birdnet.acoustic.models.v3_0 package

Submodules

birdnet.acoustic.models.v3_0.model module

class birdnet.acoustic.models.v3_0.model.AcousticDownloaderBaseV3_0

Bases: V3LabelDownloaderBase

AVAILABLE_LANGUAGES: OrderedSet[str] = OrderedSet(['bg', 'ca', 'cs', 'cy', 'da', 'de', 'en_us', 'es', 'es_ec', 'es_es', 'es_mx', 'fa', 'fi', 'fr', 'hr', 'ja', 'lt', 'nl', 'no', 'pl', 'pt', 'pt_pt', 'ru', 'sk', 'sr', 'sv', 'tr', 'uk', 'zh'])
classmethod ensure_labels_available()
Return type:

None

classmethod get_lang_file(lang)
Return type:

Path

class birdnet.acoustic.models.v3_0.model.AcousticModelV3_0(model_path, species_list, is_custom_model, backend_type, backend_kwargs)

Bases: AcousticModelBase

property backend_kwargs: dict[str, Any]
property backend_type: type[VersionedAcousticBackendProtocol]
encode(inp, /, *, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, speed=1.0, bandpass_fmin=0, bandpass_fmax=15000, half_precision=False, max_audio_duration_min=None, show_stats=None, progress_callback=None, device='CPU', segment_size_s=3.0, on_file_complete=None)

Run encoding with the BirdNET 3.0 model on files or paths to obtain embeddings.

Return type:

AcousticEncodingResultBase

Parameters:

  • inp – Path(s) or string(s) pointing to audio files to encode.

  • n_producers – Processes tasked with producing audio batches.

  • n_workers – Optional worker count for backend processing.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • device – Target device(s) for running the backend.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

  • on_file_complete – Optional callback fired once per input file as soon as that file is fully processed, receiving a single-file AcousticFileEncodingResult (invalid files are reported with their input marked unprocessable). Enables streaming per-file persistence. Invoked from a background thread with a copy of the caller’s context; file inputs only (not encode_arrays). A callback that raises cancels the run.

Returns:

Object containing embeddings for each file.

Return type:

AcousticEncodingResultBase

encode_arrays(inp, /, *, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, speed=1.0, bandpass_fmin=0, bandpass_fmax=15000, half_precision=False, max_audio_duration_min=None, show_stats=None, progress_callback=None, device='CPU', segment_size_s=3.0)

Run encoding with the BirdNET 3.0 model directly on in-memory audio arrays.

Return type:

AcousticEncodingResultBase

Parameters:

  • inp – Tuple(s) of (audio ndarray, sampling rate).

  • n_producers – Processes generating batches from the arrays.

  • n_workers – Optional worker count for backend processing.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • device – Target device(s) for running the backend.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

Returns:

Object containing embeddings for each input array.

Return type:

AcousticEncodingResultBase

encode_session(*, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, speed=1.0, bandpass_fmin=0, bandpass_fmax=15000, half_precision=False, max_audio_duration_min=None, show_stats=None, progress_callback=None, device='CPU', max_n_files=65536, segment_size_s=3.0, on_file_complete=None)

Create an encoding session with explicit resource configuration.

Return type:

AcousticEncodingSession

Parameters:

  • n_producers – Processes tasked with producing audio batches.

  • n_workers – Optional worker count for backend processing.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • device – Target device(s) for running the backend.

  • max_n_files – Upper bound on files to limit resource consumption.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

  • on_file_complete – Optional callback fired once per input file as soon as that file is fully processed, receiving a single-file AcousticFileEncodingResult (invalid files are reported with their input marked unprocessable). Enables streaming per-file persistence. Invoked from a background thread with a copy of the caller’s context; file inputs only (not run_arrays). A callback that raises cancels the run.

Returns:

Session capable of running encodings.

Return type:

AcousticEncodingSession

final classmethod get_embeddings_dim()
Return type:

int

final classmethod get_sample_rate()
Return type:

int

final classmethod get_segment_size_s()
Return type:

float

final classmethod get_segment_size_samples()
Return type:

int

final classmethod get_sig_fmax()
Return type:

int

final classmethod get_sig_fmin()
Return type:

int

final classmethod get_version()

Return the string label that identifies the acoustic model version.

Return type:

Literal['2.4', '3.0']

Returns:

Registered enum constant for the supported version.

