Finetuning a bankerbot

So here’s my problem.

I’m trying to make a BankBot chatbot. I’ve got a few hundred rows of telephone scripts to use for training data. I’ve been pretty thoughtful in creating my fine-tunes as far as prompts and completions go, but the basic template of {prompt:customer_statement, completion:agent_statement} has not been particularly fruitful. (I am using ending tags to separate everything.)

I see this in the tutorials:

{"prompt":"Summary: <summary of the interaction so far>\n\nSpecific information:<for example order details in natural language>\n\n###\n\nCustomer: <message1>\nAgent: <response1>\nCustomer: <message2>\nAgent:", "completion":" <response2>\n"}
{"prompt":"Summary: <summary of the interaction so far>\n\nSpecific information:<for example order details in natural language>\n\n###\n\nCustomer: <message1>\nAgent: <response1>\nCustomer: <message2>\nAgent: <response2>\nCustomer: <message3>\nAgent:", "completion":" <response3>\n"}

Can someone point me to some sample data so that I might see what this looks like in practice and maybe I can figure out how to get better results from it?

Any comments on this would be appreciated.