参数项

参数值

Torch.Size

[x, 1, 128]

Self Attention Hidden layer dimension

768

Dropout

0.5

Learning Rate

0.001

Epoch

10

Batch Size

8