Fix import errors and make errors more informative.

src/senselab/audio/tasks/speaker_diarization/api.py

@@ -4,12 +4,12 @@
 
 from senselab.audio.data_structures import Audio
 from senselab.audio.tasks.speaker_diarization.pyannote import diarize_audios_with_pyannote
-from senselab.utils.data_structures import DeviceType, PyannoteAudioModel, ScriptLine, SenselabModel
+from senselab.utils.data_structures import DeviceType, PyannoteAudioModel, ScriptLine
 
 
 def diarize_audios(
     audios: List[Audio],
-    model: SenselabModel = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main"),
+    model: Optional[PyannoteAudioModel] = None,
     num_speakers: Optional[int] = None,
     min_speakers: Optional[int] = None,
     max_speakers: Optional[int] = None,
@@ -29,6 +29,9 @@ def diarize_audios(
     Returns:
         List[List[ScriptLine]]: The list of script lines with speaker labels.
     """
+    if model is None:
+        model = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main")
+
     if isinstance(model, PyannoteAudioModel):
         return diarize_audios_with_pyannote(
             audios=audios,

src/senselab/audio/tasks/speaker_diarization/pyannote.py

@@ -46,7 +46,7 @@ def _get_pyannote_diarization_pipeline(
 
 def diarize_audios_with_pyannote(
     audios: List[Audio],
-    model: PyannoteAudioModel = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main"),
+    model: Optional[PyannoteAudioModel] = None,
     device: Optional[DeviceType] = None,
     num_speakers: Optional[int] = None,
     min_speakers: Optional[int] = None,
@@ -65,6 +65,8 @@ def diarize_audios_with_pyannote(
     Returns:
         List[ScriptLine]: A list of ScriptLine objects containing the diarization results.
     """
+    if model is None:
+        model = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main")
 
     def _annotation_to_script_lines(annotation: Annotation) -> List[ScriptLine]:
         """Convert a Pyannote annotation to a list of script lines.

src/senselab/audio/tasks/speaker_embeddings/api.py

@@ -6,12 +6,12 @@
 
 from senselab.audio.data_structures import Audio
 from senselab.audio.tasks.speaker_embeddings.speechbrain import SpeechBrainEmbeddings
-from senselab.utils.data_structures import DeviceType, SenselabModel, SpeechBrainModel
+from senselab.utils.data_structures import DeviceType, SpeechBrainModel
 
 
 def extract_speaker_embeddings_from_audios(
     audios: List[Audio],
-    model: SenselabModel = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main"),
+    model: Optional[SpeechBrainModel] = None,
     device: Optional[DeviceType] = None,
 ) -> List[torch.Tensor]:
     """Compute the speaker embedding of audio signals.
@@ -35,6 +35,9 @@ def extract_speaker_embeddings_from_audios(
         >>> print(embeddings[0].shape)
         torch.Size([192])
     """
+    if model is None:
+        model = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main")
+
     if isinstance(model, SpeechBrainModel):
         return SpeechBrainEmbeddings.extract_speechbrain_speaker_embeddings_from_audios(
             audios=audios, model=model, device=device

src/senselab/audio/tasks/speaker_embeddings/speechbrain.py

@@ -48,7 +48,7 @@ def _get_speechbrain_model(
     def extract_speechbrain_speaker_embeddings_from_audios(
         cls,
         audios: List[Audio],
-        model: SpeechBrainModel = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main"),
+        model: Optional[SpeechBrainModel] = None,
         device: Optional[DeviceType] = None,
     ) -> List[torch.Tensor]:
         """Compute the speaker embeddings of audio signals.
@@ -67,6 +67,9 @@ def extract_speechbrain_speaker_embeddings_from_audios(
             - Optimizing the computation by working in batches
             - Double-checking the input size of classifier.encode_batch
         """
+        if model is None:
+            model = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main")
+
         classifier = cls._get_speechbrain_model(model=model, device=device)
         # 16khz comes from the model cards of ecapa-tdnn, resnet, and xvector
         expected_sample_rate = 16000

