Hi all,
I followed the guide for installing Rasa-X running it within the docker container. I want to run a model that has spaCy with a specific model. When I run the Rasa-X I receive the following error:
rasa-production_1 | File “/opt/venv/lib/python3.8/site-packages/rasa/nlu/utils/spacy_utils.py”, line 62, in load_model
raise InvalidModelError
rasa-production_1 | rasa.nlu.model.InvalidModelError: Please confirm that en_core_web_trf is an available spaCy model. You need to download one upfront. For example: python -m spacy download en_core_web_md
I added the model to the pipeline and requirements.txt.
I guess I need to manually install it in the docker?
If so, how should I do it?
Should I run a command in the docker-compose.yml to pip install the model, should I login into the docker? What do you recommend?
and everything is inside /etc/rasa (by default). Digging deeper I see that Rasa recommends (quote) " If your model has a dependency that is not included in any of the tags (for example, a different spaCy language model), you can build a docker image that extends the rasa/rasa image."
From here
I’m currently thinking about building my own rasa image using the spaCy model I need and use that in the docker compose above. Is this the right way to proceed?
If so, I’m not sure how to build the docker image for Rasa using
In other words, whether to use the Makefile in the repo and modify the Dockerfile above or call the Docker command build directly. Any pointers on this (assuming it is the right way to proceed)?