# This tests the litelm proxy 
# it makes async Completion requests with streaming
import openai

openai.base_url = "http://0.0.0.0:8000"
openai.api_key = "temp-key"
print(openai.base_url)

async def test_async_completion():
    response = await openai.Completion.acreate(
        model="gpt-3.5-turbo", 
        prompt='this is a test request, write a short poem',
    )
    print(response)

    print("test_streaming")
    response = await openai.Completion.acreate(
        model="gpt-3.5-turbo", 
        prompt='this is a test request, write a short poem',
        stream=True
    )
    print(response)
    async for chunk in response:
        print(chunk)


import asyncio
asyncio.run(test_async_completion())