Power system state forecasting has gained more attention in real-time operations recently. Unique challenges to energy systems are emerging with the massive deployment of renewable energy resources. As a result, power system state forecasting are becoming more crucial for monitoring, operating and securing modern power systems. This paper proposes an end-to-end deep learning framework to accurately predict multi-step power system state estimations in real-time. In our model, we employ a sequence-to-sequence framework to allow for multi-step forecasting. Bidirectional gated recurrent units (BiGRUs) are incorporated into the model to achieve high prediction accuracy. The dominant performance of our model is validated using real dataset. Experimental results show the superiority of our model in predictive power compared to existing alternatives.