Return type:

ACOUSTIC_MODEL_VERSIONS

property is_custom_model: bool
classmethod load(model_path, species_list, backend_type, backend_kwargs)
Return type:

Self

classmethod load_custom(model_path, species_list, backend_type, backend_kwargs, check_validity)
Return type:

Self

property model_path: Path
property n_species: int
predict(inp, /, *, top_k=5, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, bandpass_fmin=0, bandpass_fmax=15000, speed=1.0, apply_sigmoid=True, apply_softmax=False, sigmoid_sensitivity=1.0, default_confidence_threshold=0.1, custom_confidence_thresholds=None, custom_species_list=None, half_precision=False, max_audio_duration_min=None, device='CPU', show_stats=None, progress_callback=None, segment_size_s=3.0, on_file_complete=None)

Run prediction with the BirdNET 3.0 model on files or paths.

Scores are returned as the model emits them: the V3.0 exports apply the sigmoid inside the model graph. apply_sigmoid=True (the default) returns them unchanged and apply_sigmoid=False returns the identical raw model output, so confidence thresholds are sigmoid scores either way. A sigmoid_sensitivity other than 1.0 is applied as it is to the 2.4 models, on logits recovered from the scores, and is ignored with apply_sigmoid=False as there. apply_softmax=True raises a ValueError. Custom V3.0 models are expected to apply the sigmoid as well.

This method creates one prediction session for the call. That session shuts down its producer and worker processes before this method returns, including when inference raises an exception.

Return type:

AcousticPredictionResultBase

Parameters:

  • inp – Path(s) or string(s) pointing to audio files to analyze.

  • top_k – Number of highest-confidence results to return per segment.

  • n_producers – Processes tasked with producing audio batches.

  • n_workers – Number of inference worker processes. None uses the number of physical CPU cores. Pass a fixed integer to meet a scheduler or container process limit. Each worker holds its own copy of the model, so a high count raises peak memory use. In the cases the pipeline guards against, a worker killed by the operating system to reclaim memory, including while it is processing a batch, is reported as an error naming the dead worker rather than hanging the call (see issue #73). Lowering n_workers or batch_size reduces peak memory and with it how likely such a kill is, but cannot rule it out.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • apply_sigmoid – The 3.0 exports apply the sigmoid inside the model graph, so False returns the model’s scores unchanged, as does True (the default) at the default sigmoid_sensitivity of 1.0. False ignores sigmoid_sensitivity.

  • apply_softmax – Must stay False; True raises a ValueError for the 3.0 models.

  • sigmoid_sensitivity – Shifts the logits, recovered from the model’s scores, by 10 * (sigmoid_sensitivity - 1) before the sigmoid, so values above 1.0 raise the scores and values below lower them; between 0.5 and 1.5, None means 1.0. Ignored with apply_sigmoid=False.

  • default_confidence_threshold – Base threshold to emit a detection, as a sigmoid score (range 0 to 1).

  • custom_confidence_thresholds – Species-specific override thresholds.

  • custom_species_list – Path or iterable defining a subset of species.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • device – Target device(s) for running the backend.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

  • on_file_complete – Optional callback fired once per input file as soon as that file is fully processed, receiving a single-file AcousticFilePredictionResult (invalid files are reported with their input marked unprocessable). Enables streaming per-file persistence (e.g. resumable analysis). Invoked from a background thread with a copy of the caller’s context; file inputs only (not predict_arrays). A callback that raises cancels the run.

Returns:

Object containing detected species and confidence

scores.

Return type:

AcousticPredictionResultBase

predict_arrays(inp, /, *, top_k=5, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, bandpass_fmin=0, bandpass_fmax=15000, speed=1.0, apply_sigmoid=True, apply_softmax=False, sigmoid_sensitivity=1.0, default_confidence_threshold=0.1, custom_confidence_thresholds=None, custom_species_list=None, half_precision=False, max_audio_duration_min=None, device='CPU', show_stats=None, progress_callback=None, segment_size_s=3.0)

Run prediction with the BirdNET 3.0 model directly on in-memory audio arrays.

Return type:

AcousticPredictionResultBase

Parameters:

  • inp – Tuple(s) of (audio ndarray, sampling rate).

  • top_k – Number of highest-confidence results to return per segment.

  • n_producers – Processes generating batches from the arrays.

  • n_workers – Optional worker count for backend processing.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • apply_sigmoid – The 3.0 exports apply the sigmoid inside the model graph, so False returns the model’s scores unchanged, as does True (the default) at the default sigmoid_sensitivity of 1.0. False ignores sigmoid_sensitivity.

