New Interactive Dataset: QA 🔧

Brand-new Interactive Dataset: QA :wrench:

This feature allows you to train your agent to answer certain queries predictably, no matter what pre-existing knowledge it has.

With broad or exact match options, you can control the flexibility of the question-answering model.

A broad match enables the agent to reply similarly to questions with related grammatical patterns, thus providing consistent and reliable responses.

Fine tuning

But here’s the fascinating part - this fresh update allows you to fine-tune responses for enhanced accuracy. Not only can you create a QA pair with your visitor’s question and a more refined agent’s answer, but in your conversation history, you can simply click on “fine-tune” to do so.

This update is all about giving you more control and precision in training your agent. We hope you are as excited as we are! :blush:

1 Like