Please visit https://competitions.codalab.org/competitions/20011#learn_the_details-overview for further information
The results were presented in the paper SemEval-2019 Task 6: Identifying and Categorizing Offensive Language in Social Media (OffensEval) which can be found at https://arxiv.org/abs/1903.08983. Place: 10 out of 800 registered teams. See the team 'Vadym' for details.
- Please download FastText Embeddings: https://s3-us-west-1.amazonaws.com/fasttext-vectors/crawl-300d-2M.vec.zip and extract it to the project folder.
- Download the data and extract it to the project folder.
- Install all required packages:
pip install -r requirements.txt
. It is recommended to use a virtual environment. - The project folder should contain:
- 'training-v1' directory
- 'model_output' directory
- Tensorboard can be run by executing the command
tensorboard --logdir ./model_output/
- Run
model.py
See also Paper.pdf
for more details