Why `small = -1000` ?
gaoisbest opened this issue · 1 comments
gaoisbest commented
In file model.py
line 288, small = -1000
is used to construct
b_s = np.array([[small] * n_tags + [0, small]]).astype(np.float32)
e_s = np.array([[small] * n_tags + [small, 0]]).astype(np.float32)
I have two questions:
[1] Why does small
equal to -1000
?
[2] Why b_s
and e_s
are construct like this ? What's the difference between [[small, small, ..., 0, small]]
and [[small, small, ..., small, 0]]
Thanks.
gaoisbest commented
Have understand the principles.