Average Model Cost: $0.0000
Number of Runs: 31,215
Models by this creator
The russian-sensitive-topics model is trained on a dataset of sensitive topics in the Russian language. It predicts combinations of 18 sensitive topics and is trained on a mixture of manually and semi-automatically labeled samples. The model's performance has been tested on the manually labeled data. The dataset is available on the model's GitHub or Kaggle page, and it is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
The Russian-inappropriate-messages model is a text classification model trained on a dataset of inappropriate messages in the Russian language. Inappropriate messages are those that may not contain obscenities or toxicity but can still damage the reputation of the speaker. The model serves as an additional layer of filtering for inappropriateness after toxicity and obscenity filtration. It can be used to detect sensitive topics in text. The model has been trained, validated, and tested on samples with 100% confidence and has achieved good metrics on the test set. The model and dataset are open-source and can be found on GitHub or Kaggle.