src/senselab/audio/tasks/speaker_verification/speaker_verification.py

@@ -18,7 +18,7 @@
 
 def verify_speaker(
     audios: List[Tuple[Audio, Audio]],
-    model: SpeechBrainModel = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main"),
+    model: Optional[SpeechBrainModel] = None,
     device: Optional[DeviceType] = None,
     threshold: float = 0.25,
 ) -> List[Tuple[float, bool]]:
@@ -38,6 +38,8 @@ def verify_speaker(
                                   between the two samples, and the prediction is a boolean
                                   indicating if the two samples are from the same speaker.
     """
+    if model is None:
+        model = SpeechBrainModel(path_or_uri="speechbrain/spkrec-ecapa-voxceleb", revision="main")
     device = _select_device_and_dtype(compatible_devices=[DeviceType.CPU, DeviceType.CUDA])[0]
 
     scores_and_predictions = []

src/senselab/audio/tasks/speech_enhancement/api.py

@@ -4,12 +4,12 @@
 
 from senselab.audio.data_structures import Audio
 from senselab.audio.tasks.speech_enhancement.speechbrain import SpeechBrainEnhancer
-from senselab.utils.data_structures import DeviceType, SenselabModel, SpeechBrainModel
+from senselab.utils.data_structures import DeviceType, SpeechBrainModel
 
 
 def enhance_audios(
     audios: List[Audio],
-    model: SenselabModel = SpeechBrainModel(path_or_uri="speechbrain/sepformer-wham16k-enhancement", revision="main"),
+    model: Optional[SpeechBrainModel] = None,
     device: Optional[DeviceType] = None,
 ) -> List[Audio]:
     """Enhances all audios using the given model.
@@ -23,6 +23,9 @@ def enhance_audios(
     Returns:
         List[Audio]: The list of enhanced audio objects.
     """
+    if model is None:
+        model = SpeechBrainModel(path_or_uri="speechbrain/sepformer-wham16k-enhancement", revision="main")
+
     if isinstance(model, SpeechBrainModel):
         return SpeechBrainEnhancer.enhance_audios_with_speechbrain(audios=audios, model=model, device=device)
     else:

src/senselab/audio/tasks/speech_enhancement/speechbrain.py

@@ -46,9 +46,7 @@ def _get_speechbrain_model(
     def enhance_audios_with_speechbrain(
         cls,
         audios: List[Audio],
-        model: SpeechBrainModel = SpeechBrainModel(
-            path_or_uri="speechbrain/sepformer-wham16k-enhancement", revision="main"
-        ),
+        model: Optional[SpeechBrainModel] = None,
         device: Optional[DeviceType] = None,
         batch_size: int = 16,
     ) -> List[Audio]:
@@ -63,6 +61,9 @@ def enhance_audios_with_speechbrain(
         Returns:
             List[Audio]: The list of enhanced audio objects.
         """
+        if model is None:
+            model = SpeechBrainModel(path_or_uri="speechbrain/sepformer-wham16k-enhancement", revision="main")
+
         # Take the start time of the model initialization
         start_time_model = time.time()
         enhancer, device, _ = cls._get_speechbrain_model(model=model, device=device)

src/senselab/audio/tasks/speech_to_text/huggingface.py

@@ -68,7 +68,7 @@ def _get_hf_asr_pipeline(
     def transcribe_audios_with_transformers(
         cls,
         audios: List[Audio],
-        model: HFModel = HFModel(path_or_uri="openai/whisper-tiny"),
+        model: Optional[HFModel] = None,
         language: Optional[Language] = None,
         return_timestamps: Optional[str] = "word",
         max_new_tokens: int = 128,
@@ -80,7 +80,7 @@ def transcribe_audios_with_transformers(
 
