Update train_vits-0.py
Browse files- train_vits-0.py +7 -5
train_vits-0.py
CHANGED
@@ -36,15 +36,15 @@ def mozilla_with_speaker(root_path, meta_file, **kwargs): # pylint: disable=unu
|
|
36 |
|
37 |
|
38 |
dataset_config1 = BaseDatasetConfig(
|
39 |
-
meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset-famale"
|
40 |
)
|
41 |
|
42 |
dataset_config2 = BaseDatasetConfig(
|
43 |
-
meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset"
|
44 |
)
|
45 |
|
46 |
dataset_config3 = BaseDatasetConfig(
|
47 |
-
meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset-male"
|
48 |
)
|
49 |
|
50 |
|
@@ -126,8 +126,10 @@ train_samples, eval_samples = load_tts_samples(
|
|
126 |
speaker_manager = SpeakerManager()
|
127 |
speaker_manager.set_ids_from_data(train_samples + eval_samples, parse_key="speaker_name")
|
128 |
config.num_speakers = speaker_manager.num_speakers
|
129 |
-
|
130 |
-
|
|
|
|
|
131 |
|
132 |
# init model
|
133 |
model = Vits(config, ap, tokenizer, speaker_manager=speaker_manager)
|
|
|
36 |
|
37 |
|
38 |
dataset_config1 = BaseDatasetConfig(
|
39 |
+
formatter="mozilla" ,meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset-famale"
|
40 |
)
|
41 |
|
42 |
dataset_config2 = BaseDatasetConfig(
|
43 |
+
formatter="mozilla" ,meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset"
|
44 |
)
|
45 |
|
46 |
dataset_config3 = BaseDatasetConfig(
|
47 |
+
formatter="mozilla" ,meta_file_train="metadata.csv", path="/kaggle/input/persian-tts-dataset-male"
|
48 |
)
|
49 |
|
50 |
|
|
|
126 |
speaker_manager = SpeakerManager()
|
127 |
speaker_manager.set_ids_from_data(train_samples + eval_samples, parse_key="speaker_name")
|
128 |
config.num_speakers = speaker_manager.num_speakers
|
129 |
+
print("\n"*10)
|
130 |
+
print("#>"*10)
|
131 |
+
print(speaker_manager.speaker_names)
|
132 |
+
print("\n"*10)
|
133 |
|
134 |
# init model
|
135 |
model = Vits(config, ap, tokenizer, speaker_manager=speaker_manager)
|