Using openai@azure as backend

[julietKiloRomeo]

I would like to use azure as the backend. How do you accomplish that?

This is how to do it with pure requests:

import requests

endpoint_url = "https://xxxxx.openai.azure.com/openai/deployments/xxxxx/completions?api-version=2022-12-01"
OPENAI_API_KEY = "xxxxxxxxxx"

response = requests.post(endpoint_url, 
    json={"prompt": "Once upon a time", "max_tokens":100},
    headers={
        "Content-Type": "application/json",
        "api-key": OPENAI_API_KEY,
    },
)

print(response.json())

I tried to hack my way through, by starting with lambdaprompt:

import os
os.environ["OPENAI_API_KEY"] = "xxxxx"

from lambdaprompt import GPT3Prompt

example = GPT3Prompt(
    "Sally had {{ number }} of {{ thing }}. Sally sold ",
)
example.backend.endpoint_url = "https://xxxx.openai.azure.com/openai/deployments/xxxx/completions?api-version=2022-12-01"
def dummy(answer, *args, **kwargs):
    print(answer)
    
example.backend.parse_response = dummy
example(number=12, thing="apples")

but get an error printed:

{'statusCode': 401, 'message': 'Unauthorized. Access token is missing, invalid, audience is incorrect (https://cognitiveservices.azure.com/), or have expired.'}
Is there an official way to use azure that I just haven't found in the docs? :)

[julietKiloRomeo]

Ok. I was looking at adding it as a backend in lambdaprompt and fortunately saw that there already is an open pull request for that 👍
I will either wait for that or create my environment from that branch.