         Args:
             audios (List[Audio]): The list of audio objects to be transcribed.
-            model (HFModel): The Hugging Face model used for transcription.
+            model (HFModel): The Hugging Face model used for transcription. (default is `openai/whisper-tiny`).
             language (Optional[Language]): The language of the audio (default is None).
             return_timestamps (Optional[str]): The level of timestamp details (default is "word").
             max_new_tokens (int): The maximum number of new tokens (default is 128).
@@ -91,6 +91,8 @@ def transcribe_audios_with_transformers(
         Returns:
             List[ScritpLine]: The list of script lines.
         """
+        if model is None:
+            model = HFModel(path_or_uri="openai/whisper-tiny")
 
         def _audio_to_huggingface_dict(audio: Audio) -> Dict:
             """Convert an Audio object to a dictionary that can be used by the transformers pipeline.

src/senselab/audio/tasks/text_to_speech/api.py

@@ -10,7 +10,7 @@
 
 def synthesize_texts(
     texts: List[str],
-    model: SenselabModel = HFModel(path_or_uri="suno/bark", revision="main"),
+    model: Optional[SenselabModel] = None,
     language: Optional[Language] = None,
     device: Optional[DeviceType] = None,
     targets: Optional[List[Audio | Tuple[Audio, str]]] = None,
@@ -43,6 +43,9 @@ def synthesize_texts(
     Returns:
         List[Audio]: The list of synthesized audio objects.
     """
+    if model is None:
+        model = HFModel(path_or_uri="suno/bark", revision="main")
+
     if targets is not None:
         assert len(targets) == len(texts), ValueError("Provided targets should be same length as texts")
 

src/senselab/audio/tasks/text_to_speech/huggingface.py

@@ -46,7 +46,7 @@ def _get_hf_tts_pipeline(
     def synthesize_texts_with_transformers(
         cls,
         texts: List[str],
-        model: HFModel = HFModel(path_or_uri="suno/bark", revision="main"),
+        model: Optional[HFModel] = None,
         device: Optional[DeviceType] = None,
         forward_params: Optional[Dict[str, Any]] = None,
     ) -> List[Audio]:
@@ -57,7 +57,7 @@ def synthesize_texts_with_transformers(
 
         Args:
             texts (List[str]): The list of text strings to be synthesized.
-            model (HFModel): The Hugging Face model used for synthesis.
+            model (HFModel): The Hugging Face model used for synthesis (default is `suno/bark`).
             device (Optional[DeviceType]): The device to run the model on (default is None).
             forward_params (Optional[Dict[str, Any]]): Additional parameters to pass to the forward function.
 
@@ -68,6 +68,8 @@ def synthesize_texts_with_transformers(
             - Add speaker embeddings as they do in here:
             https://huggingface.co/docs/transformers/tasks/text-to-speech
         """
+        if model is None:
+            model = HFModel(path_or_uri="suno/bark", revision="main")
         pipe = HuggingFaceTTS._get_hf_tts_pipeline(model=model, device=device)
 
         synthesized_audios = pipe(texts, forward_params=forward_params)

src/senselab/audio/tasks/text_to_speech/marstts.py

@@ -18,14 +18,14 @@ class Mars5TTS:
     @classmethod
     def _get_torch_tts_model(
         cls,
-        model: TorchModel = TorchModel(path_or_uri="Camb-ai/mars5-tts", revision="master"),
+        model: Optional[TorchModel] = None,
         language: Optional[Language] = Language(language_code="en"),
         device: Optional[DeviceType] = None,
     ) -> Tuple[torch.nn.Module, type]:
         """Get or create a Torch-based Mars5TTS model.
 
