Rasa Intent Router
This is a Rasa Open Source NLU model for intent classification in an assistant/router setting.
Model Details
- Framework: Rasa Open Source 3.6.21
- Base model:
distilbert/distilbert-base-uncased - Architecture:
WhitespaceTokenizer+LanguageModelFeaturizer+DIETClassifier - Training data:
data/nlu.yml - Training examples: 631
- Intents: 9
Intents
daily_news, family_album, music_playback, outfit_advice, proverb_game, reminder, small_talk, teleconference, unsupported_command
Usage
This model is intended to be served with Rasa:
rasa run --enable-api --port ${SERVER_PORT} -m distil_bert_1000.tar.gz
Limitations
The model is trained for English utterances in a limited assistant domain. It may perform poorly on unsupported languages, unseen intents, or commands outside the training distribution.
Model tree for LaraSg/rasa-router
Base model
distilbert/distilbert-base-uncased