Open michaellerickson opened 3 weeks ago
I cannot answer some of your questions unfortunately, but as you mentioned token limit issue with Open AI and have a different list of TPMs I felt I could at least offer some help in that area.
Open AI has the concept of usage tiers and based on the fact that your response said 30000 limit for gpt-4-turbo-preview I believe you are effectively on Open AI's tier 1 usage. Here is the link to the Open AI usage tier 1 guide: Tier 1 Rate Limits. Including a small screenshot for quick reference:
To workaround this I can offer a couple of suggestions:
I understand that 1-2 might not be an option and 3-5 may not give you as high a quality results as you are getting from gpt4-turbo. Though personally I have been very happy with the quality I have been getting from Google Gemini. I did want to at least give you a couple of options though in case any of these could work for you.
For reference: Gemini Flash is 1,000,000TPM and Gemini Pro is only marginally higher than gpt4-turbo at 32,000TPM. But both of these are on their free tier. There are pay as you go options for higher amounts.
They are forcing a present day solopreneur to a Team plan which at most costs $720 and is meant for two people versus $600 billed annually. You ain't got that, do you?
Ape of the sea over here. I'm after reason. #Q*Gate
I think Fabric is PIVOTAL in shaping how the world needs to see.
I'm still trying to get this to work with OpenAI. There's a token limit issue at the time.
Can django make fabric wed computer vision and text?
Why do we see double?
OpenAI Token limits