OpenAI Customized Model Deployment Failure

Question

OpenAI Customized Model Deployment Failure

Borhade, Pooja 30

Hi, We are using Azure OpenAI service. We tried creating our own customized model from Azure OpenAI Studio. We can successfully create a customized fine-tuned model from the base model (Curie) by providing our own training data. However, it fails during deployment. We need help with the above issue.

0 comments

Answer accepted by question author

0 additional answers

Your answer

Answer 1

YutongTie-9091 54,026 Moderator

Hello @Borhade, Pooja

Thank you for your patience. Are you working on EastUS region? I have checked with product team regarding the issue you experienced. It appears that the root cause of the problem was a temporary shortage of available GPUs in the EastUS region. However, the team is actively working to free up resources, when there is one freed up their GPU, new deployments possible again.

We understand that this situation has been inconvenient for you, and we sincerely apologize for any frustration it may have caused. To minimize the impact of this issue, the product team has decided to open up additional deployment options in both the West Europe and South Central US regions. We hope that this will provide you with greater flexibility and enable you to continue using our services without further interruption.

Thank you for your understanding, and please do not hesitate to reach out to us if you have any further questions or concerns.

Regards, Yutong

-Please kindly accept the answer if you feel helpful and vote 'Yes' to support the community, thanks a lot.

Borhade, Pooja 30 Reputation points

2023-04-21T18:03:45.6266667+00:00

Thanks Yutong. We are using the EastUS region currently. We will check for other deployment options as well. Hopefully the issue gets resolved soon.
YutongTie-9091 54,026 Reputation points Moderator

2023-04-21T18:05:40.5633333+00:00

Hello @Borhade, Pooja Yes, I can understand your inconvenience, I hope we can get more GPU soon. Thanks for your understanding.
Ranjan goal 0 Reputation points

2023-05-07T22:35:33.9066667+00:00

I have the same problem even on US South Central region. I am using an ada model.

Was able to successfully fine-tune the model. During deployment, it errors out saying another model with the same name exists.

Definitely, this is the only deployment I am attempting on the South Central region.

I have done multiple retries with the same error. Also, my train and validation files are quite small - 20KB range. The 1MB file-based training was taking way too long.
Trieu Hong Truong 40 Reputation points

2023-05-09T09:32:37.0133333+00:00

Hi Yutong,

May I know if this improvement has been applied to the South Central region?

I have a customized model based on the curie model in this region, but it still can't deploy with no errors why it failed.

Share via

OpenAI Customized Model Deployment Failure

0 additional answers

Your answer