Hi! Could you share the command you're running and an excerpt from the patterns you created using the insults terminology dataset? One possible explanation is that the insults patterns aren't matched, so all you see are the models suggestions, which are completely random (since the model hasn't really learned anything about insults yet).
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
rookie error following insults textcat video? | 2 | 409 | August 4, 2019 | |
German Insult Classifier | 3 | 912 | January 14, 2019 | |
Prodigy terms.teach does not give any meaningful suggestions with en_vectors_web_lg , en_core_web_lg | 5 | 735 | September 2, 2020 | |
Training Insults classifier video out of date (--seeds argument) and moved documentation | 4 | 669 | February 8, 2019 | |
Entity Recognition vs. Text Classification
|
2 | 3050 | May 11, 2018 |