metadata
language:
- en
About
Waiter dataset while ordering menu. This dataset aims to be used for fine-tuning LLM model to become a waiter that will help customers in the ordering process. The final output will be a chatbot that can be deployed in a restaurant's website or app to help user in ordering process.
The data was generated from various source of conversations between waiter and customer (45 rows) and then upscaled with GPT-4 model. There is some cleaning and refining of the upscaled data and the quality has been checked. All of the 389 rows are unique.
Features
instruction: describes the task the model should perform. Each of the 389 instructions are unique.
input: optional context or input for the task.
output: the answer to the instruction.