search

andabi / voice-vector

Deep neural networks for getting text-independent speaker embedding written in TensorFlow
MIT License
307 stars 83 forks source link

I think take the last timestep output should use code like 'out = out[:, -1, :]' #9

Closed shuguang101 closed 5 years ago

shuguang101 commented 5 years ago

Hi, @andabi in master branch model.py line 54 your code is: out = out[..., -1] # (n, h)

is this want to take the last timestep output?

andabi commented 5 years ago

@shuguang101 actually it was fixed but not updated ;( It should be out = out[..., -1, :].