Merge pull request #111 from iclementine/develop

fix a config key error
PaddlePaddle · May 18, 2021 · 87a0cec · 87a0cec
2 parents 4f288a6 + 3aa6aed
commit 87a0cec
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 3 deletions.
diff --git a/examples/transformer_tts/config.py b/examples/transformer_tts/config.py
@@ -39,7 +39,7 @@
         d_ffn=1024,  # encoder_d_ffn & decoder_d_ffn
         encoder_layers=4,  # number of transformer encoder layer
         decoder_layers=4,  # number of transformer decoder layer
-        d_prenet=256,  # decprenet's hidden size (d_mel=>d_prenet=>d_decoder)
+        d_prenet=256,  # decoder prenet's hidden size (n_mels=>d_prenet=>d_decoder)
         d_postnet=256,  # decoder postnet(cnn)'s internal channel
         postnet_layers=5,  # decoder postnet(cnn)'s layer
         postnet_kernel_size=5,  # decoder postnet(cnn)'s kernel size

diff --git a/examples/transformer_tts/preprocess.py b/examples/transformer_tts/preprocess.py
@@ -38,7 +38,7 @@ def create_dataset(config, source_path, target_path, verbose=False):
     processor = AudioProcessor(
         sample_rate=config.data.sample_rate,
         n_fft=config.data.n_fft,
-        n_mels=config.data.d_mel,
+        n_mels=config.data.n_mels,
         win_length=config.data.win_length,
         hop_length=config.data.hop_length,
         fmax=config.data.fmax,

diff --git a/parakeet/models/transformer_tts.py b/parakeet/models/transformer_tts.py
@@ -571,7 +571,7 @@ def from_pretrained(cls, frontend, config, checkpoint_path):
             frontend,
             d_encoder=config.model.d_encoder,
             d_decoder=config.model.d_decoder,
-            d_mel=config.data.d_mel,
+            d_mel=config.data.n_mels,
             n_heads=config.model.n_heads,
             d_ffn=config.model.d_ffn,
             encoder_layers=config.model.encoder_layers,