joeddav/distilbert-base-uncased-go-emotions-student

This model is distilled from the zero-shot classification pipeline on the unlabeled GoEmotions dataset . It was trained with mixed precision for 10 epochs and otherwise used the default script arguments . The model can be used like any other model trained on GoEmotion, but will likely not perform as well as a model trained with full supervision . It is primarily intended as a demo of how an expensive NLI-based zero-shooting model could be distilled to a more efficient student . The teacher used single-label classification to create psuedo-labels . Note that although the GoEmotional dataset allow multiple labels per instance, the teacher used one label per instance to create Psued,