Since my files are really big and training takes a lot of resources I was wondering whether there is the possibility of uploading a trained RASA model instead of training on in the cloud?
I trained a model locally and would like to use it in Okteto deployment. It runs well in my docker containers locally but when I want to deploy it in Okteto I get this error message:
rasa-server/usr/local/lib/python3.7/site-packages/rasa/shared/utils/io.py:99: UserWarning: No valid model found at models!
This is the command I use in docker-compose to run the model (and which works nicely locally) but as I said I get the above error message.
command: bash -c "rasa run --model finalmodel.tar.gz --enable-api --cors \"*\" --debug --port 5006"
Before that it says:
Loading model models/finalmodel.tar.gz…
Hope someone can help me with this!