues on how to actually view resources in these rows.
The one thing mentioned was finding an appropriate region, and setting all resources to the same region, which I have.
Any clues on how to proceed? 🙂
I have had a similar experience with my Vertex AI Node JS client. It works perfectly fine locally. But once I deploy on Cloud Run GCP, it returns error "exception posting request to model". I initially thought it was a service account permissions issue because it happened on a local container I built and ran as well. However, I've given all of the necessary permissions including Service Account User, Vertex AI agent, Vertex AI Admin, Cloud Run user etc to the service account. Still doesn't solve the issue. However, it runs perfectly locally with my service account.
