distilbert-base-cased

This model is a distilled version of the BERT base model . It is cased: it does make a difference between English and English . The training details on the pre-training, the uses, limitations and potential biases are the same as for DistilBERT-base-uncased. The code for the distillation process can be found here. It was introduced in this paper. The model achieves the following results:. When fine-tuned on downstream tasks, this model achieves . the following . results: . when fine-tuneing on downstream . tasks, . this model . achieves the . following results . when . the . test results::. Glue test results,