This file is also where responses are defined. For example, use this to include intents defined in the nlu.yml file or any created actions. The official documentation has a page on Connecting to Messaging and Voice Channels that provides more information.Īlso refer to our tutorial on How to Use WebSockets with Socket.IO for an example of a Socket.IO chat application that can integrate with a Rasa assistant.ĭomain.yml specifies what components from the configurations to include in the Rasa assistant’s “world”. The default file includes placeholders for numerous platforms, including Facebook, Slack, and Socket.IO. Learn more on the Model Configuration page of the official documentation.Ĭredentials.yml stores credentials used by the Rasa assistant for interfacing with text and voice chat platforms. Without specification, Rasa uses a default approach. See the official documentation’s page on Testing Your Assistant for information on constructing effective Rasa test stories.Ĭonfig.yml specifies the configuration for training the Rasa assistant. Test_stories.yml defines test stories to verify that the Rasa assistant responds as expected. See the official documentation’s Stories page for more on the roles and details of stories. These models are used for training the assistant for conversation, and consist of user intention and/or information annotations, alongside sequences of assistant actions. Stories.yml models dialogues that the Rasa assistant is expected to engage in. See the official documentation’s Rules page for further context. These should define rule-like behavior, or actions to always take when certain intentions or information are provided by the user. Rules.yml defines a set of specific actions to take given specific conditions. See the official documentation’s page on NLU Training Data for more details on these models. This gives the assistant structures to use in identifying user intention and communicated information. Nlu.yml defines Natural Language Understanding (NLU) models for the Rasa assistant. This is where most of the assistant’s development likely takes place. Rasa’s documentation covers actions and gives context for using them.ĭata/ contains the core models for the Rasa assistant. The contents should resemble the following outline excluding files like _init_.py that are not likely needed in developing the example Rasa assistant.Īctions.py defines custom actions for the Rasa assistant, using Python code that can be activated upon certain conditions. The following section breaks down the default project contents to help understand the structure and navigate Rasa’s components. The new project’s directory contains a basic Rasa project structure.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |