Add eval_split and eval_split_size in the call of load_tts_samples for all recipes

recipes/ljspeech/align_tts/train_aligntts.py

@@ -49,7 +49,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init model
 model = AlignTTS(config, ap, tokenizer)

recipes/ljspeech/fast_pitch/train_fast_pitch.py

@@ -84,7 +84,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init the model
 model = ForwardTTS(config, ap, tokenizer, speaker_manager=None)

recipes/ljspeech/fast_speech/train_fast_speech.py

@@ -83,7 +83,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init the model
 model = ForwardTTS(config, ap, tokenizer)

recipes/ljspeech/glow_tts/train_glowtts.py

@@ -60,7 +60,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # INITIALIZE THE MODEL
 # Models take a config object and a speaker manager as input

recipes/ljspeech/speedy_speech/train_speedy_speech.py

@@ -67,7 +67,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init model
 model = ForwardTTS(config, ap, tokenizer)

recipes/ljspeech/tacotron2-DCA/train_tacotron_dca.py

@@ -77,7 +77,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # INITIALIZE THE MODEL
 # Models take a config object and a speaker manager as input

recipes/ljspeech/tacotron2-DDC/train_tacotron_ddc.py

@@ -74,7 +74,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # INITIALIZE THE MODEL
 # Models take a config object and a speaker manager as input

recipes/ljspeech/vits_tts/train_vits.py

@@ -69,7 +69,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init model
 model = Vits(config, ap, tokenizer, speaker_manager=None)

recipes/multilingual/vits_tts/train_vits_tts.py

@@ -109,7 +109,7 @@ config.from_dict(config.to_dict())
 ap = AudioProcessor(**config.audio.to_dict())
 
 # load training samples
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader

recipes/vctk/fast_pitch/train_fast_pitch.py

@@ -71,7 +71,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader

recipes/vctk/fast_speech/train_fast_speech.py

@@ -69,7 +69,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader

recipes/vctk/glow_tts/train_glow_tts.py

@@ -69,7 +69,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader

recipes/vctk/speedy_speech/train_speedy_speech.py

@@ -69,7 +69,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader

recipes/vctk/tacotron-DDC/train_tacotron-DDC.py

@@ -72,7 +72,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it mainly handles speaker-id to speaker-name for the model and the data-loader

recipes/vctk/tacotron2-DDC/train_tacotron2-ddc.py

@@ -78,7 +78,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it mainly handles speaker-id to speaker-name for the model and the data-loader

recipes/vctk/tacotron2/train_tacotron2.py

@@ -78,7 +78,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it mainly handles speaker-id to speaker-name for the model and the data-loader

recipes/vctk/vits/train_vits.py

@@ -79,7 +79,7 @@ tokenizer, config = TTSTokenizer.init_from_config(config)
 # You can define your custom sample loader returning the list of samples.
 # Or define your custom formatter and pass it to the `load_tts_samples`.
 # Check `TTS.tts.datasets.load_tts_samples` for more details.
-train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True, eval_split_max_size=config.eval_split_max_size, eval_split_size=config.eval_split_size)
 
 # init speaker manager for multi-speaker training
 # it maps speaker-id to speaker-name in the model and data-loader