DeepPavlov is on the first place of DialoGLUE benchmark!!
DialoGLUE benchmark was presented at the fall of 2020. Challenge includes benchmark models from Amazon Alexa, which achieve state-of-the-art performance on five of the challenge tasks. To outperform baselines we trained the same models with augmented data.
DialoGLUE Leaderboard - https://eval.ai/.../challenge-page/708/leaderboard/1944
Post about DialoGLUE - https://www.amazon.science/.../amazon-launches-online...
DialoGLUE paper - https://arxiv.org/abs/2009.13570
DialoGLUE GitHub - https://github.com/alexa/dialoglue