Modified data.py to generate masks as models inputs

examples/fastspeech/README.md

@@ -55,7 +55,7 @@ python -m paddle.distributed.launch --selected_gpus=0,1,2,3 --log_dir ./mylog tr
 --config_path='config/fastspeech.yaml' \
 ```
 
-if you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--fastspeech_step``
+If you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--fastspeech_step``
 
 For more help on arguments: 
 ``python train.py --help``.

examples/fastspeech/configs/synthesis.yaml

@@ -3,8 +3,8 @@ audio:
   n_fft: 2048
   sr: 22050
   preemphasis: 0.97
-  hop_length: 275
+  hop_length: 256
-  win_length: 1102
+  win_length: 1024
   power: 1.2
   min_level_db: -100
   ref_level_db: 20

examples/fastspeech/parse.py

@@ -17,6 +17,9 @@ def add_config_options_to_parser(parser):
         help="use gpu or not during training.")
     parser.add_argument('--use_data_parallel', type=int, default=0,
         help="use data parallel or not during training.")
+    parser.add_argument('--alpha', type=float, default=1.0,
+        help="The hyperparameter to determine the length of the expanded sequence \
+                mel, thereby controlling the voice speed.")
 
     parser.add_argument('--data_path', type=str, default='./dataset/LJSpeech-1.1',
         help="the path of dataset.")

examples/fastspeech/synthesis.py

@@ -11,6 +11,7 @@ import paddle.fluid.dygraph as dg
 from parakeet.g2p.en import text_to_sequence
 from parakeet import audio
 from parakeet.models.fastspeech.fastspeech import FastSpeech
+from parakeet.models.transformer_tts.utils import *
 
 def load_checkpoint(step, model_path):
     model_dict, _ = fluid.dygraph.load_dygraph(os.path.join(model_path, step))
@@ -41,11 +42,22 @@ def synthesis(text_input, args):
         model.eval()
 
         text = np.asarray(text_to_sequence(text_input))
-        text = fluid.layers.unsqueeze(dg.to_variable(text),[0])
+        text = np.expand_dims(text, axis=0)
         pos_text = np.arange(1, text.shape[1]+1)
-        pos_text = fluid.layers.unsqueeze(dg.to_variable(pos_text),[0])
+        pos_text = np.expand_dims(pos_text, axis=0)
+        enc_non_pad_mask = get_non_pad_mask(pos_text).astype(np.float32)
+        enc_slf_attn_mask = get_attn_key_pad_mask(pos_text, text).astype(np.float32)
+        
+        text = dg.to_variable(text)
+        pos_text = dg.to_variable(pos_text)
+        enc_non_pad_mask = dg.to_variable(enc_non_pad_mask)
+        enc_slf_attn_mask = dg.to_variable(enc_slf_attn_mask)
 
-        mel_output, mel_output_postnet = model(text, pos_text, alpha=args.alpha)
+        mel_output, mel_output_postnet = model(text, pos_text, alpha=args.alpha,
+                                               enc_non_pad_mask=enc_non_pad_mask,
+                                                enc_slf_attn_mask=enc_slf_attn_mask,
+                                                dec_non_pad_mask=None,
+                                                dec_slf_attn_mask=None)
 
         _ljspeech_processor = audio.AudioProcessor(
             sample_rate=cfg['audio']['sr'], 

examples/fastspeech/train.py

@@ -8,6 +8,7 @@ from parse import add_config_options_to_parser
 from pprint import pprint
 from ruamel import yaml
 from tqdm import tqdm
+from matplotlib import cm
 from collections import OrderedDict
 from tensorboardX import SummaryWriter
 import paddle.fluid.dygraph as dg
@@ -77,18 +78,32 @@ def main(args):
 
             for i, data in enumerate(pbar):
                 pbar.set_description('Processing at epoch %d'%epoch)
-                character, mel, mel_input, pos_text, pos_mel, text_length, mel_lens = data
+                (character, mel, mel_input, pos_text, pos_mel, text_length, mel_lens,
+                enc_slf_mask, enc_query_mask, dec_slf_mask, enc_dec_mask, dec_query_slf_mask, dec_query_mask) = data
 
-                _, _, attn_probs, _, _, _ = transformerTTS(character, mel_input, pos_text, pos_mel)
+                _, _, attn_probs, _, _, _ = transformerTTS(character, mel_input, pos_text, pos_mel,
-                alignment = dg.to_variable(get_alignment(attn_probs, mel_lens, cfg['transformer_head'])).astype(np.float32)
+                                                           dec_slf_mask=dec_slf_mask, 
+                                                           enc_slf_mask=enc_slf_mask, enc_query_mask=enc_query_mask, 
+                                                           enc_dec_mask=enc_dec_mask, dec_query_slf_mask=dec_query_slf_mask,
+                                                           dec_query_mask=dec_query_mask)
+                alignment, max_attn = get_alignment(attn_probs, mel_lens, cfg['transformer_head'])
+                alignment = dg.to_variable(alignment).astype(np.float32)
 
+                if local_rank==0 and global_step % 5 == 1:
+                    x = np.uint8(cm.viridis(max_attn[8,:mel_lens.numpy()[8]]) * 255)
+                    writer.add_image('Attention_%d_0'%global_step, x, 0, dataformats="HWC")
+                
                 global_step += 1
                     
                 #Forward
                 result= model(character, 
                               pos_text, 
                               mel_pos=pos_mel,  
-                              length_target=alignment)
+                              length_target=alignment,
+                              enc_non_pad_mask=enc_query_mask,
+                              enc_slf_attn_mask=enc_slf_mask,
+                              dec_non_pad_mask=dec_query_slf_mask,
+                              dec_slf_attn_mask=dec_slf_mask)
                 mel_output, mel_output_postnet, duration_predictor_output, _, _ = result
                 mel_loss = layers.mse_loss(mel_output, mel)
                 mel_postnet_loss = layers.mse_loss(mel_output_postnet, mel)

examples/fastspeech/train.sh

@@ -1,6 +1,6 @@
 # train model
 # if you wish to resume from an exists model, uncomment --checkpoint_path and --fastspeech_step
-CUDA_VISIBLE_DEVICES=0\
+export CUDA_VISIBLE_DEVICES=0
 python -u train.py \
 --batch_size=32 \
 --epochs=10000 \

examples/transformer_tts/README.md

@@ -1,5 +1,5 @@
 # TransformerTTS
-Paddle fluid implementation of TransformerTTS, a neural TTS with Transformer. The implementation is based on [Neural Speech Synthesis with Transformer Network](https://arxiv.org/abs/1809.08895).
