Hello, There is something I do not understand: I have Number of intent examples: 264 (9 distinct intents) When I split the data for training by 80% training and 20% testing (rasa data split nlu), it became: training (211) and testing (53) and this is correct I also have: Number of entity examples: 160 (14 distinct entities) When I split the data for training by 80% training and 20% testing, it became: training (128) and this is correct…test (171) and this is incorrect…what is the reason? intents.yml (17.9 KB) lookups.yml (11.1 KB) synonyms.yml (13.8 KB) config.yml (706 Bytes)
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Large number of entities per example & cross validation expectations | 0 | 773 | April 5, 2019 | |
| Is it a problem, if i have more nlu examples | 2 | 407 | October 5, 2021 | |
| Wrong intent identification | 3 | 306 | December 5, 2021 | |
| Choosing NLU pipeline | 6 | 1314 | December 16, 2019 | |
| Training on large intent examples | 5 | 1319 | March 17, 2022 |