WASSA-2021-Shared-Task

Empathy Results:

Mode	Training Correlation	Validation Correlation
Plain RoBERTa model	0.56	0.47
Plain multi-input model	0.64	0.50
Multi-input, multi-task model	0.74	0.57
Multi-input, multi-task model+ NRC + Empath	0.64	0.54

Distress Results:

Mode	Training Correlation	Validation Correlation
Plain RoBERTa model	0.56	0.41
Plain multi-input model	0.64	0.49
Multi-input, multi-task model	0.66	0.53
Multi-input, multi-task model+ NRC + Empath	0.66	0.56