I did a chatbot using my pdf files with the gpt3.5 turbo model. Everything works fine. But in the API usage it shows request from text-embedding-ada-002-v2, which is ok because i use embeddings for the pdfs but it also shows requests from text-davinci.
I checked the correct initialisationof the model with giving me the values back with print and it showed the 3.5 turbo model is used.
Maybe i don’t understand something right. But if i use the gpt3.5 -turbo model is this then right that there are requests from text-davinci?