I followed each step of Deploy to a Server and everything seems to be ok. Then, I used Rasa X HTTP API Documentation to add my nlu data, stories, domain, and config. But I couldn’t train a model because my config file contains a custom component named datetime_parser
.
In my local, I placed datetime_parser.py
next to config.yml
but there is no config.yml
file inside any of my containers. I tried to send datetime_parser.py
file to /app
folder of rasa-x
, rasa-production
and rasa-worker
containers but nothing changed and I still have this error. What should I do now?
RASA_X_VERSION= 0.22.2
RASA_VERSION=1.4.3
Error:
ERROR:rasax.community.api.blueprints.models:500, Internal Server Error, body='b'{"version":"1.4.3","status":"failure","message":"An unexpected error occurred during training. Error: Failed to find module \'datetime_parser\'. \\nNo module named \'regex\'","reason":"TrainingError","details":{},"help":null,"code":500}''
Config file:
language: fa
pipeline:
- name: WhitespaceTokenizer
- name: CRFEntityExtractor
- name: EntitySynonymMapper
- name: CountVectorsFeaturizer
- name: EmbeddingIntentClassifier
- name: datetime_parser.ParserComponent
policies:
- name: KerasPolicy
- name: MemoizationPolicy
- name: FallbackPolicy
nlu_threshold: 0.3
core_threshold: 0.3
fallback_action_name: utter_not_clear_query
- name: FormPolicy
- name: MappingPolicy