Question about GPT4 context passing

API
,
1

In the GPT3.5 version, the API caller needs the caller to pass in context information, for example:

openai.ChatCompletion.create(
   model="gpt-3.5-turbo",
   messages=[
         {"role": "system", "content": "You are a helpful assistant."},
         {"role": "user", "content": "Who won the world series in 2020?"},
         {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
         {"role": "user", "content": "Where was it played?"}
     ]
)

But now that 4.0 is open, it is said that the caller does not pass in the context information. The GPT server will automatically process the context information. I want to know how to implement it through API calls.

2

You always need to pass any context you wish to be considered in the response, there is no way around that. Where was it said that you no longer have to?

1 Like
3

I just saw similar words in some unofficial media reports, so I am not sure, so I will confirm it officially.

1 Like