Prompt engineering can dramatically speed up building and testing classification models. You can use a model to generate datasets automatically, tune model parameters and even interpret classification results.
Task:
Generate a labeled dataset for training a text classification model on the categories "Spam" and "Not spam".
Prompt:
Generate 20 text examples for classification into the categories 'Spam' and 'Not spam'. Half of the examples must be spam. Use the following format: Text: Category: "
Result:
1. Text: "Buy a new smartphone at 50% off! Follow the link!"
Category: Spam
2. Text: "Don't forget to send the report by the end of the day."
Category: Not spam
3. Text: "Get free access to premium services, just click here!"
Category: Spam
4. Text: "Team, the meeting has been moved to 3:00 p.m."
Category: Not spam
When building datasets, try to include varied examples so the model trains better. Make sure the examples are realistic and cover a wide range of possible situations.