All - Activity - bradyneal - Rasa Community Forum

Restarting Rasa X erases uncommitted labeled data

I’m using the Docker Compose installation method. Have you found that you’re able to recover the lost changes (e.g. via git stash pop on any of the installation types)?

After I’ve labeled data and before starting Rasa X (shows that data has been labeled):

After I restart Rasa X (no longer has my labeled data):

Is it the case that if I restart Rasa X, it erases all my uncommitted labeled data (resets it to the most recent commit)?

Apple M1: illegal hardware instruction (using any rasa command)

I’ve installed TensorFlow and sklearn using the above links (important part seems to be using miniforge), and they both work. However, the Rasa installation doesn’t seem to work. Is it supposed to work on M1 now, given that I’ve already successfully installed TensorFlow and sklearn?

Job to build ML that helps people with decision making (initially, restaurant decisions)!

Hi! Oogway’s mission is to help people make better decisions faster. We’re focusing on helping people with the “what restaurant should I choose?” decision first, but we plan to utilize the rapid pace of AI advancement to expand to all decisions over time (starting with decisions that are most similar).

Which Rasa chatbot widget to use

You mean the official Rasa Chat Widget (in the poll), right? It already seems to be online, no?

poll
Which one are y’all using? The Botfront one is no longer maintained, right? How do these compare in terms of features, customizability, maintenance, etc.? For example, I’m guessing the official Rasa Chat Widget will be the best maintained, but isn’t as customizable or open-source.

Is it possible to use nlu part of previously trained model only and use a newly trained core?

Gotcha, sounds like there’s no way to get it working with rasa interactive. I haven’t been careful about hooking training and git together, but it sounds like the thing you suggest is maybe the best solution for now:

Hook git commits and training together (always have a data commit associated with a training commit).

Get all Rasa X labeled data to go to a single file

It definitely writes to a file where the intent is already defined. I have the same intent used in my different files, though. Usually, my entities change across files (e.g. each file is for a different entity).

I have many files that my NLU data is split across. Data that I label in Rasa X seems to be going to different arbitrary files. How can I make all the data I label in Rasa X go to a single file?

How does the CRF for entity extraction work? Best practices?

I think the main thing that brought doubt into my mind is that I saw a Rasa team member saying that the entity extractor doesn’t really take into account the current token (current word for anyone using the default config with WhitespaceTokenizer) like it takes into account the tokens around it.

By “any part of the input”, I mean the current message. Because of the transformer and neural network layers before the linear chain CRF, entity extraction should be able to take into account any part of the current message, right?

Perfect! For anyone else reading, this is a linear chain CRF, which means that if this were applied directly on top of the tokens, it would mean that only the given token and the two tokens that are adjacent to it would be used in the entity prediction. However, because in DIET there’s a transformer layer, it can utilize a broader context.

Plan for nested entities support

Hi! Is there a plan to support nested entities in Rasa? Or are they already supported?
Example for clarity:
Text: Pad Thai
Entities: