Need guidance or advice on train model

[privatejava]

Hi,
I have been trying to create custom model which will help me to search and summarize my library of rules using ada , davinci for my big library of rules and regulation. I have created prompt like

prompt: "For the Rule 2 Section A : \n\n"
completion:  " <content>#END#"

prompt: "For the Rule 2 Section A : \n\n"
completion:  " <content>#END#"

I did it for all of my rules and sections but when I tried to use that custom model with completion api . It just does not give me correctly. I think I am missing something or I am doing something wrong. Could any body please help me on this for me to perfect out the model ?

Thanks ,

[ruby_coder]

@privatejava

Could any body please help me on this for me to perfect out the model ?

Sure. Here ya go, @privatejava

Fine-Tuning In a Nutshell with a Single Line JSONL File and n_epochs

Those who frequent this community easily notice that developers struggle to get good fine-tunings results. In this multi-part tutorial, I’m going to use the davinci base model and fine-tune it multiple times with the same single-line prompt using different values for n_epochs. I will take a completion with each of these models, and demonstrate that if you set your n_epochs value high enough, you can get excellent results fine-tuning with a single-line prompt! Before we get started, I would …

HTH