+PaddlePaddle fluid implementation of TransformerTTS, a neural TTS with Transformer. The implementation is based on [Neural Speech Synthesis with Transformer Network](https://arxiv.org/abs/1809.08895).
 
 ## Dataset
 
@@ -48,7 +48,7 @@ python -m paddle.distributed.launch --selected_gpus=0,1,2,3 --log_dir ./mylog tr
 --config_path='config/train_transformer.yaml' \
 ```
 
-if you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--transformer_step``
+If you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--transformer_step``
 
 For more help on arguments: 
 ``python train_transformer.py --help``.
@@ -76,7 +76,7 @@ python -m paddle.distributed.launch --selected_gpus=0,1,2,3 --log_dir ./mylog tr
 --data_path=${DATAPATH} \
 --config_path='config/train_vocoder.yaml' \
 ```
-if you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--vocoder_step``
+If you wish to resume from an exists model, please set ``--checkpoint_path`` and ``--vocoder_step``
 
 For more help on arguments: 
 ``python train_vocoder.py --help``.

examples/transformer_tts/configs/synthesis.yaml

@@ -8,4 +8,7 @@ audio:
   power: 1.2
   min_level_db: -100
   ref_level_db: 20
-  outputs_per_step: 1
+  outputs_per_step: 1
+
+hidden_size: 256
+embedding_size: 512

examples/transformer_tts/data.py

@@ -10,7 +10,8 @@ from parakeet import audio
 from parakeet.data.sampler import *
 from parakeet.data.datacargo import DataCargo
 from parakeet.data.batch import TextIDBatcher, SpecBatcher
-from parakeet.data.dataset import DatasetMixin, TransformDataset
+from parakeet.data.dataset import DatasetMixin, TransformDataset, CacheDataset
+from parakeet.models.transformer_tts.utils import *
 
 class LJSpeechLoader:
     def __init__(self, config, args, nranks, rank, is_vocoder=False, shuffle=True):
@@ -20,6 +21,8 @@ class LJSpeechLoader:
         metadata = LJSpeechMetaData(LJSPEECH_ROOT)
         transformer = LJSpeech(config)
         dataset = TransformDataset(metadata, transformer)
+        dataset = CacheDataset(dataset)
+
         sampler = DistributedSampler(len(metadata), nranks, rank, shuffle=shuffle)
 
         assert args.batch_size % nranks == 0
@@ -132,7 +135,15 @@ def batch_examples(batch):
     pos_mels = TextIDBatcher(pad_id=0)(pos_mels) #(B,T)
     mels = np.transpose(SpecBatcher(pad_value=0.)(mels), axes=(0,2,1)) #(B,T,num_mels)
     mel_inputs = np.transpose(SpecBatcher(pad_value=0.)(mel_inputs), axes=(0,2,1))#(B,T,num_mels)
-    return (texts, mels, mel_inputs, pos_texts, pos_mels, np.array(text_lens), np.array(mel_lens))
+    enc_slf_mask = get_attn_key_pad_mask(pos_texts, texts).astype(np.float32)
+    enc_query_mask = get_non_pad_mask(pos_texts).astype(np.float32)
+    dec_slf_mask = get_dec_attn_key_pad_mask(pos_mels,mel_inputs).astype(np.float32)
+    enc_dec_mask = get_attn_key_pad_mask(enc_query_mask[:,:,0], mel_inputs).astype(np.float32)
+    dec_query_slf_mask = get_non_pad_mask(pos_mels).astype(np.float32)
+    dec_query_mask = get_non_pad_mask(pos_mels).astype(np.float32)
+
+    return (texts, mels, mel_inputs, pos_texts, pos_mels, np.array(text_lens), np.array(mel_lens),
+            enc_slf_mask, enc_query_mask, dec_slf_mask, enc_dec_mask, dec_query_slf_mask, dec_query_mask)
 
 def batch_examples_vocoder(batch):
     mels=[]

examples/transformer_tts/synthesis.py

@@ -3,6 +3,7 @@ from scipy.io.wavfile import write
 from parakeet.g2p.en import text_to_sequence
 import numpy as np
 from tqdm import tqdm
+from matplotlib import cm
 from tensorboardX import SummaryWriter
 from ruamel import yaml
 import paddle.fluid as fluid
@@ -12,6 +13,7 @@ import argparse
 from parse import add_config_options_to_parser
 from pprint import pprint
 from collections import OrderedDict
+from parakeet.models.transformer_tts.utils import *
 from parakeet import audio
 from parakeet.models.transformer_tts.vocoder import Vocoder
 from parakeet.models.transformer_tts.transformer_tts import TransformerTTS
@@ -55,15 +57,17 @@ def synthesis(text_input, args):
         mel_input = dg.to_variable(np.zeros([1,1,80])).astype(np.float32)
         pos_text = np.arange(1, text.shape[1]+1)
         pos_text = fluid.layers.unsqueeze(dg.to_variable(pos_text),[0])
-        
+    
 
         pbar = tqdm(range(args.max_len))
