Rasa X model training

Hi, I am new to Rasa X. I deployed Rasa X on Kubernetes with docker container on AWS EKS environment. I have connected Rasa X to my GitHub repo where I have the project files for my Rasa assistant. When I hit “Train model” button under “Add model”, the training fails. Why is that? Please help!!! Thank you!

What’s your Rasa and Rasa X version? Is it incremental training or fresh training?

It is incremental training. I already have a model that was pretrained with a small set of data on my local machine. Now I want to expose that model to the test users through Rasa X. Rasa version 2.5.0 and Rasa X version 0.39.3

The versions looks good. Have you already uploaded the model to Rasa X?

Well, I connected Rasa X to my GitHub repo where I have my project files including the pretrained models (the model files with .tar.gz extension). My understanding is that once this connection is established, Rasa X should be able to fetch the model from the GitHub repo and train the model. No?

Ah. Can you see your model in Rasa X?

Rasa X only syncs training data. It does not sync models. You will need to either upload a model to Rasa X via the UI or API. We recommend storing your model in a separate store.

1 Like

Oh I see. So Rasa X doesn’t synch models? Then it makes sense. Thank you!