Rasa Shell No NLU model found

[Deprecated] Rasa X Community Edition
1

Hi, Been updating when i run

rasa shell --debug

It always gives me

oot  - Starting Rasa Core server on http://localhost:5005
2019-05-25 11:13:23 INFO     root  - Enabling coroutine debugging. Loop id 5049651224.
2019-05-25 11:13:23 DEBUG    rasa.model  - Extracted model to '/var/folders/jl/4jvvp46s22gcddjv7_2tj5nm0000gn/T/tmpsr9oefjp'.
2019-05-25 11:13:23 WARNING  rasa.core.interpreter  - No NLU model found. Using RegexInterpreter instead.
2019-05-25 11:13:23 DEBUG    rasa.model  - Extracted model to '/var/folders/jl/4jvvp46s22gcddjv7_2tj5nm0000gn/T/tmpq66bdax7'.
2019-05-25 11:13:23.706700: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

Why can’t find my NLU model, it is also in the models folder

if i run

rasa shell nlu

it only gives me nlu predictions

how to enable NLU for the bot shell

2

Hi @souvikg10. Do you have a compressed model in your models directory or separate NLU and Core models? What happens when you run rasa train?

3

Hi @Juste, I ran them separately to train indicating the flag for nlu and core I guess the shell works when the train happens together It works when Rasa train

Which also means that now it is best to run Rasa nlu and core together