[Bug]: Add Model Providers:Azure-OpenAI error

[guoyuhao2330]

Is there an existing issue for the same bug?

• I have checked the existing issues.

Branch name

main

Commit ID

main

Other environment information

No response

Actual behavior

ValueError('Must provide one of the base_url or azure_endpoint arguments, or the AZURE_OPENAI_ENDPOINT environment variable')

Expected behavior

No response

Steps to reproduce

Add Azure-OpenAI API-Key

Additional information

No response

[rpkpavan]

we are also facing same issue..please help.

[rpkpavan]

Dear @KevinHuSh : Thanks for the commit. I have done gitpull and i face the following error. kindly guide

while trying to add the Azure Open API from the portal ValueError('Must provide one of the base_url or azure_endpoint arguments, or the AZURE_OPENAI_ENDPOINT environment variable')

on backend logs i get this error:
"LLM factory gpt-4-32k has not supported yet, switch to 'Tongyi-Qianwen/QWen' automatically, and please check the API_KEY in service_conf.yaml."

[VietNguyenHoang812]

I add a GPT-4o key to with both API key and base_url but got this message error on Web:
"Hint 102
Fail to access embedding model(azure-text-embedding-3-large) using this api key.Unsupported data type Fail to access embedding model(azure-text-embedding-3-small) using this api key.Unsupported data type Fail to access embedding model(azure-text-embedding-ada-002) using this api key.Unsupported data type"

[rendybjunior]

I still got the same issue. I suspect this is because the Azure OpenAI configuration only allows full url chat completion as parameter. Meanwhile Azure openAI has different url for each deployment model for each different tasks.

Raised new issue here #2840