  • apply_softmax – Must stay False; True raises a ValueError for the 3.0 models.

  • sigmoid_sensitivity – Shifts the logits, recovered from the model’s scores, by 10 * (sigmoid_sensitivity - 1) before the sigmoid, so values above 1.0 raise the scores and values below lower them; between 0.5 and 1.5, None means 1.0. Ignored with apply_sigmoid=False.

  • default_confidence_threshold – Base threshold to emit a detection, as a sigmoid score (range 0 to 1).

  • custom_confidence_thresholds – Species-specific override thresholds.

  • custom_species_list – Path or iterable defining a subset of species.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • device – Target device(s) for running the backend.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

Returns:

Object containing detected species and confidence

scores.

Return type:

AcousticPredictionResultBase

predict_session(*, top_k=5, n_producers=1, n_workers=None, batch_size=1, prefetch_ratio=1, overlap_duration_s=0, speed=1.0, bandpass_fmin=0, bandpass_fmax=15000, apply_sigmoid=True, apply_softmax=False, sigmoid_sensitivity=1.0, default_confidence_threshold=0.1, custom_confidence_thresholds=None, custom_species_list=None, half_precision=False, max_audio_duration_min=None, show_stats=None, progress_callback=None, device='CPU', max_n_files=65536, segment_size_s=3.0, on_file_complete=None)

Create a prediction session for the BirdNET 3.0 model.

Scores: every V3.0 export (tf, pb, pt, onnx — official and custom alike) applies the sigmoid inside the model graph, so scores leave the model with the sigmoid applied. apply_sigmoid=True (the default) returns them unchanged — no second sigmoid is applied — and apply_sigmoid=False returns the identical raw model output. A sigmoid_sensitivity other than 1.0 is applied as it is to the 2.4 models: the logits are recovered from the model’s scores and the pipeline’s sigmoid runs on them, so thresholds see the transformed scores; with apply_sigmoid=False the sensitivity is ignored, as for the 2.4 models. apply_softmax=True raises a ValueError.

The tf and pb exports (official and custom) have a static batch dimension of 1, so batch_size above 1 raises a ValueError for them; onnx and pt take any batch size.

Return type:

AcousticPredictionSession

Parameters:

  • top_k – Number of highest-confidence results to return per segment.

  • n_producers – Processes tasked with producing audio batches.

  • n_workers – Optional worker count for backend processing.

  • batch_size – Number of records evaluated per inference call.

  • prefetch_ratio – How many batches to decode ahead of processing.

  • overlap_duration_s – Seconds of overlap between sliding windows.

  • speed – Resampling multiplier to accommodate different recording speeds.

  • bandpass_fmin – Lower bound for the bandpass filter in Hz.

  • bandpass_fmax – Upper bound for the bandpass filter in Hz.

  • apply_sigmoid – The 3.0 exports apply the sigmoid inside the model graph, so False returns the model’s scores unchanged, as does True (the default) at the default sigmoid_sensitivity of 1.0. False ignores sigmoid_sensitivity.

  • apply_softmax – Must stay False; True raises a ValueError for the 3.0 models.

  • sigmoid_sensitivity – Shifts the logits, recovered from the model’s scores, by 10 * (sigmoid_sensitivity - 1) before the sigmoid, so values above 1.0 raise the scores and values below lower them; between 0.5 and 1.5, None means 1.0. Ignored with apply_sigmoid=False.

  • default_confidence_threshold – Base threshold to emit a detection, as a sigmoid score (range 0 to 1).

  • custom_confidence_thresholds – Species-specific override thresholds.

  • custom_species_list – Path or iterable defining a subset of species.

  • half_precision – Cast the model’s outputs to float16 after inference; inference itself runs at the model’s precision.

  • max_audio_duration_min – Longest single input, in minutes, the segment indices are sized for; None sizes them for any length. An input with more segments than they can hold raises an error.

  • show_stats – Level of statistics logging to emit.

  • progress_callback – Optional callback to report progress. Invoked from a background worker thread, inheriting a copy of the caller’s context (contextvars) as captured when the call starts.

  • device – Target device(s) for running the backend.

  • max_n_files – Upper bound on files to limit resource consumption.

  • segment_size_s – Length in seconds of the audio segments passed to the model. The official tf, onnx and pt exports accept lengths other than the default 3.0 s. pb and the custom classifier exports take only 3.0 s and raise a ValueError for any other length; a custom TFLite export loaded with check_validity=False and an explicit version fails during the run instead. overlap_duration_s must be smaller than this value.

