---
license: mit
---
Model trained on IBMArgRank30k for 2 epochs with a learning rate of 3e-5 (optimised via grid search)