fix missing imports, fix ljspeech.yaml config key: encoder_channels

examples/deepvoice3/ljspeech.yaml

@@ -21,6 +21,7 @@ transform:
   # db scale
   min_level_db: -100
   ref_level_db: 20
+  clip_norm: true
 
 
 loss:
@@ -48,20 +49,20 @@ model:
   embedding_weight_std: 0.1
   freeze_embedding: false
   padding_idx: 0
-  encoder_channels: 256
+  encoder_channels: 512
 
   # decoder
   query_position_rate: 1.0
   key_position_rate: 1.29
   trainable_positional_encodings: false
   kernel_size: 3
-  decoder_channels: 512
+  decoder_channels: 256
   downsample_factor: 4
   outputs_per_step: 1
   
   # attention
-  key_position_rate: true
-  value_position_rate: true
+  key_projection: true
+  value_projection: true
   force_monotonic_attention: true
   window_backward: -1
   window_ahead: 3
@@ -88,16 +89,3 @@ train:
   snap_interval: 1000
   eval_interval: 10000
   save_interval: 10000
-
-
-
-
-
-
-
-
-
-
-
-
-

examples/deepvoice3/synthesis.py

@@ -1,6 +1,6 @@
 import os
 import argparse
-import ruamel.yamls
+import ruamel.yaml
 import numpy as np
 import soundfile as sf
 
@@ -22,6 +22,11 @@ if __name__ == "__main__":
     parser.add_argument("checkpoint", type=str, help="checkpoint to load.")
     parser.add_argument("text", type=str, help="text file to synthesize")
     parser.add_argument("output_path", type=str, help="path to save results")
+    parser.add_argument("-g",
+                        "--device",
+                        type=int,
+                        default=-1,
+                        help="device to use")
 
     args = parser.parse_args()
     with open(args.config, 'rt') as f:
@@ -67,7 +72,7 @@ if __name__ == "__main__":
         use_memory_mask = model_config["use_memory_mask"]
         query_position_rate = model_config["query_position_rate"]
         key_position_rate = model_config["key_position_rate"]
-        window_behind = model_config["window_behind"]
+        window_backward = model_config["window_backward"]
         window_ahead = model_config["window_ahead"]
         key_projection = model_config["key_projection"]
         value_projection = model_config["value_projection"]
@@ -76,11 +81,12 @@ if __name__ == "__main__":
                          freeze_embedding, filter_size, encoder_channels,
                          n_mels, decoder_channels, r,
                          trainable_positional_encodings, use_memory_mask,
-                         query_position_rate, key_position_rate, window_behind,
-                         window_ahead, key_projection, value_projection,
-                         downsample_factor, linear_dim, use_decoder_states,
-                         converter_channels, dropout)
+                         query_position_rate, key_position_rate,
+                         window_backward, window_ahead, key_projection,
+                         value_projection, downsample_factor, linear_dim,
+                         use_decoder_states, converter_channels, dropout)
 
+        summary(dv3)
         state, _ = dg.load_dygraph(args.checkpoint)
         dv3.set_dict(state)
 

examples/deepvoice3/train.py

@@ -1,6 +1,6 @@
 import os
 import argparse
-import ruamel.yamls
+import ruamel.yaml
 import numpy as np
 from matplotlib import cm
 import matplotlib.pyplot as plt
@@ -15,10 +15,9 @@ import paddle.fluid.layers as F
 import paddle.fluid.dygraph as dg
 
 from parakeet.g2p import en
-from parakeet.models.deepvoice3.encoder import ConvSpec
 from parakeet.data import FilterDataset, TransformDataset, FilterDataset
 from parakeet.data import DataCargo, PartialyRandomizedSimilarTimeLengthSampler, SequentialSampler
-from parakeet.models.deepvoice3 import Encoder, Decoder, Converter, DeepVoice3
+from parakeet.models.deepvoice3 import Encoder, Decoder, Converter, DeepVoice3, ConvSpec
 from parakeet.models.deepvoice3.loss import TTSLoss
 from parakeet.utils.layer_tools import summary
 
@@ -128,7 +127,7 @@ if __name__ == "__main__":
         use_memory_mask = model_config["use_memory_mask"]
         query_position_rate = model_config["query_position_rate"]
         key_position_rate = model_config["key_position_rate"]
-        window_behind = model_config["window_behind"]
+        window_backward = model_config["window_backward"]
         window_ahead = model_config["window_ahead"]
         key_projection = model_config["key_projection"]
         value_projection = model_config["value_projection"]
@@ -137,10 +136,10 @@ if __name__ == "__main__":
                          freeze_embedding, filter_size, encoder_channels,
                          n_mels, decoder_channels, r,
                          trainable_positional_encodings, use_memory_mask,
-                         query_position_rate, key_position_rate, window_behind,
-                         window_ahead, key_projection, value_projection,
-                         downsample_factor, linear_dim, use_decoder_states,
-                         converter_channels, dropout)
+                         query_position_rate, key_position_rate,
+                         window_backward, window_ahead, key_projection,
+                         value_projection, downsample_factor, linear_dim,
+                         use_decoder_states, converter_channels, dropout)
 
         # =========================loss=========================
         loss_config = config["loss"]

parakeet/data/__init__.py

@@ -0,0 +1,4 @@
+from .dataset import *
+from .datacargo import *
+from .sampler import *
+from .batch import *

parakeet/models/deepvoice3/__init__.py

@@ -1,4 +1,4 @@
-from parakeet.models.deepvoice3.encoder import Encoder
-from parakeet.models.deepvoice3.decoder import Decoder
+from parakeet.models.deepvoice3.encoder import Encoder, ConvSpec
+from parakeet.models.deepvoice3.decoder import Decoder, WindowRange
 from parakeet.models.deepvoice3.converter import Converter
 from parakeet.models.deepvoice3.model import DeepVoice3