-
         for i in pbar:
+            dec_slf_mask = get_triu_tensor(mel_input.numpy(), mel_input.numpy()).astype(np.float32)
+            dec_slf_mask = fluid.layers.cast(dg.to_variable(dec_slf_mask == 0), np.float32)
             pos_mel = np.arange(1, mel_input.shape[1]+1)
             pos_mel = fluid.layers.unsqueeze(dg.to_variable(pos_mel),[0])
-            mel_pred, postnet_pred, attn_probs, stop_preds, attn_enc, attn_dec = model(text, mel_input, pos_text, pos_mel)
+            mel_pred, postnet_pred, attn_probs, stop_preds, attn_enc, attn_dec = model(text, mel_input, pos_text, pos_mel, dec_slf_mask)
             mel_input = fluid.layers.concat([mel_input, postnet_pred[:,-1:,:]], axis=1)
+
         mag_pred = model_vocoder(postnet_pred)
 
         _ljspeech_processor = audio.AudioProcessor(
@@ -87,6 +91,21 @@ def synthesis(text_input, args):
             sound_norm=False)
 
         wav = _ljspeech_processor.inv_spectrogram(fluid.layers.transpose(fluid.layers.squeeze(mag_pred,[0]), [1,0]).numpy())
+        global_step = 0
+        for i, prob in enumerate(attn_probs):
+            for j in range(4):
+                x = np.uint8(cm.viridis(prob.numpy()[j]) * 255)
+                writer.add_image('Attention_%d_0'%global_step, x, i*4+j, dataformats="HWC")
+
+        for i, prob in enumerate(attn_enc):
+            for j in range(4):
+                x = np.uint8(cm.viridis(prob.numpy()[j]) * 255)
+                writer.add_image('Attention_enc_%d_0'%global_step, x, i*4+j, dataformats="HWC")
+
+        for i, prob in enumerate(attn_dec):
+            for j in range(4):
+                x = np.uint8(cm.viridis(prob.numpy()[j]) * 255)
+                writer.add_image('Attention_dec_%d_0'%global_step, x, i*4+j, dataformats="HWC")
         writer.add_audio(text_input, wav, 0, cfg['audio']['sr'])
         if not os.path.exists(args.sample_path):
             os.mkdir(args.sample_path)
@@ -97,4 +116,4 @@ if __name__ == '__main__':
     parser = argparse.ArgumentParser(description="Synthesis model")
     add_config_options_to_parser(parser)
     args = parser.parse_args()
-    synthesis("Transformer model is so fast!", args)
+    synthesis("They emphasized the necessity that the information now being furnished be handled with judgment and care.", args)

examples/transformer_tts/synthesis.sh

@@ -2,10 +2,10 @@
 # train model
 CUDA_VISIBLE_DEVICES=0 \
 python -u synthesis.py \
---max_len=50 \
+--max_len=600 \
 --transformer_step=160000 \
---vocoder_step=70000 \
+--vocoder_step=90000 \
---use_gpu=1
+--use_gpu=1 \
 --checkpoint_path='./checkpoint' \
 --log_dir='./log' \
 --sample_path='./sample' \

examples/transformer_tts/train_transformer.py

@@ -69,19 +69,24 @@ def main(args):
             pbar = tqdm(reader)
             for i, data in enumerate(pbar):
                 pbar.set_description('Processing at epoch %d'%epoch)
-                character, mel, mel_input, pos_text, pos_mel, text_length, _ = data
+                character, mel, mel_input, pos_text, pos_mel, text_length, _, enc_slf_mask, enc_query_mask, dec_slf_mask, enc_dec_mask, dec_query_slf_mask, dec_query_mask= data
 
                 global_step += 1
-                mel_pred, postnet_pred, attn_probs, stop_preds, attn_enc, attn_dec = model(character, mel_input, pos_text, pos_mel)
                 
-
+                mel_pred, postnet_pred, attn_probs, stop_preds, attn_enc, attn_dec = model(character, mel_input, pos_text, pos_mel, dec_slf_mask=dec_slf_mask, 
-                label = (pos_mel == 0).astype(np.float32)
+                                                                                           enc_slf_mask=enc_slf_mask, enc_query_mask=enc_query_mask, 
-                    
+                                                                                           enc_dec_mask=enc_dec_mask, dec_query_slf_mask=dec_query_slf_mask,
+                                                                                           dec_query_mask=dec_query_mask)
+                
+ 
                 mel_loss = layers.mean(layers.abs(layers.elementwise_sub(mel_pred, mel)))
                 post_mel_loss = layers.mean(layers.abs(layers.elementwise_sub(postnet_pred, mel)))
