Open DYF-AI opened 4 years ago
In attention_recognition_head.py def smape: y_prev = predicted? I can't find predicted,and why state = torch.zeros(1, batch_size, self.sDim)?
In attention_recognition_head.py def smape: y_prev = predicted? I can't find predicted,and why state = torch.zeros(1, batch_size, self.sDim)?