Open AgostinoA opened 11 months ago
hey @yil532 ! hoping you can triage this to the right person on your team
@yil532 Guys, is there any news?
@yil532 Guys, is there any news? We need to solve this problem, for a corporate job, if you have other faster means of communication let us know, we are still paying for a service on Google Vortex that doesn't work properly
@leahecole Kindly can you solicit
Hi, any news about this issue?
@leahecole Excuse me I saw that @yil532, left the post, but can you give us support for this issue. I name for a company that pay for your service and for an issues, you can not get support, we urgently need to solve, thank you
@AgostinoA What model_id are you using? Did you verify that
model_path = "projects/{}/locations/us-central1/models/{}".format(
project_id, model_id
)
Sets model_path
to point at a model that exists and is accessible to the user/service account executing the script?
In which file did you encounter the issue?
Describe the issue
https://cloud.google.com/video-intelligence/docs/streaming/video-classification
I am conducting a series of tests on AutoML, using the code provided in the official documentation, and I am encountering an issue. After training the model with VortexAI, completing the process, and retrieving the Model ID, the Python APIs consistently return 'Requested model could not be loaded.' Honestly, the demos are poorly written, and the exceptions in the code are very implicit. Some demos don't work, and you have to make corrections in certain areas to get them to function properly.
The model was specifically trained for Video Intelligence in the VortexAI section and, in turn, did not function correctly. In the documentation, it's not clear whether the model necessarily requires an Endpoint. Both the code and the documentation have a comprehension issue.