Capture Whisper usage

[jpanay-TNM]

Summary

Pretty much the title, lol. Would be nice to be able to see OpenAI usage for the whisper-1 model in New Relic. If this is already a feature then maybe it is just a bug on my end because it seems to be capturing calls for the gpt-3.5-turbo and gpt-4 models but not whisper-1. Sample code below.

Desired Behavior

Include the same functionality for whisper that we have for the gpt models.

Additional context

import openai
from nr_openai_observability import monitor

openai.api_key = "NUNYA_BEEZWAX"
monitor.initialization(application_name="feature-request-example", license_key="DONT_HACK_ME_PLS")

# Model: gpt-4 (Captured by New Relic)
response_chat_completion = openai.ChatCompletion.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "What is Observability?"}],
    max_tokens=100,
    temperature=0
)
print(response_chat_completion)

# Model: whisper-1 (Not captured by New Relic)
audio_file = open("./testing-clip.mp3", "rb")
response_transcribe = openai.Audio.transcribe("whisper-1", audio_file)
print(response_transcribe)