model-for-DROP

train

allennlp train bert5.jsonnet -s models --include-package code 

Details

We employ a python library decimal for decimal fixed point and floating point arithmetic. The model is trained for 20 epochs with batch size 10 Implementation is done with allennlp[21], and all models are trained on two GTX 1080tis.