Fix import

TTS/tts/configs/tacotron_config.py

@@ -83,6 +83,8 @@ class TacotronConfig(BaseTTSConfig):
         ddc_r (int):
             reduction rate used by the coarse decoder when `double_decoder_consistency` is in use. Set this
             as a multiple of the `r` value. Defaults to 6.
+        speakers_file (str):
+            Path to the speaker mapping file for the Speaker Manager. Defaults to None.
         use_speaker_embedding (bool):
             enable / disable using speaker embeddings for multi-speaker models. If set True, the model is
             in the multi-speaker mode. Defaults to False.
@@ -176,6 +178,7 @@ class TacotronConfig(BaseTTSConfig):
     ddc_r: int = 6
 
     # multi-speaker settings
+    speakers_file: str = None
     use_speaker_embedding: bool = False
     speaker_embedding_dim: int = 512
     use_d_vector_file: bool = False

TTS/tts/models/forward_tts.py

@@ -261,7 +261,7 @@ class ForwardTTS(BaseTTS):
         # init speaker embedding layer
         if config.use_speaker_embedding and not config.use_d_vector_file:
             print(" > Init speaker_embedding layer.")
-            self.emb_g = nn.Embedding(self.args.num_speakers, self.args.hidden_channels)
+            self.emb_g = nn.Embedding(self.num_speakers, self.args.hidden_channels)
             nn.init.uniform_(self.emb_g.weight, -0.1, 0.1)
 
     @staticmethod

tests/data_tests/test_loader.py

@@ -44,13 +44,13 @@ class TestTTSDataset(unittest.TestCase):
         self.max_loader_iter = 4
         self.ap = AudioProcessor(**c.audio)
 
-    def _create_dataloader(self, batch_size, r, bgs):
+    def _create_dataloader(self, batch_size, r, bgs, start_by_longest=False):
 
         # load dataset
         meta_data_train, meta_data_eval = load_tts_samples(dataset_config, eval_split=True, eval_split_size=0.2)
         items = meta_data_train + meta_data_eval
 
-        tokenizer = TTSTokenizer.init_from_config(c)
+        tokenizer, _ = TTSTokenizer.init_from_config(c)
         dataset = TTSDataset(
             outputs_per_step=r,
             compute_linear_spec=True,
@@ -77,7 +77,7 @@ class TestTTSDataset(unittest.TestCase):
 
     def test_loader(self):
         if ok_ljspeech:
-            dataloader, dataset = self._create_dataloader(2, c.r, 0)
+            dataloader, dataset = self._create_dataloader(1, 1, 0)
 
             for i, data in enumerate(dataloader):
                 if i == self.max_loader_iter:

tests/tts_tests/test_align_tts_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.align_tts_config import AlignTTSConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_fast_pitch_speaker_emb_train.py

@@ -5,9 +5,9 @@ import shutil
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.config.shared_configs import BaseAudioConfig
 from TTS.tts.configs.fast_pitch_config import FastPitchConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
-config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
+config_path = os.path.join(get_tests_output_path(), "fast_pitch_speaker_emb_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")
 
 audio_config = BaseAudioConfig(

tests/tts_tests/test_fast_pitch_train.py

@@ -5,7 +5,7 @@ import shutil
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.config.shared_configs import BaseAudioConfig
 from TTS.tts.configs.fast_pitch_config import FastPitchConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_glow_tts_d-vectors_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.glow_tts_config import GlowTTSConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")
@@ -17,7 +17,6 @@ config = GlowTTSConfig(
     num_eval_loader_workers=0,
     text_cleaner="english_cleaners",
     use_phonemes=True,
-    use_espeak_phonemes=True,
     phoneme_language="en-us",
     phoneme_cache_path="tests/data/ljspeech/phoneme_cache/",
     run_eval=True,

tests/tts_tests/test_glow_tts_speaker_emb_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.glow_tts_config import GlowTTSConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")
@@ -17,7 +17,6 @@ config = GlowTTSConfig(
     num_eval_loader_workers=0,
     text_cleaner="english_cleaners",
     use_phonemes=True,
-    use_espeak_phonemes=True,
     phoneme_language="en-us",
     phoneme_cache_path="tests/data/ljspeech/phoneme_cache/",
     run_eval=True,

tests/tts_tests/test_glow_tts_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.glow_tts_config import GlowTTSConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_speedy_speech_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.speedy_speech_config import SpeedySpeechConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_speedy_speech_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_tacotron2_d-vectors_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.tacotron2_config import Tacotron2Config
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_tacotron2_speaker_emb_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.tacotron2_config import Tacotron2Config
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")
@@ -28,6 +28,7 @@ config = Tacotron2Config(
         "Be a voice, not an echo.",
     ],
     use_speaker_embedding=True,
+    num_speakers=4,
     max_decoder_steps=50,
 )
 

tests/tts_tests/test_tacotron2_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.tacotron2_config import Tacotron2Config
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_tacotron_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.tacotron_config import TacotronConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_vits_multilingual_speaker_emb_train.py

@@ -5,7 +5,7 @@ import shutil
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.config.shared_configs import BaseDatasetConfig
 from TTS.tts.configs.vits_config import VitsConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_vits_multilingual_train-d_vectors.py

@@ -5,7 +5,7 @@ import shutil
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.config.shared_configs import BaseDatasetConfig
 from TTS.tts.configs.vits_config import VitsConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_vits_speaker_emb_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.vits_config import VitsConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")

tests/tts_tests/test_vits_train.py

@@ -4,7 +4,7 @@ import shutil
 
 from tests import get_device_id, get_tests_output_path, run_cli
 from TTS.tts.configs.vits_config import VitsConfig
-from TTS.utils.trainer_utils import get_last_checkpoint
+from trainer import get_last_checkpoint
 
 config_path = os.path.join(get_tests_output_path(), "test_model_config.json")
 output_path = os.path.join(get_tests_output_path(), "train_outputs")