How modules (mentioned in config.yaml file under pipeline) are getting connected to the training file? Where is the import call for these modules made in the GitHub files?
I want to understand how rasa works but couldn’t find the above call in the repo. Could you please mention the file name or line number? It would be a great help. Thanks
@NileshSuthar can you please elaborate more what types of modules you want to connect and import call module GitHub files too?
@nik202 I don’t want to connect anything new. I just want to understand how the modules mentioned in config.yaml are being called for training? Exa: I add a module unde pipeline in config.yaml. Let’s say it is whitespace tokenizer. Now I can see that this tokenizer is described in rasa/nlu/tokenizer repo on GitHub. But I couldn’t find any call made to the tokenizer repo in any other file of the source code. Basically can you show me how Rasa imports the modules that we mention in config file under pipeline? Thanks
@NileshSuthar Firstly, in the config.yml file we add the components (not modules) i.e Components for our model configuration i.e Model Configuration
@NileshSuthar Secondly, in this rasa Github repo: https://github.com/RasaHQ/rasa/tree/main/rasa/nlu/tokenizers this is just a code at the backend rasa is using for the different tokenizers.
Note: As you know rasa is open-source, that is why all codes are visible for customization, research, academic learning, and experimental purposes.
@NileshSuthar Third, if you want to run any code it should always have __init___.py please see the Github for the reference.
@NileshSuthar Finally, if you want to learn more about the Algorithm of rasa you can ref this video tutorial of Rasa Dashboard: Introducing The Algorithm Whiteboard - YouTube
@NileshSuthar You question about importing: please see this code : https://github.com/RasaHQ/rasa/tree/main/rasa/core/policies
You need to see every codes, to get to know how rasa is working and what code rasa is using to bind the rasa open source. Good Luck!