hml18

```python self.mask_embedding.weight.data.copy_(torch.FloatTensor([ [1, 0, 0], [0, 1, 0], [0, 0, 1], [0, 0, 0]])) ``` 这里的mask是不是应该是 ```python self.mask_embedding.weight.data.copy_(torch.FloatTensor([[0, 0, 0], [1, 0, 0], [0, 1, 0], [0, 0, 1]])) ```