Hey all,
wondering if you could shine some light on my issue. Currently the ChatGPT connection will only output two lines of text no matter what I seem to do, I’ve tried different input connections and it is always the same, only putting out 2 lines tops. The finish_reason is always length every time as well.
Can anyone shine some light on my my prompt outputs are not completing? Thank you!
@WilliamO This is quite a common thing; until the model understands your requirement, it will give small or incomplete outputs.
You should apply a filter check where the content is sufficiently long. If not then rerun the module.
This can be achieved using the filters and length conditions.
I’m sorry but that does not make any sense, my prompt is definitely an average length and so is the input. The same prompt and input works just fine outside of make and on Zapier, and even through Google sheets.
Apply a filter check where the content is sufficiently long?
The default max tokens is 16, I’d guess that is what is causing it to return early. If you haven’t, I’d check the “show advanced” toggle on the bottom and set a larger number in the Max Tokens field.
Yup that was it. Unreal, I probably read that a dozen times and I even understand it but didn’t make the connection that was the issue. Thank you HessCJ!
