mirrors/icefall
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Simplify the dropout mask, no non-dropped-out sequences

Browse Source
This commit is contained in:
Daniel Povey 2022-10-16 19:14:24 +08:00
parent ae0067c384
commit 325f5539f9
1 changed files with 3 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
egs/librispeech/ASR/pruned_transducer_stateless7/conformer.py
View File

@ -163,15 +163,10 @@ class Conformer(EncoderInterface):
# self.encoder_unmasked_dim.
feature_mask_dropout_prob = 0.15
# we only apply the random frame masking on 90% of sequences; we leave the remaining 10%
# un-masked so that the model has seen un-masked data.
sequence_mask_dropout_prob = 0.9
# frame_mask is 0 with probability `feature_mask_dropout_prob`
# frame_mask1 shape: (num_frames1, batch_size, 1)
frame_mask1 = torch.logical_or(
torch.rand(num_frames1, batch_size, 1, device=x.device) > feature_mask_dropout_prob,
torch.rand(1, batch_size, 1, device=x.device) > sequence_mask_dropout_prob).to(x.dtype)
frame_mask1 = (torch.rand(num_frames1, batch_size, 1, device=x.device) >
feature_mask_dropout_prob).to(x.dtype)
feature_mask1 = torch.ones(num_frames1, batch_size, self.d_model[1],
dtype=x.dtype, device=x.device)