         Args:
-            model (TorchModel): The Torch model (default is "Camb-ai/mars5-tts").
+            model (TorchModel): The Torch model (currently only supports "Camb-ai/mars5-tts").
             language (Optional[Language]): The language of the text (default is Language(language_code="en")).
                 The only supported language is "en" for now.
             device (DeviceType): The device to run the model on (default is None). Supported devices are CPU and CUDA.
@@ -34,6 +34,9 @@ def _get_torch_tts_model(
             model: The Torch-based Mars5TTS model.
             config_class: The configuration class used by the model.
         """
+        if model is None:
+            model = TorchModel(path_or_uri="Camb-ai/mars5-tts", revision="master")
+
         if model.path_or_uri != "Camb-ai/mars5-tts" or model.revision != "master":
             raise NotImplementedError("Only the 'Camb-ai/mars5-tts' model is supported for now.")
         if language == Language(language_code="en"):
@@ -57,7 +60,7 @@ def synthesize_texts_with_mars5tts(
         cls,
         texts: List[str],
         targets: List[Tuple[Audio, str]],
-        model: TorchModel = TorchModel(path_or_uri="Camb-ai/mars5-tts", revision="master"),
+        model: Optional[TorchModel] = None,
         language: Optional[Language] = None,
         device: Optional[DeviceType] = None,
         deep_clone: bool = True,
@@ -95,6 +98,9 @@ def synthesize_texts_with_mars5tts(
 
         The original repo of the model is: https://github.com/Camb-ai/MARS5-TTS.
         """
+        if model is None:
+            model = TorchModel(path_or_uri="Camb-ai/mars5-tts", revision="master")
+
         # Take the start time of the model initialization
         start_time_model = time.time()
         my_model, config_class = cls._get_torch_tts_model(model, language, device)

src/senselab/audio/tasks/text_to_speech/styletts2.py

@@ -32,7 +32,7 @@ class StyleTTS2:
     @classmethod
     def _get_style_tts_2_model(
         cls,
-        model: TorchModel = TorchModel(path_or_uri="wilke0818/StyleTTS2-TorchHub", revision="main"),
+        model: Optional[TorchModel] = None,
         language: Optional[Language] = None,
         device: Optional[DeviceType] = None,
         pretrain_data: Optional[Literal["LibriTTS", "LJSpeech"]] = "LibriTTS",
@@ -54,6 +54,8 @@ def _get_style_tts_2_model(
         Returns:
             model: The Torch-based StyleTTS2 model.
         """
+        if model is None:
+            model = TorchModel(path_or_uri="wilke0818/StyleTTS2-TorchHub", revision="main")
         if language == Language(language_code="en"):
             model_name: str = "styletts2"  # This is the default model they have for English.
         else:
@@ -80,7 +82,7 @@ def synthesize_texts_with_style_tts_2(
         texts: List[str],
         target_audios: List[Audio],
         target_transcripts: List[Optional[str]],
-        model: TorchModel = TorchModel(path_or_uri="wilke0818/StyleTTS2-TorchHub", revision="main"),
+        model: Optional[TorchModel] = None,
         language: Optional[Language] = None,
         device: Optional[DeviceType] = None,
         pretrain_data: Optional[Literal["LibriTTS", "LJSpeech"]] = "LibriTTS",
@@ -132,6 +134,8 @@ def synthesize_texts_with_style_tts_2(
 
         The original repo of the model is: https://github.com/yl4579/StyleTTS2.
         """
+        if model is None:
+            model = TorchModel(path_or_uri="wilke0818/StyleTTS2-TorchHub", revision="main")
         nltk.download("punkt")
         nltk.download("punkt_tab")
         # Take the start time of the model initialization

src/senselab/audio/tasks/voice_activity_detection/api.py

@@ -4,29 +4,35 @@
 
 from senselab.audio.data_structures import Audio
 from senselab.audio.tasks.speaker_diarization.pyannote import diarize_audios_with_pyannote
-from senselab.utils.data_structures import DeviceType, PyannoteAudioModel, ScriptLine, SenselabModel
+from senselab.utils.data_structures import DeviceType, PyannoteAudioModel, ScriptLine
 
 
 def detect_human_voice_activity_in_audios(
     audios: List[Audio],
-    model: SenselabModel = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main"),
+    model: Optional[PyannoteAudioModel] = None,
     device: Optional[DeviceType] = None,
 ) -> List[List[ScriptLine]]:
     """Diarizes all audios using the given model.
 
