[Feature]: Azure OpenAI Support

[Mshz2]

Is your feature request related to a problem? Please describe. The server url fpr openai api is not applicable to azure openai Endpoint url.

Describe the solution you'd like OPENAI_TYPE=Azure PROXY_API_KEY=Azure key PROXY_SERVER_URL=https://resoucename.openai.azure.com

Describe alternatives you've considered I tried to do PROXY_API_KEY=**** PROXY_SERVER_URL=https://****.openai.azure.com/openai/deployments/chatgpt35/chat/completions?api-version=2023-03-15-preview But got bellow error in the frontend:

[csunny]

Azure is not supported currently， can you give a pr for that?

[Mshz2]

Azure is not supported currently， can you give a pr for that?

I would love to, but unfortunately I am not sure if I have sufficient skill. Anyway, I believe many would like to deploy your nice product in the cloud, which for instance cloud providers like azure is popular due to openai gpt model integration.

[cason0126]

i can handle it @csunny

[cason0126]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo

u can use config like this 。 it works @Mshz2

[yihong0618]

@cason0126 nice

[riverind]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo

u can use config like this 。 it works @Mshz2

PROXY_SERVER_URL is represented for？ thks @cason0126

[cason0126]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo u can use config like this 。 it works @Mshz2

PROXY_SERVER_URL is represented for？ thks @cason0126

I guess it should be redundant configuration items. You can leave this part blank, or just fill in XXX as the value. As for why and iterative improvement, I am also learning.

[riverind]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo u can use config like this 。 it works @Mshz2

PROXY_SERVER_URL is represented for？ thks @cason0126

I guess it should be redundant configuration items. You can leave this part blank, or just fill in XXX as the value. As for why and iterative improvement, I am also learning.

thx, you run success or not ?

if success, can i know your para about azure?

if set the proxy_server_url to space, it will occur error, error info is as follows: dbgpt_server.py: error: the following arguments are required: --proxy_server_url

[cason0126]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo u can use config like this 。 it works @Mshz2

PROXY_SERVER_URL is represented for？ thks @cason0126

I guess it should be redundant configuration items. You can leave this part blank, or just fill in XXX as the value. As for why and iterative improvement, I am also learning.

thx, you run success or not ?

if success, can i know your para about azure?

if set the proxy_server_url to space, it will occur error, error info is as follows: dbgpt_server.py: error: the following arguments are required: --proxy_server_url

i can work success . my para is :
LLM_MODEL=proxyllm PROXY_API_KEY=your key PROXY_API_BASE=https://{your_domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx
PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo

[Mshz2]

@csunny @cason0126 thanks for the feedback. It is working now ;)

By any chance, how can I change similarly for using the azure embedding model? I tried to do like this, but did not worked:

EMBEDDING_MODEL=proxy_openai
proxy_openai_proxy_server_url=https://myresourcename.openai.azure.com/openai/deployments/text-embedding-ada-002/embeddings)
proxy_openai_proxy_api_key=password
proxy_openai_proxy_backend=text-embedding-ada-002

[AbhijitManepatil]

LLM_MODEL=proxyllm PROXY_API_KEY={your key} PROXY_API_BASE=https://{your domain}.openai.azure.com/ PROXY_API_TYPE=azure PROXY_SERVER_URL=xxxx PROXY_API_VERSION=2023-05-15 PROXYLLM_BACKEND=gpt-35-turbo

Only change in above parameters : I used PROXYLLM_BACKEND="mydeplymentname" works for me

[Cualacin0]

EMBEDDING_MODEL=proxy_openai proxy_openai_proxy_server_url=https://myresourcename.openai.azure.com/openai/deployments/text-embedding-ada-002/embeddings) proxy_openai_proxy_api_key=password proxy_openai_proxy_backend=text-embedding-ada-002

@csunny @cason0126 thanks for the feedback. It is working now ;)

By any chance, how can I change similarly for using the azure embedding model? I tried to do like this, but did not worked:

EMBEDDING_MODEL=proxy_openai
proxy_openai_proxy_server_url=https://myresourcename.openai.azure.com/openai/deployments/text-embedding-ada-002/embeddings)
proxy_openai_proxy_api_key=password
proxy_openai_proxy_backend=text-embedding-ada-002

@Mshz2 Hi, have you figured out how to use azure embedding model?

[Mshz2]

EMBEDDING_MODEL=proxy_openai proxy_openai_proxy_server_url=https://myresourcename.openai.azure.com/openai/deployments/text-embedding-ada-002/embeddings) proxy_openai_proxy_api_key=password proxy_openai_proxy_backend=text-embedding-ada-002

@csunny @cason0126 thanks for the feedback. It is working now ;) By any chance, how can I change similarly for using the azure embedding model? I tried to do like this, but did not worked:

EMBEDDING_MODEL=proxy_openai
proxy_openai_proxy_server_url=https://myresourcename.openai.azure.com/openai/deployments/text-embedding-ada-002/embeddings)
proxy_openai_proxy_api_key=password
proxy_openai_proxy_backend=text-embedding-ada-002

@Mshz2 Hi, have you figured out how to use azure embedding model?

I'm not using the tool anymore))