Hello,
I have been trying to train an AutoML Image Classification model in the us-central1 region.
My dataset (drowsy_model_final) successfully imported 368,227 images. The data import seems to be 100% successful.
However, every time I try to start training, the pipeline fails immediately (within 10 minutes) with the error: Error Messages: INTERNAL
This has happened 3 times in a row. I have tried:
A large budget (125 node hours)
A minimum test budget (8 node hours)
Both failed with the same INTERNAL error.
My project ID is project-(PII Removed by Staff). The dataset seems to be perfectly fine. Could you please check if there is an internal issue (like a quota or resource allocation problem) with my project in the us-central1 region?
Here are the Failed Pipeline Resource Names:
projects/(PII Removed by Staff)/locations/us-central1/trainingPipelines/5170562229728182272 (8-hour test)
projects/(PII Removed by Staff)/locations/us-central1/trainingPipelines/8075383989382152192 (v2)
projects/(PII Removed by Staff)/locations/us-central1/trainingPipelines/6568929914026721280 (v1)
Thank you for your help.