                 loss = mel_loss + post_mel_loss
+
+                
                 # Note: When used stop token loss the learning did not work.
                 if args.stop_token:
+                    label = (pos_mel == 0).astype(np.float32)
                     stop_loss = cross_entropy(stop_preds, label)
                     loss = loss + stop_loss
 
@@ -123,6 +128,7 @@ def main(args):
                                 x = np.uint8(cm.viridis(prob.numpy()[j*16]) * 255)
                                 writer.add_image('Attention_dec_%d_0'%global_step, x, i*4+j, dataformats="HWC")
                                 
+                
                 if args.use_data_parallel:
                     loss = model.scale_loss(loss)
                     loss.backward()
@@ -132,6 +138,7 @@ def main(args):
                 optimizer.minimize(loss, grad_clip = fluid.dygraph_grad_clip.GradClipByGlobalNorm(cfg['grad_clip_thresh']))
                 model.clear_gradients()
                 
+                
                 # save checkpoint
                 if local_rank==0 and global_step % args.save_step == 0:
                     if not os.path.exists(args.save_path):

examples/transformer_tts/train_transformer.sh

@@ -1,7 +1,7 @@
 
 # train model
 # if you wish to resume from an exists model, uncomment --checkpoint_path and --transformer_step
-CUDA_VISIBLE_DEVICES=0 \
+export CUDA_VISIBLE_DEVICES=2
 python -u train_transformer.py \
 --batch_size=32 \
 --epochs=10000 \

parakeet/data/dataset.py

@@ -1,5 +1,6 @@
 import six
 import numpy as np
+from tqdm import tqdm
 
 
 class DatasetMixin(object):
@@ -45,6 +46,17 @@ class TransformDataset(DatasetMixin):
         in_data = self._dataset[i]
         return self._transform(in_data)
 
+class CacheDataset(DatasetMixin):
+    def __init__(self, dataset):
+        self._dataset = dataset
+        pbar = tqdm(range(len(self._dataset)))
+        self._cache = [self._dataset[i] for i in pbar]
+
+    def __len__(self):
+        return len(self._dataset)
+
+    def get_example(self, i):
+        return self._cache[i]
 
 class TupleDataset(object):
     def __init__(self, *datasets):

parakeet/models/fastspeech/decoder.py

@@ -18,6 +18,7 @@ class Decoder(dg.Layer):
         super(Decoder, self).__init__()
 
         n_position = len_max_seq + 1
+        self.n_head = n_head
         self.pos_inp = get_sinusoid_encoding_table(n_position, d_model, padding_idx=0)
         self.position_enc = dg.Embedding(size=[n_position, d_model],
                                  padding_idx=0,
@@ -28,7 +29,7 @@ class Decoder(dg.Layer):
         for i, layer in enumerate(self.layer_stack):
             self.add_sublayer('fft_{}'.format(i), layer)
     
-    def forward(self, enc_seq, enc_pos):
+    def forward(self, enc_seq, enc_pos, non_pad_mask, slf_attn_mask=None):
         """
         Decoder layer of FastSpeech.
         
@@ -42,10 +43,7 @@ class Decoder(dg.Layer):
             dec_slf_attn_list (Variable), Shape(B, mel_T, mel_T), the decoder self attention list.
         """
         dec_slf_attn_list = []
-
+        slf_attn_mask = layers.expand(slf_attn_mask, [self.n_head, 1, 1])
-        # -- Prepare masks
-        slf_attn_mask = get_attn_key_pad_mask(seq_k=enc_pos, seq_q=enc_pos)
-        non_pad_mask = get_non_pad_mask(enc_pos)
 
         # -- Forward
         dec_output = enc_seq + self.position_enc(enc_pos)

parakeet/models/fastspeech/encoder.py

@@ -18,11 +18,12 @@ class Encoder(dg.Layer):
                  dropout=0.1):
         super(Encoder, self).__init__()
         n_position = len_max_seq + 1
+        self.n_head = n_head
 
-        self.src_word_emb = dg.Embedding(size=[n_src_vocab, d_model], padding_idx=0)
+        self.src_word_emb = dg.Embedding(size=[n_src_vocab, d_model], padding_idx=0,
+                                         param_attr=fluid.initializer.Normal(loc=0.0, scale=1.0))
         self.pos_inp = get_sinusoid_encoding_table(n_position, d_model, padding_idx=0)
         self.position_enc = dg.Embedding(size=[n_position, d_model],
-                                 padding_idx=0,
                                  param_attr=fluid.ParamAttr(
                                      initializer=fluid.initializer.NumpyArrayInitializer(self.pos_inp),
                                      trainable=False))
@@ -30,7 +31,7 @@ class Encoder(dg.Layer):
         for i, layer in enumerate(self.layer_stack):
             self.add_sublayer('fft_{}'.format(i), layer)
 
-    def forward(self, character, text_pos):
+    def forward(self, character, text_pos, non_pad_mask, slf_attn_mask=None):
         """
         Encoder layer of FastSpeech.
         
@@ -46,10 +47,7 @@ class Encoder(dg.Layer):
             enc_slf_attn_list (list<Variable>), Len(n_layers), Shape(B * n_head, text_T, text_T), the encoder self attention list.
         """
         enc_slf_attn_list = []
-        # -- prepare masks
+        slf_attn_mask = layers.expand(slf_attn_mask, [self.n_head, 1, 1])
-        # shape character (N, T)
-        slf_attn_mask = get_attn_key_pad_mask(seq_k=character, seq_q=character)
-        non_pad_mask = get_non_pad_mask(character)
 
         # -- Forward
         enc_output = self.src_word_emb(character) + self.position_enc(text_pos) #(N, T, C)
@@ -61,4 +59,4 @@ class Encoder(dg.Layer):
                 slf_attn_mask=slf_attn_mask)
             enc_slf_attn_list += [enc_slf_attn]
         
-        return enc_output, non_pad_mask, enc_slf_attn_list
+        return enc_output, enc_slf_attn_list

parakeet/models/fastspeech/fastspeech.py

@@ -1,7 +1,9 @@
 import math
+import numpy as np
 import paddle.fluid.dygraph as dg
 import paddle.fluid as fluid
 from parakeet.g2p.text.symbols import symbols
+from parakeet.models.transformer_tts.utils import *
 from parakeet.models.transformer_tts.post_convnet import PostConvNet
 from parakeet.models.fastspeech.length_regulator import LengthRegulator
 from parakeet.models.fastspeech.encoder import Encoder
@@ -54,7 +56,9 @@ class FastSpeech(dg.Layer):
                  dropout=0.1,
                  batchnorm_last=True)
 
-    def forward(self, character, text_pos, mel_pos=None, length_target=None, alpha=1.0):
+    def forward(self, character, text_pos, enc_non_pad_mask, dec_non_pad_mask,
+                 enc_slf_attn_mask=None, dec_slf_attn_mask=None,
+                 mel_pos=None, length_target=None, alpha=1.0):
         """
         FastSpeech model.
