Hi @jing huang ,
Thank you for reaching out to the Microsoft Q&A forum. I understand that you want to make fine tuning in the model of GPT3.5.
Currently, Fine tuning is not available, as it is turned off for all regions due to capacity constraints while using an Azure OpenAI resource.
Unfortunately, we don't have any information at this moment when fine tuning opens again.
As more regions are added to Azure OpenAI, this restriction might be removed and fine tuning will be available.
The models page will be updated as soon as we open up regions for fine tuning and users should be able to submit the requests when it is made available.
I hope this information helps!