To celebrate our major version launch, we held a free weeklong event on our YouTube channel. The event included office hours where people could ask their questions about new features but also a live coding session and a community showcase. This launch event is available as a playlist and can be viewed here. In particular, you can rewatch the following events:
- Rasa 3.0: Launch Livestream
- Rasa 3.0: Architecture Office Hours
- Rasa 3.0: Slots Office Hours
- Rasa 3.0: Markers Office Hours
- Rasa 3.0: Livecoding
- Rasa Community Showcase: 3.0 Edition
If you go to our YouTube channel then you may notice that some of our playlists have been updated. The Rasa Tutorials playlist has replaced some of the old form tutorials such that the new slot mappings feature is taken into account. We didn’t remove the videos for Rasa Open Source 2.x and have chosen to move them into an archived tutorial playlist. That way, if you’re still using an older version of Rasa, you can still benefit from the tutorials.
You can watch the new tutorials on forms on Youtube. In particular, here are the updated videos.
- Introduction to Slots
- Constraining Slot Values
- Introduction to Rasa Forms
- Slots filled by Text
- Form Validation
Similarly, we’ve also updated our Conversation AI with Rasa playlist. All the tutorials that discussed slots have now been updated to reflect the changes introduced by the new slot mappings feature. This includes the following lessons:
- Conversational AI with Rasa: Slots
- Conversational AI with Rasa: Custom Actions
- Conversational AI with Rasa: Basic Forms
- Conversational AI with Rasa: Custom Forms
Just like with the Rasa tutorials playlist, we’ve also created an archived playlist that is compatible with Rasa Open Source 2.x. This playlist can be found here.
Rasa Learning Center
We didn’t just update the videos on Youtube, we’ve also updated the lessons on our learning center. That means that all the code examples and appendices also received updates. In particular, the new “Rasa Forms”-course can be found here and the new “Conversational AI with Rasa Course” can be found here.
You can also find a new archive on the learning center that contains the material for Rasa 2.x. If you’re watching content that has been updated you’ll also see a warning that can take you to the updated material right away.
It’s also good to know that the Rasa certification study guide is up to date and compatible with both Rasa Open Source 3.0 and 2.x. If you’re ready to take the certification exam then you can take it here.
Rasa Documentation and Migration Guides
The Rasa docs have also been improved with sections that weren’t around before. There’s a migration guide that will help you upgrade to Rasa Open Source 3.x but there are also completely new sections that highlight the new features. Here’s a list of new resources.
- The docs now contain a section that highlights the new graph architecture. This document also comes with a tutorial on how to create your own custom NLU components. It even has some examples that you can use as a source of inspiration.
- The docs now list a new section on slot mappings. The new slot mapping changes are also reflected in the forms documentation page.
- The docs now also show a new section on our new markers feature.
The docs have also been simplified with regards to versions of Rasa. We now host the docs for 1.x, 2.x, 3.x, and the main unreleased branch on Github. This simplifies the docs immensely and should make it easier to find relevant sections of older Rasa versions.
When we announced Rasa Open Source 3.0 we immediately wrote a blogpost that explains what to look forward to. The features that we’ve mentioned in this blog post are now also explained in more detail on other posts on our blog. In particular:
- Slots in Rasa Open Source 3.0 talks about the changes that were introduced to slots.
- Markers in Rasa Open Source 3.0 contains a small tutorial on how to get started tracking the performance of your assistant as it is interacting with users.
- Bending the ML Pipeline in Rasa Open Source 3.0 explains how the new graph architecture will make re-training much faster.
There are many resources available to you if you want to learn more about the new features in Rasa Open Source 3.0. We’re keeping the material for Rasa Open Source 2.x around in an archive so you can always consult these if you need them. As always, more content on new features is on the way.