yueyu1030 / COSINE

[NAACL 2021] This is the code for our paper `Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach'.
200Updated 2 years ago

Related projects: