Could the OpenAI team consider keeping the text-davinci-003 model in the completion API? I understand that all of the things that model can do can be done in chat mode, and that 97% of API usage is from the chat mode, but there are still users who prefer the text-davinci-003 model for certain use case(s). Perhaps you could keep it as a legacy model while still getting rid of all of the others like you plan to.
Hello, you can still access the text-davinci-003 model through their API. If you want to continue using it here is a basic Python script that I created to help you access DaVinci. It would still be nice to have DaVinci as an option while using Openai’s chat system.
//PYTHON CODE
from cmd import PROMPT
import os
from pyexpat.errors import messages
import openai
openai.api_key = "YOUR_API_KEY"
print("Ask DaVinci a question.")
question = input()
text = openai.Completion.create(
model="text-davinci-003",
prompt=question,
max_tokens=15,
temperature=1
)
print(text['choices'][0]['text'])