-
Notifications
You must be signed in to change notification settings - Fork 4.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add full intent name to ranked responses #5690
Conversation
…esponse_intents * 'response_intents' of github.com:hotzenklotz/rasa: (37 commits) fix style issue in test_evaluation fix deepsource sequence lengths error fix golden inputs in case mask loss is on in diet Update docs/user-guide/building-assistants.rst fix typo in type def, plus remove redundant comment fix typo in forms doc add changelog set regularization constant to 0.001 prepared release of version 1.9.6 rename changelog file use correct config param apply activation if transformer used ci cd docs based on 1.9 add changelog move activation to transformer block fix formatting issues Update stories_trip_circuit_breaker.md Update end_to_end_trips_circuit_breaker.md Update tests/core/test_evaluation.py Update tests/core/test_evaluation.py ...
Thanks for submitting a pull request 🚀 @melindaloubser1 will take a look at it as soon as possible ✨ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code looks good 👍 can you please add a test to tests/nlu/selectors/test_selectors.py
?
Co-Authored-By: Tom Bocklisch <tom@rasa.com>
…esponse_intents * 'response_intents' of github.com:hotzenklotz/rasa: Update changelog/5690.improvement.rst fix defaults in docs fix number reference for changelog for preveous PR only apply input dropout to text add dense input dropout as a parameter Update changelog/5636.bugfix.rst add pass training to input dropout update changelog add dropout before ffnn
@tmbo I extended a test to cover the |
Proposed changes:
When using a
ResponseSelector
a user action has access to thefull_retrieval_intent
property for the predicted response. Additionally, theResponseSelector
provides aranking
of all predicted responses with only the properties ´nameand
confidence`.For downstream tasks it would be useful to have the
full_retrieval_intent
property also available for theranked
responses. This PR adds this feature.Status (please check what you already did):
black
(please check Readme for instructions)