Has anyone been doing anything with the OpenAI API...
# cfml-general
d
Has anyone been doing anything with the OpenAI API? How much time are your spending training the data? How much ongoing time are your spending making sure the results it's returning are reliable? Are you training it on very specific known data? Or are you trying to have it learn from the data in your system?
r
I've done a lot with the API. I am not spending anytime 'training' - I am using it as is
'zero-shot' training as it is apparently called. I write a prompt that sets up the question on each request, without using any 'fine tuning' as they call it