Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

renamde text_encoder_output_dim in capacitron_layers for clarity #7

Merged
merged 1 commit into from
Sep 22, 2021
Merged

renamde text_encoder_output_dim in capacitron_layers for clarity #7

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions TTS/tts/layers/tacotron/capacitron_layers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ class CapacitronVAE(nn.Module):

See https://arxiv.org/abs/1906.03402 """

def __init__(self, num_mel, capacitron_embedding_dim, encoder_output_dim=256, reference_encoder_out_dim=128, speaker_embedding_dim=None, text_summary_embedding_dim=None):
def __init__(self, num_mel, capacitron_embedding_dim, text_encoder_output_dim=256, reference_encoder_out_dim=128, speaker_embedding_dim=None, text_summary_embedding_dim=None):
super().__init__()
# Init distributions
self.prior_distribution = MVN(torch.zeros(capacitron_embedding_dim), torch.eye(capacitron_embedding_dim))
Expand All @@ -21,7 +21,7 @@ def __init__(self, num_mel, capacitron_embedding_dim, encoder_output_dim=256, re
mlp_input_dimension = reference_encoder_out_dim

if text_summary_embedding_dim is not None:
self.text_summary_net = TextSummary(text_summary_embedding_dim, encoder_output_dim=encoder_output_dim)
self.text_summary_net = TextSummary(text_summary_embedding_dim, text_encoder_output_dim=text_encoder_output_dim)
mlp_input_dimension += text_summary_embedding_dim
if speaker_embedding_dim is not None:
# TODO: Figure out what to do with speaker_embedding_dim
Expand Down Expand Up @@ -157,9 +157,9 @@ def calculate_post_conv_height(height, kernel_size, stride, pad,
return height

class TextSummary(nn.Module):
def __init__(self, embedding_dim, encoder_output_dim):
def __init__(self, embedding_dim, text_encoder_output_dim):
super().__init__()
self.lstm = nn.LSTM(encoder_output_dim, # text embedding dimension from the text encoder
self.lstm = nn.LSTM(text_encoder_output_dim, # text embedding dimension from the text encoder
embedding_dim, # fixed length output summary the lstm creates from the input
batch_first=True,
bidirectional=False)
Expand Down