gpt 3.5 turbo

[dsakura]

Here in git it says that it uses gpt 4o, but when testing the tool, it is at 3.5 turbo. Do I need to configure something to use 4o? Thank you very much!

[adamlui]

Hey what test shows it is 3.5 turbo? Right now it randomly cycles thru APIs per query, like Free Chat which offers:

'deepseek-ai/deepseek-llm-67b-chat', 'gemma2-9b-it', 'THUDM/glm-4-9b-chat', 'gpt-4o-mini-2024-07-18', 'llama3-70b-8192', 'mixtral-8x7b-32768', 'nous-hermes-2-mixtral-8x7b-dpo', 'Qwen/Qwen2-57B-A14B-Instruct', '01-ai/Yi-1.5-34B-Chat-16K'

So for example if Free Chat is the chosen API for a query, then it picks a random model (but I will add a model selector eventually)

[dsakura]

I ask it to inform its model. Have some way to verifiy? Thanks

[adamlui]

Delete all APIs in the apis obj except Free Chat, then delete all availModels except gpt-4o and ask him