     Args:
         audios (List[Audio]): The list of audio objects to be processed.
-        model (SenselabModel): The model used for voice activity detection.
+        model (Optional[PyannoteAudioModel]): The model used for voice activity detection
+            (default is `pyannote/speaker-diarization-3.1`).
         device (Optional[DeviceType]): The device to run the model on (default is None).
 
     Returns:
         List[List[ScriptLine]]: The list of script lines with voice label.
     """
+    if model is None:
+        model = PyannoteAudioModel(path_or_uri="pyannote/speaker-diarization-3.1", revision="main")
+
     if isinstance(model, PyannoteAudioModel):
         results = diarize_audios_with_pyannote(audios=audios, model=model, device=device)
         for sample in results:
             for chunk in sample:
                 chunk.speaker = "VOICE"
         return results
     else:
-        raise NotImplementedError("Only Pyannote models are supported for now.")
+        raise NotImplementedError(
+            "Only Pyannote models are supported for now. We aim to support more models in the future."
+        )

src/senselab/audio/tasks/voice_cloning/api.py

@@ -4,13 +4,13 @@
 
 from senselab.audio.data_structures import Audio
 from senselab.audio.tasks.voice_cloning.knnvc import KNNVC
-from senselab.utils.data_structures import DeviceType, SenselabModel, TorchModel
+from senselab.utils.data_structures import DeviceType, TorchModel
 
 
 def clone_voices(
     source_audios: List[Audio],
     target_audios: List[Audio],
-    model: SenselabModel = TorchModel(path_or_uri="bshall/knn-vc", revision="master"),
+    model: Optional[TorchModel] = None,
     device: Optional[DeviceType] = None,
     **kwargs: Any,  # noqa:ANN401
 ) -> List[Audio]:
@@ -28,7 +28,7 @@ def clone_voices(
             (e.g., words) will remain the same, but the voice sounds like the target.
         target_audios (List[Audio]): A list of audio samples whose voices will be extracted
             and used to replace the voices in the corresponding source audio samples.
-        model (SenselabModel, optional): The model to use for voice cloning. Currently,
+        model (TorchModel, optional): The model to use for voice cloning. Currently,
             only KNNVC (K-Nearest Neighbors Voice Conversion) is supported, encapsulated
             by the `TorchModel` class. `TorchModel` is a child class of `SenselabModel`
             and specifies the model and revision for cloning. Defaults to
@@ -53,6 +53,8 @@ def clone_voices(
     Todo:
         Add logging with timestamps.
     """
+    if model is None:
+        model = TorchModel(path_or_uri="bshall/knn-vc", revision="master")
     if len(source_audios) != len(target_audios):
         raise ValueError("The list of source and target audios must have the same length.")
 

src/senselab/audio/tasks/voice_cloning/knnvc.py

@@ -53,7 +53,7 @@ def clone_voices_with_knn_vc(
         cls,
         source_audios: List[Audio],
         target_audios: List[Audio],
-        model: TorchModel = TorchModel(path_or_uri="bshall/knn-vc", revision="master"),
+        model: Optional[TorchModel] = None,
         prematched_vocoder: bool = True,
         topk: int = 4,
         device: Optional[DeviceType] = None,
@@ -75,6 +75,8 @@ def clone_voices_with_knn_vc(
         Raises:
             ValueError: If the audio files are not mono or if the sampling rates are not supported.
         """
+        if model is None:
+            model = TorchModel(path_or_uri="bshall/knn-vc", revision="master")
         if not isinstance(prematched_vocoder, bool):
             raise TypeError("prematched_vocoder must be a boolean.")