The repository consists of the code for creation of the dataset and training the small language models via model distillation.
-