Thomas
The first three items in `verbnet.lemmas()` are "December", "FedEx" and "UPS" (using NLTK 3.3). The following is a minimal example to reproduce the issue: ``` In [3]: import nltk In...
_goddamn you autocorrect_
**Definition of Done:** - [ ] Bot upgraded to Rasa 3.0 - [ ] Validate nothing broke (i.e. run with user simulator pre- and post-upgrade and compare results)
Currently, `carbon-bot` is one big NLU file, but doesn't have a dedicated evaluation and validation set. We have annotated 1200 additional NLU examples - data that has been collected from...
annotated more flu (yes autocorrect flu data!) data
There are a couple of insults and other offensive responses in `domain.yml` (see `utter_insult` and `utter_vulgar`) that we should remove - esp. now that we are more actively using `carbon-bot`...
The model regression tests have been failing on some datasets due to changes in `domain.yml` for the 3.0 release (see Slack thread [here](https://rasa-hq.slack.com/archives/C36SS4N8M/p1635439839103300)). I have run the recommended `rasa data...
Some things I have encountered during setup * I know this is an obvious step but it might be worthwhile mentioning that the repository needs to be cloned and the...