         
@@ -80,13 +84,15 @@ class FastSpeech(dg.Layer):
             dec_slf_attn_list (Variable), Shape(B, mel_T, mel_T), the decoder self attention list.
         """
 
-        encoder_output, non_pad_mask, enc_slf_attn_list = self.encoder(character, text_pos)
+        encoder_output, enc_slf_attn_list = self.encoder(character, text_pos, enc_non_pad_mask, slf_attn_mask=enc_slf_attn_mask)
         if fluid.framework._dygraph_tracer()._train_mode:
             
             length_regulator_output, duration_predictor_output = self.length_regulator(encoder_output,
                                                                                        target=length_target,
                                                                                        alpha=alpha)
-            decoder_output, dec_slf_attn_list = self.decoder(length_regulator_output, mel_pos)
+            decoder_output, dec_slf_attn_list = self.decoder(length_regulator_output, mel_pos, 
+                                                             dec_non_pad_mask, 
+                                                             slf_attn_mask=dec_slf_attn_mask)
 
             mel_output = self.mel_linear(decoder_output)
             mel_output_postnet = self.postnet(mel_output) + mel_output
@@ -94,7 +100,12 @@ class FastSpeech(dg.Layer):
             return mel_output, mel_output_postnet, duration_predictor_output, enc_slf_attn_list, dec_slf_attn_list
         else:
             length_regulator_output, decoder_pos = self.length_regulator(encoder_output, alpha=alpha)
-            decoder_output, _ = self.decoder(length_regulator_output, decoder_pos)
+            slf_attn_mask = get_triu_tensor(decoder_pos.numpy(), decoder_pos.numpy()).astype(np.float32)
+            slf_attn_mask = fluid.layers.cast(dg.to_variable(slf_attn_mask == 0), np.float32)
+            slf_attn_mask = dg.to_variable(slf_attn_mask)
+            dec_non_pad_mask = fluid.layers.unsqueeze((decoder_pos != 0).astype(np.float32), [-1])
+            decoder_output, _ = self.decoder(length_regulator_output, decoder_pos, dec_non_pad_mask, 
+                                             slf_attn_mask=slf_attn_mask)
             mel_output = self.mel_linear(decoder_output)
             mel_output_postnet = self.postnet(mel_output) + mel_output
 

parakeet/models/fastspeech/fft_block.py

@@ -12,7 +12,7 @@ class FFTBlock(dg.Layer):
         self.slf_attn = MultiheadAttention(d_model, d_k, d_v, num_head=n_head, is_bias=True, dropout=dropout, is_concat=False)
         self.pos_ffn = PositionwiseFeedForward(d_model, d_inner, filter_size =filter_size, padding =padding, dropout=dropout)
 
-    def forward(self, enc_input, non_pad_mask=None, slf_attn_mask=None):
+    def forward(self, enc_input, non_pad_mask, slf_attn_mask=None):
         """
         Feed Forward Transformer block in FastSpeech.
         
@@ -28,6 +28,7 @@ class FFTBlock(dg.Layer):
             slf_attn (Variable), Shape(B * n_head, T, T), the self attention.
         """
         output, slf_attn = self.slf_attn(enc_input, enc_input, enc_input, mask=slf_attn_mask)
+
         output *= non_pad_mask
 
         output = self.pos_ffn(output)

parakeet/models/fastspeech/length_regulator.py

@@ -121,11 +121,11 @@ class DurationPredictor(dg.Layer):
         out = layers.transpose(encoder_output, [0,2,1])
         out = self.conv1(out)
         out = layers.transpose(out, [0,2,1])
-        out = layers.dropout(layers.relu(self.layer_norm1(out)), self.dropout)
+        out = layers.dropout(layers.relu(self.layer_norm1(out)), self.dropout, dropout_implementation='upscale_in_train')
         out = layers.transpose(out, [0,2,1])
         out = self.conv2(out)
         out = layers.transpose(out, [0,2,1])
-        out = layers.dropout(layers.relu(self.layer_norm2(out)), self.dropout)
+        out = layers.dropout(layers.relu(self.layer_norm2(out)), self.dropout, dropout_implementation='upscale_in_train')
         out = layers.relu(self.linear(out))
         out = layers.squeeze(out, axes=[-1])
         

parakeet/models/fastspeech/utils.py

@@ -14,7 +14,7 @@ def get_alignment(attn_probs, mel_lens, n_head):
                 max_F = F
                 max_attn = attn
     alignment = compute_duration(max_attn, mel_lens)
-    return alignment
+    return alignment, max_attn
     
 def score_F(attn):
     max = np.max(attn, axis=-1)

parakeet/models/transformer_tts/cbhg.py

@@ -124,6 +124,7 @@ class CBHG(dg.Layer):
         conv_list = []
         conv_input = input_
         
+        
         for i, (conv, batchnorm) in enumerate(zip(self.conv_list, self.batchnorm_list)):
             conv_input = self._conv_fit_dim(conv(conv_input), i+1)
             conv_input = layers.relu(batchnorm(conv_input))

parakeet/models/transformer_tts/decoder.py

@@ -1,7 +1,7 @@
 import math
 import paddle.fluid.dygraph as dg
 import paddle.fluid as fluid
-from parakeet.modules.utils import *
+from parakeet.models.transformer_tts.utils import *
 from parakeet.modules.multihead_attention import MultiheadAttention
 from parakeet.modules.ffn import PositionwiseFeedForward
 from parakeet.models.transformer_tts.prenet import PreNet
@@ -11,6 +11,7 @@ class Decoder(dg.Layer):
     def __init__(self, num_hidden, config, num_head=4):
         super(Decoder, self).__init__()
         self.num_hidden = num_hidden
+        self.num_head = num_head
         param = fluid.ParamAttr()
         self.alpha = self.create_parameter(shape=(1,), attr=param, dtype='float32',
