Using gpt-engineer with Azure OpenAI service endpoint

[HermanCodes]

I have an Azure OpenAI endpoint with a GPT-35-Turbo model deployed. Is it possible for me to use this endpoint and key with gpt-engineer?

[smitzlroy]

Any insight on this question would be great.

[machmuel]

I have an Azure OpenAI endpoint with a GPT-35-Turbo model deployed. Is it possible for me to use this endpoint and key with gpt-engineer?

I´m also using my own Azure OpenAI Service (https://..... .openai.azure.com/).
My goal is to use it for training of own data in my own separated area. The used openai library seems actually to also support usage of Azure OpenAI Service.

So, it would be very beneficial (for me and others) to also add support of Azure OpenAI Service as openai.api_type.

I already checked the code to find the code location, where it´d needed to be extended.
If someone could give me a hint, I´d provide even a prepared pull request for it.

Many thanks in advance!

@smitzlroy

[magnusahlden]

I have an Azure OpenAI endpoint with a GPT-35-Turbo model deployed. Is it possible for me to use this endpoint and key with gpt-engineer?

yes, just use --azure your-azure-url yourdirectory-with-prompt NAME_OF_YOUR_DEPLOYED_MODEL

for me it works with gpt-engineer --azure mga.openai.azure.com ./example/ magnus_model

where magnus_model is the deployed GPT3.5-turbo model