Hello rasa folks!
I’m trying to deploy my very-simple assistant in server using kubernetes.
I have Rasax connected to my git repo but I am having problems to upload the trained model and I cant train it eather from rasa x, it seems that the problem is related to the nlu training data, all the files where loaded from the git repo except the nlu.yml file.
Rasa Open Source version 2.0.3
Rasa x version 0.32.4
When cheking the appi version I found that production and worker are not running, could that be the root of the problem?
Thanks in advance
Ana
Hi @anarucu, I would suspect that the issue is due to the 2.0 training data format. Take a look at the compatibility matrix of rasa and rasa x versions, as 0.32 does not have support for 2.0. Upgrading to 0.33 should help!
thanks a lot for your reply @erohmensing , upgrade was a good enough solution concerning the problem of nlu training data.
The problem with production and worker was due to the virtual machine where the Docker compose for rasa x was installed. the CPU must support Sandy Bridge other wise tensor flow module will dump. thank you again
1 Like