I am trying to use Rasa in a new personal project. I want to use the CORE HTTP API but the /conversations/{sender}/parse route is not available. It returns a 404.
Has some someone succeeded to make it work? If yes can I have some help?
I can see that you save your NLU model inside the directory called ‘projects’, which means that after you run the first command your model is saved in ./projects/default/model_xxx . However, when you start the sever (running rasa_core.run function), the path you provide for your nlu model is models/current/nlu. I would expect you to get the 'No such file or directory: 'models/current/nlu\\metadata.json'' error here. To fix that you should specificy the correct path for your nlu model which is projects/default/model_xxxxxxxx.
Alternatively, you can change the directory where you save the nlu model after it’s trained. For example, using the following command:
Hello @Juste,
Thank you very much for your reply.
I indeed, run again the nlu_train command with what you suggested before being able to launch the server, you’re right.
Once I did that I was still facing the issue mentioned that the parse route was not available…
Thank you very much for your response.
I’ve seen this changelog was published today (for version 0.11.2) but documentation latest version is 0.11.1. Maybe it will be available later today. I’ll wait a bit and see.