Stories

These docs are for version 1.x of Rasa Open Source.

User Guide

NLU

Core

Conversation Design

API Reference

Migrate from (beta)

Reference

Glossary

Versions

viewing: 1.10.5

Stories

Rasa stories are a form of training data used to train Rasa’s dialogue management models.

A story is a representation of a conversation between a user and an AI assistant, converted into a specific format where user inputs are expressed as corresponding intents (and entities where necessary) while the responses of an assistant are expressed as corresponding action names.

Format

Here’s an example of a dialogue in the Rasa story format:

## greet + location/price + cuisine + num people
* greet
   - action_ask_howcanhelp
* inform{"location": "rome", "price": "cheap"}
   - action_on_it
   - action_ask_cuisine
* inform{"cuisine": "spanish"}
   - action_ask_numpeople
* inform{"people": "six"}
   - action_ack_dosearch

What makes up a story?

A story starts with a name preceded by ##.
The end of a story is denoted by a newline.
Messages sent by the user start with *.
Actions executed by the bot start with -.

User Messages

While writing stories, you do not have to deal with the specific contents of the messages.

Actions

Utterance actions are hardcoded messages. Custom actions involve custom code being executed.

Events

Events have to be explicitly written as part of the stories.

Slot Events

Slot events are written as - slot{"slot_name": "value"}.

Form Events

Several kinds of events need consideration when dealing with forms in stories.

Checkpoints and OR statements

Checkpoints and OR statements should be used with caution.

Checkpoints

You can use > checkpoints to modularize your training data.

OR Statements

Another way to write shorter stories is to use an OR statement.

End-to-End Story Evaluation Format

The end-to-end story format combines both NLU and Core training data into a single file for evaluation.