                         default_initializer = fluid.initializer.ConstantInitializer(value=1.0))
@@ -48,25 +49,21 @@ class Decoder(dg.Layer):
         self.postconvnet = PostConvNet(config['audio']['num_mels'], config['hidden_size'], 
                                        filter_size = 5, padding = 4, num_conv=5, 
                                        outputs_per_step=config['audio']['outputs_per_step'], 
-                                       use_cudnn = True)
+                                       use_cudnn=True)
 
-    def forward(self, key, value, query, c_mask, positional):
+    def forward(self, key, value, query, positional,  mask, m_mask=None, m_self_mask=None, zero_mask=None):
 
         # get decoder mask with triangular matrix
         
         if fluid.framework._dygraph_tracer()._train_mode:
-            m_mask = get_non_pad_mask(positional)
+            m_mask = layers.expand(m_mask, [self.num_head, 1, key.shape[1]])
-            mask = get_attn_key_pad_mask((positional==0).astype(np.float32), query)
+            m_self_mask = layers.expand(m_self_mask, [self.num_head, 1, query.shape[1]])
-            triu_tensor = dg.to_variable(get_triu_tensor(query.numpy(), query.numpy())).astype(np.float32)
+            mask = layers.expand(mask, [self.num_head, 1, 1])
-            mask = mask + triu_tensor
+            zero_mask = layers.expand(zero_mask, [self.num_head, 1, 1])
-            mask = fluid.layers.cast(mask == 0, np.float32)
+
-            
-            # (batch_size, decoder_len, encoder_len)
-            zero_mask = get_attn_key_pad_mask(layers.squeeze(c_mask,[-1]), query)
         else:
-            mask = get_triu_tensor(query.numpy(), query.numpy()).astype(np.float32)
+            m_mask, m_self_mask, zero_mask = None, None, None
-            mask = fluid.layers.cast(dg.to_variable(mask == 0), np.float32)
+        
-            m_mask, zero_mask = None, None
 
         # Decoder pre-network
         query = self.decoder_prenet(query)
@@ -79,18 +76,22 @@ class Decoder(dg.Layer):
         query = positional * self.alpha + query
 
         #positional dropout
-        query = fluid.layers.dropout(query, 0.1)
+        query = fluid.layers.dropout(query, 0.1, dropout_implementation='upscale_in_train')
+       
 
         # Attention decoder-decoder, encoder-decoder
         selfattn_list = list()
         attn_list = list()
         
+        
         for selfattn, attn, ffn in zip(self.selfattn_layers, self.attn_layers, self.ffns):
-            query, attn_dec = selfattn(query, query, query, mask = mask, query_mask = m_mask)
+            query, attn_dec = selfattn(query, query, query, mask = mask, query_mask = m_self_mask)
             query, attn_dot = attn(key, value, query, mask = zero_mask, query_mask = m_mask)
             query = ffn(query)
             selfattn_list.append(attn_dec)
             attn_list.append(attn_dot)
+            
+        
         # Mel linear projection
         mel_out = self.mel_linear(query)
         # Post Mel Network

parakeet/models/transformer_tts/encoder.py

@@ -9,11 +9,11 @@ class Encoder(dg.Layer):
     def __init__(self, embedding_size, num_hidden, num_head=4):
         super(Encoder, self).__init__()
         self.num_hidden = num_hidden
+        self.num_head = num_head
         param = fluid.ParamAttr(initializer=fluid.initializer.Constant(value=1.0))
         self.alpha = self.create_parameter(shape=(1, ), attr=param, dtype='float32')
         self.pos_inp = get_sinusoid_encoding_table(1024, self.num_hidden, padding_idx=0)
         self.pos_emb = dg.Embedding(size=[1024, num_hidden],
-                                 padding_idx=0,
                                  param_attr=fluid.ParamAttr(
                                      initializer=fluid.initializer.NumpyArrayInitializer(self.pos_inp),
                                      trainable=False))
@@ -23,17 +23,20 @@ class Encoder(dg.Layer):
         self.layers = [MultiheadAttention(num_hidden, num_hidden//num_head, num_hidden//num_head) for _ in range(3)]
         for i, layer in enumerate(self.layers):
             self.add_sublayer("self_attn_{}".format(i), layer)
-        self.ffns = [PositionwiseFeedForward(num_hidden, num_hidden*num_head, filter_size=1, use_cudnn = True) for _ in range(3)]
+        self.ffns = [PositionwiseFeedForward(num_hidden, num_hidden*num_head, filter_size=1, use_cudnn=True) for _ in range(3)]
         for i, layer in enumerate(self.ffns):
             self.add_sublayer("ffns_{}".format(i), layer)
 
-    def forward(self, x, positional):
+    def forward(self, x, positional, mask=None, query_mask=None):
+        
         if fluid.framework._dygraph_tracer()._train_mode:
-            query_mask = get_non_pad_mask(positional)
+            seq_len_key = x.shape[1]
-            mask = get_attn_key_pad_mask(positional, x)
+            query_mask = layers.expand(query_mask, [self.num_head, 1, seq_len_key])
+            mask = layers.expand(mask, [self.num_head, 1, 1])
         else:
             query_mask, mask = None, None
         
+    
         # Encoder pre_network
         x = self.encoder_prenet(x) #(N,T,C)
         
@@ -43,9 +46,9 @@ class Encoder(dg.Layer):
         
         x = positional * self.alpha + x #(N, T, C)
        
-
+        
         # Positional dropout
-        x = layers.dropout(x, 0.1)
+        x = layers.dropout(x, 0.1, dropout_implementation='upscale_in_train')
         
