Error “Invalid Cloud Storage URI for artifact_uri” when deploying gte-qwen2-7b-instruct model on Ver

[LudovicoP]

Hi everyone,

I’m trying to deploy the gte-qwen2-7b-instruct from huggingface found in model garden to use it for embeddings, but during the deployment I get this instant error:

Field: model.artifact_uri; Message: Invalid Cloud Storage URI for artifact_uri.

How could I solve this problem?

Thanks!!

[MarvinLlamas]

Hi @LudovicoP,

Welcome to the Google Cloud Community!

It looks like you are trying to deploy a model in Vertex AI, but the process is failing because it cannot copy the model's files into a storage bucket in your project. 

Here are the potential ways that might help with your use case:

• Use the Vertex AI Model Garden "DEPLOY" Button: Navigate to the gte-qwen2-7b-instruct model page within the Google Cloud Console's Model Garden, then select the "Deploy" button. This action will automatically copy your model to a valid Google Cloud Storage (GCS) location, ensuring that you avoid deployment errors related to an invalid  artifact_uri.
• Permissions: Ensure that the Vertex AI Service Account has the Storage Object Admin role, or confirm it has its default Vertex AI Service Agent role, to allow it to write the model to Google Cloud Storage.
• Validating Your Google Cloud Storage URI: Verify that your artifact_uri is correctly formatted as a valid Google Cloud Storage (GCS) path, such as ‘gs://your-bucket-name/path-to-artifact’.

Was this helpful? If so, please accept this answer as “Solution”. If you need additional assistance, reply here within 2 business days and I’ll be happy to help.