Helicone Community Page

Updated last year

All of my requests show as invalid prompt

I'm willing to share more information privately but all of my prompts are showing as invalid and not pretty-printing out. They are all valid JSON and I get valid request/response from OpenAI.

I'm using gpt3.5-turbo, if I make the same response using gpt4, it parses properly.

Attachments

12 comments

CCole

Hey , could you add me to your org? cole@helicone.ai

Thanks

aayoKho

What is the model here? And can you send a screenshot of the JSON output?

CCole

in the future, please use the OpenAI model names. Here are the gpt 3.5 ones:

https://platform.openai.com/docs/models/gpt-3-5
I suspect you are attempting to use gpt-3.5-turbo

SSteve

That's not possible, we have a custom deployment that has a custom model name, it looks like they're returning the underlying model name.

SSteve

But I need to send the custom model name in my request so they use the right deployment.

SSteve

I'd appreciate it if ya'll removed our custom model name from this chat

CCole

Apologies! Removed

SSteve

thanks!

CCole

I will look into this and get back to you!