Can the gpt4 model be used now? I setted my model as .gpt4, but it doesn't work...

[donfishman]

Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

Describe the solution you'd like
A clear and concise description of what you want to happen.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

[SunburstEnzo]

Hi @donfishman, what error are you getting when using .gpt4? Do either gpt4_0613 or gpt3_5Turbo work for you or are they failing as well?

[donfishman]

Hi@SunburstEnzo, thanks for your reply, the code when I using .gpt4 worked perfectly without error. But When I asked him who he was, he always replied that he was gpt3, but actually, I had set the model to .gpt4 or .gpt4_0613, and I tried to ask him if he knew who the winner of the 2022 World Cup was, and he told me that he only studied until 2021 and could not predict the future....

[donfishman]

Hi, @SunburstEnzo ,now I got an error when using .gpt4:
Error Domain=NSURLErrorDomain Code=-1017 "cannot parse response" UserInfo={_kCFStreamErrorCodeKey=-1, NSUnderlyingError=0x6000015dcd80 {Error Domain=kCFErrorDomainCFNetwork Code=-1017 "(null)" UserInfo={_kCFStreamErrorCodeKey=-1, _kCFStreamErrorDomainKey=4}}, _NSURLErrorFailingURLSessionTaskErrorKey=LocalDataTask .<3>, _NSURLErrorRelatedURLSessionTaskErrorKey=(
"LocalDataTask .<3>"
), NSLocalizedDescription=cannot parse response, NSErrorFailingURLStringKey=https://api.openai.com/v1/chat/completions, NSErrorFailingURLKey=https://api.openai.com/v1/chat/completions, _kCFStreamErrorDomainKey=4}
but when I use .gpt-3.5-turbo everything is perfect

[SunburstEnzo]

I believe if you're using Completions you have to use gpt-3.5-turbo-instruct I'm afraid — it's a discussion point in a few other issues on here

[kalafus]

completions supports:
"gpt-3.5-turbo-instruct"
"babbage-002"
"davinci-002"

typesafe model support proposed in this pull obviates the question.

[chrlyck]

Don't think I understand this, completions on gpt4 are supported by OpenAI:

Is it because this framework is using the legacy api?

https://platform.openai.com/docs/guides/text-generation