Update train_vits-2.py

train_vits-2.py

@@ -56,8 +56,8 @@ def mozilla_with_speaker(root_path, meta_file, **kwargs):
 
 
 dataset_config = BaseDatasetConfig(
-    # formatter="mozilla_with_speaker",
-    formatter="mozilla",
+    formatter="mozilla_with_speaker",
+    # formatter="mozilla",
     dataset_name="multi_persian",
     meta_file_train="metadata.csv",
     language="fa",
@@ -143,7 +143,7 @@ config = VitsConfig(
         ["یکی اسبی به عاریت خواست","changiz",None,"fa"],
     ],
     output_path=output_path,
-    datasets=[audio_config],
+    datasets=[dataset_config],
     
     # Enable the weighted sampler
     use_weighted_sampler=True,
@@ -175,7 +175,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # Load all the datasets samples and split traning and evaluation sets
 train_samples, eval_samples = load_tts_samples(
     config.datasets,
-    formatter=mozilla_with_speaker,
+    # formatter=mozilla_with_speaker,
     eval_split=True,
     eval_split_max_size=config.eval_split_max_size,
     eval_split_size=config.eval_split_size,