Hello,
I am trying to generate large output after fine tuning gpt-3.5-turbo-0125, but my output is being cut in half. The limit is 2048, so
How can i increase the output?
If not is there any work around with that.
I am a tier 2 user.
Any suggestion and ideas are welcome.
Thank You in Advance.
Hi!
The tokens per minute rate limit should not be the relevant factor here.
Have you checked the max_tokens parameter?
Note that with GPT 3.5 the absolute max is set to 4096 for both input and output.
Otherwise you would need to choose another model.
You can read up on the limits here:
https://platform.openai.com/docs/models/gpt-3-5-turbo
1 Like
Thank you for your help! I will look into it.