Hello, I am unsure about how tokens are counted for functions. I am assuming the whole json schema is used to count for the token or does it do something in the server and count it a different way?
Thank you,
Wai
Hello, I am unsure about how tokens are counted for functions. I am assuming the whole json schema is used to count for the token or does it do something in the server and count it a different way?
Thank you,
Wai
Token count for context window is always input + output. So any input prompt that you are giving plus the output that you get back from GPT will count towards the total tokens used
Hi,
You can take a look at this post