         # Self attention encoder
         attentions = list()
@@ -54,4 +57,4 @@ class Encoder(dg.Layer):
             x = ffn(x)
             attentions.append(attention)
 
-        return x, query_mask, attentions
+        return x, attentions

parakeet/models/transformer_tts/encoderprenet.py

@@ -14,7 +14,8 @@ class EncoderPrenet(dg.Layer):
         self.num_hidden = num_hidden
         self.use_cudnn = use_cudnn
         self.embedding = dg.Embedding( size = [len(symbols), embedding_size],
-                                        padding_idx = None)
+                                    padding_idx = 0,
+                                    param_attr=fluid.initializer.Normal(loc=0.0, scale=1.0))
         self.conv_list = []
         k = math.sqrt(1 / embedding_size)
         self.conv_list.append(Conv1D(num_channels = embedding_size, 
@@ -49,10 +50,12 @@ class EncoderPrenet(dg.Layer):
                                 bias_attr=fluid.ParamAttr(initializer = fluid.initializer.Uniform(low=-k, high=k)))
 
     def forward(self, x):
+        
         x = self.embedding(x) #(batch_size, seq_len, embending_size)
         x = layers.transpose(x,[0,2,1])
         for batch_norm, conv in zip(self.batch_norm_list, self.conv_list):
-            x = layers.dropout(layers.relu(batch_norm(conv(x))), 0.2)
+            x = layers.dropout(layers.relu(batch_norm(conv(x))), 0.2, 
+                               dropout_implementation='upscale_in_train')
         x = layers.transpose(x,[0,2,1]) #(N,T,C)
         x = self.projection(x)
 

parakeet/models/transformer_tts/post_convnet.py

@@ -76,11 +76,13 @@ class PostConvNet(dg.Layer):
             batch_norm = self.batch_norm_list[i]
             conv = self.conv_list[i]
             
-            input = layers.dropout(layers.tanh(batch_norm(conv(input)[:,:,:len])), self.dropout)
+            input = layers.dropout(layers.tanh(batch_norm(conv(input)[:,:,:len])), self.dropout, 
+                                   dropout_implementation='upscale_in_train')
         conv = self.conv_list[self.num_conv-1]
         input = conv(input)[:,:,:len]
         if self.batchnorm_last:
             batch_norm = self.batch_norm_list[self.num_conv-1]
-            input = layers.dropout(batch_norm(input), self.dropout)
+            input = layers.dropout(batch_norm(input), self.dropout, 
+                                   dropout_implementation='upscale_in_train')
         output = layers.transpose(input, [0,2,1])
         return output

parakeet/models/transformer_tts/prenet.py

@@ -34,6 +34,6 @@ class PreNet(dg.Layer):
         Returns:
             x (Variable), Shape(B, T, C), the result after pernet.
         """
-        x = layers.dropout(layers.relu(self.linear1(x)), self.dropout_rate)
+        x = layers.dropout(layers.relu(self.linear1(x)), self.dropout_rate, dropout_implementation='upscale_in_train')
-        x = layers.dropout(layers.relu(self.linear2(x)), self.dropout_rate)
+        x = layers.dropout(layers.relu(self.linear2(x)), self.dropout_rate, dropout_implementation='upscale_in_train')
         return x

parakeet/models/transformer_tts/transformer_tts.py

@@ -10,15 +10,14 @@ class TransformerTTS(dg.Layer):
         self.decoder = Decoder(config['hidden_size'], config)
         self.config = config
 
-    def forward(self, characters, mel_input, pos_text, pos_mel):
+    def forward(self, characters, mel_input, pos_text, pos_mel, dec_slf_mask, enc_slf_mask=None, enc_query_mask=None, enc_dec_mask=None, dec_query_slf_mask=None, dec_query_mask=None):
-        
+        key, attns_enc = self.encoder(characters, pos_text, mask=enc_slf_mask, query_mask=enc_query_mask)   
-        key, c_mask, attns_enc = self.encoder(characters, pos_text)
+       
-        
+        mel_output, postnet_output, attn_probs, stop_preds, attns_dec = self.decoder(key, key, mel_input, pos_mel, 
-        mel_output, postnet_output, attn_probs, stop_preds, attns_dec = self.decoder(key, key, mel_input, c_mask, pos_mel)
+                                                                        mask=dec_slf_mask, zero_mask=enc_dec_mask, 
-
+                                                                        m_self_mask=dec_query_slf_mask, m_mask=dec_query_mask )
         return mel_output, postnet_output, attn_probs, stop_preds, attns_enc, attns_dec
 
-
     
 
 

parakeet/models/transformer_tts/utils.py

@@ -35,7 +35,9 @@ def get_sinusoid_encoding_table(n_position, d_hid, padding_idx=None):
     return sinusoid_table
 
 def get_non_pad_mask(seq):
-    return layers.unsqueeze((seq != 0).astype(np.float32),[-1])
+    mask = (seq != 0).astype(np.float32)
+    mask = np.expand_dims(mask, axis=-1)
+    return mask
 
 def get_attn_key_pad_mask(seq_k, seq_q):
     ''' For masking out the padding part of key sequence. '''
@@ -43,7 +45,21 @@ def get_attn_key_pad_mask(seq_k, seq_q):
     # Expand to fit the shape of key query attention matrix.
     len_q = seq_q.shape[1]
     padding_mask = (seq_k != 0).astype(np.float32)
-    padding_mask = layers.expand(layers.unsqueeze(padding_mask,[1]), [1, len_q, 1]) 
+    padding_mask = np.expand_dims(padding_mask, axis=1)
+    padding_mask = padding_mask.repeat([len_q],axis=1)
+    padding_mask = (padding_mask == 0).astype(np.float32) * (-2 ** 32 + 1)
+    return padding_mask
+
+def get_dec_attn_key_pad_mask(seq_k, seq_q):
+    ''' For masking out the padding part of key sequence. '''
+
+    # Expand to fit the shape of key query attention matrix.
