Hi guys,
please, I am trying to train the model in weatherbot tutorial and I am getting this error on CMD:
Traceback (most recent call last):
File “nlu_model.py”, line 1, in
from rasa_nlu.converters import load_data
ModuleNotFoundError: No module named ‘rasa_nlu.converters’
I am using Python 3.6 and installed the updated dependencies on requirements.txt. Anything else I should inform here?
Hi @EPedrotti , I appreciate your reply and will read carefully the link you sent. Thanks a lot man!
EDIT:
I did check the github link, Justina says that training the nlu_model did not change much from the original video. There is a a new updated code to use, slightly different from the fist one, but my question is, which part of this new code should I use to train my model? It is not clear. Do you know something about that?
Hey @Vinicivs. I understand the frustration with the video being compatible with the older versions of Core and NLU. I would suggest you taking the code from the up-to-date directory of the tutorial repo and watch the video as a reference to understand the concepts. In this case, you don’t really need to add/remove any code - everything should work the way it’s written. The readme of the directory has the instructions on how to use the updated code as well. If you have any issues, let me know
Thanks for replying Justina! Look, it’s completely ok. The point is, for complete beginners like me, I never know where is exactly the mistake: on my code, or my pc, on the provided code (unlikely) etc, etc…it can be so many things, even running a command on the wrong folder!
But let me say something else. videos like the one you did are THE BEST kind of resource for people like me. I don’t know how many, but we do need tutorials which says clearly how, where and when run each single line of code, or create folders and everything else. 99% of so called “tutorials” on the net and medium don’t do it at all.
So, I really want to thank you for your hard work here. And please, if you find some time, after most important duties on Rasa, or even for someone else on Rasa team or community, I believe it worth to update that video .
Thanks a lot for your feedback @Vinicivs, I really appreciate it! Updating the video is on my radar, hopefully I will scope out some time for that during the holiday season.