Open the Tool

Use the default Fruit Finder example, then paste the chatbot's CSV rows into Add training examples from CSV.

Text Data Teachable Machine

Chatbot instruction

Copy this instruction and paste it into any generative AI chatbot.

I am making training data for a simple classroom machine learning activity. The clue columns are: Color Shape Size The Label/Class should be the fruit name. Here are the training rows I already have: Red, Round, Medium, Apple Green, Round, Medium, Apple Yellow, Long, Medium, Banana Yellow, Curved, Medium, Banana Generate a varied CSV dataset with many different fruits. Rules: - Use exactly 4 comma-separated values per row: Color, Shape, Size, Label/Class. - Use a variety of fruit labels, not only Apple and Banana. - Its okay to repeat the same fruit label in multiple rows, but vary the other values. - Do not include a header row. - Do not include explanations. - Use Title Case for every value in every CSV row. For example, write "Red, Round, Medium, Apple" instead of "red, round, medium, apple". - Wrap the answer in a code block. - Output only copyable plain text or code, with one CSV row per line. - Make the rows useful for teaching the model to recognize different fruits from Color, Shape, and Size. - Avoid exact duplicates of the rows I already have. Copy Chatbot Instruction