+    len_q = seq_q.shape[1]
+    padding_mask = (seq_k == 0).astype(np.float32)
+    padding_mask = np.expand_dims(padding_mask, axis=1)
+    triu_tensor = get_triu_tensor(seq_q, seq_q)
+    padding_mask = padding_mask.repeat([len_q],axis=1) + triu_tensor
+    padding_mask = (padding_mask != 0).astype(np.float32) * (-2 ** 32 + 1)
     return padding_mask
 
 def get_triu_tensor(seq_k, seq_q):

parakeet/modules/dynamic_gru.py

@@ -40,10 +40,10 @@ class DynamicGRU(dg.Layer):
                 i = inputs.shape[1] - 1 - i
             input_ = inputs[:, i:i + 1, :]
             input_ = layers.reshape(
-                input_, [-1, input_.shape[2]], inplace=False)
+                input_, [-1, input_.shape[2]])
             hidden, reset, gate = self.gru_unit(input_, hidden)
             hidden_ = layers.reshape(
-                hidden, [-1, 1, hidden.shape[1]], inplace=False)
+                hidden, [-1, 1, hidden.shape[1]])
             res.append(hidden_)
         if self.is_reverse:
             res = res[::-1]

parakeet/modules/ffn.py

@@ -45,7 +45,7 @@ class PositionwiseFeedForward(dg.Layer):
         x = self.w_2(layers.relu(self.w_1(x)))
         
         # dropout
-        x = layers.dropout(x, self.dropout)
+        x = layers.dropout(x, self.dropout, dropout_implementation='upscale_in_train')
 
         x = layers.transpose(x, [0,2,1])
         # residual connection

parakeet/modules/modules.py

@@ -211,7 +211,7 @@ class Conv1DGLU(dg.Layer):
 
         residual = x
         x = fluid.layers.dropout(
-            x, self.dropout, dropout_implementation="upscale_in_train")
+            x, self.dropout)
         x = self.conv(x)
 
         content, gate = fluid.layers.split(x, num_or_sections=2, dim=1)
@@ -241,7 +241,7 @@ class Conv1DGLU(dg.Layer):
 
         # add step input and produce step output
         x = fluid.layers.dropout(
-            x, self.dropout, dropout_implementation="upscale_in_train")
+            x, self.dropout)
         x = self.conv.add_input(x)
 
         content, gate = fluid.layers.split(x, num_or_sections=2, dim=1)

parakeet/modules/multihead_attention.py

@@ -47,17 +47,13 @@ class ScaledDotProductAttention(dg.Layer):
             attention (Variable), Shape(n_head * B, T, C), the attention of key.
         """
         # Compute attention score
-        attention = layers.matmul(query, key, transpose_y=True) #transpose the last dim in y
+        attention = layers.matmul(query, key, transpose_y=True, alpha=self.d_key**-0.5) #transpose the last dim in y
-        attention = attention / math.sqrt(self.d_key)
 
         # Mask key to ignore padding
         if mask is not None:
-            attention = attention * mask
-            mask = (mask == 0).astype(np.float32) * (-2 ** 32 + 1)
             attention = attention + mask
-        
         attention = layers.softmax(attention)
-        attention = layers.dropout(attention, dropout)
+        attention = layers.dropout(attention, dropout, dropout_implementation='upscale_in_train')
         
         # Mask query to ignore padding
         if query_mask is not None:
@@ -103,15 +99,10 @@ class MultiheadAttention(dg.Layer):
             result (Variable), Shape(B, T, C), the result of mutihead attention.
             attention (Variable), Shape(n_head * B, T, C), the attention of key.
         """
+
         batch_size = key.shape[0]
         seq_len_key = key.shape[1]
         seq_len_query = query_input.shape[1]
-
-        # repeat masks h times
-        if query_mask is not None:
-            query_mask = layers.expand(query_mask, [self.num_head, 1, seq_len_key])
-        if mask is not None:
-            mask = layers.expand(mask, (self.num_head, 1, 1))
         
         
         # Make multihead attention
@@ -123,15 +114,15 @@ class MultiheadAttention(dg.Layer):
         key = layers.reshape(layers.transpose(key, [2, 0, 1, 3]), [-1, seq_len_key, self.d_k])
         value = layers.reshape(layers.transpose(value, [2, 0, 1, 3]), [-1, seq_len_key, self.d_k])
         query = layers.reshape(layers.transpose(query, [2, 0, 1, 3]), [-1, seq_len_query, self.d_q])
-        
+    
         result, attention = self.scal_attn(key, value, query, mask=mask, query_mask=query_mask)
-        
+    
         # concat all multihead result
         result = layers.reshape(result, [self.num_head, batch_size, seq_len_query, self.d_q])
         result = layers.reshape(layers.transpose(result, [1,2,0,3]),[batch_size, seq_len_query, -1])
         if self.is_concat:
             result = layers.concat([query_input,result], axis=-1)
-        result = layers.dropout(self.fc(result), self.dropout)
+        result = layers.dropout(self.fc(result), self.dropout, dropout_implementation='upscale_in_train')
         result = result + query_input
         
         result = self.layer_norm(result)