  • on_file_complete – Optional callback fired once per input file as soon as that file is fully processed, receiving a single-file AcousticFilePredictionResult (invalid files are reported with their input marked unprocessable). Enables streaming per-file persistence (e.g. resumable analysis). Invoked from a background thread with a copy of the caller’s context; file inputs only (not run_arrays). A callback that raises cancels the run.

Returns:

Session capable of running predictions.

Return type:

AcousticPredictionSession

property species_list: OrderedSet[str]

birdnet.acoustic.models.v3_0.pb module

class birdnet.acoustic.models.v3_0.pb.AcousticPBBackendFP32V3_0(model_path, device_name, half_precision, **kwargs)

Bases: PBBackend, VersionedAcousticBackendProtocol

classmethod applies_sigmoid()

Whether the sigmoid is applied inside the model graph.

A model that applies the sigmoid inside its graph must not get the pipeline’s sigmoid on top: that squashes every score into [0.5, 0.73]. The prediction worker recovers the logits from such a model’s scores when a non-default sigmoid sensitivity asks for the pipeline sigmoid after all.

Return type:

bool

classmethod encoding_key()
Return type:

str | None

classmethod encoding_signature_name()
Return type:

str | None

classmethod fixed_input_samples()

The only input length, in samples, the model accepts, or None for any.

Exports compiled with a static input length fail inside the worker for any other segment length, so the session rejects such a length up front.

Return type:

int | None

classmethod input_key()
Return type:

str

classmethod max_batch_size()

The largest batch the model accepts, or None when the batch is dynamic.

Exports compiled with a static batch dimension fail inside the worker for anything larger, so the session rejects such a batch size up front.

Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_key()
Return type:

str

classmethod prediction_signature_name()
Return type:

str

classmethod supports_encoding()
Return type:

bool

class birdnet.acoustic.models.v3_0.pb.AcousticPBDownloaderV3_0

Bases: AcousticDownloaderBaseV3_0

classmethod get_model_path_and_labels(lang)
Return type:

tuple[Path, OrderedSet[str]]

class birdnet.acoustic.models.v3_0.pb.AcousticRavenBackendFP32V3_0(model_path, device_name, half_precision, **kwargs)

Bases: PBBackend, VersionedAcousticBackendProtocol

A Raven export from BirdNET-Analyzer on the 3.0 base.

Signature “basic” maps “inputs” (batch, 96000 samples) to “scores”, with the sigmoid applied; there is no embeddings signature. It wraps the official SavedModel, so it inherits its static batch of 1 and its fixed input length.

classmethod applies_sigmoid()

Whether the sigmoid is applied inside the model graph.

A model that applies the sigmoid inside its graph must not get the pipeline’s sigmoid on top: that squashes every score into [0.5, 0.73]. The prediction worker recovers the logits from such a model’s scores when a non-default sigmoid sensitivity asks for the pipeline sigmoid after all.

Return type:

bool

classmethod encoding_key()
Return type:

str | None

classmethod encoding_signature_name()
Return type:

str | None

classmethod fixed_input_samples()

The only input length, in samples, the model accepts, or None for any.

Exports compiled with a static input length fail inside the worker for any other segment length, so the session rejects such a length up front.

Return type:

int | None

classmethod input_key()
Return type:

str

classmethod max_batch_size()

The largest batch the model accepts, or None when the batch is dynamic.

Exports compiled with a static batch dimension fail inside the worker for anything larger, so the session rejects such a batch size up front.

Return type:

int | None

classmethod name()
Return type:

str

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_key()
Return type:

str

classmethod prediction_signature_name()
Return type:

str

classmethod supports_encoding()
Return type:

bool

birdnet.acoustic.models.v3_0.tf module

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: AcousticTFBackendCustomV3_0

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomFixedInputFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: _FixedInputV3_0, AcousticTFBackendCustomFP16V3_0

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomFixedInputNoEmbeddingsFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: _FixedInputV3_0, AcousticTFBackendCustomNoEmbeddingsFP16V3_0

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomFixedInputNoEmbeddingsV3_0(model_path, device_name, half_precision, **kwargs)

Bases: _FixedInputV3_0, AcousticTFBackendCustomNoEmbeddingsV3_0

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomFixedInputV3_0(model_path, device_name, half_precision, **kwargs)

Bases: _FixedInputV3_0, AcousticTFBackendCustomV3_0

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomNoEmbeddingsFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: AcousticTFBackendCustomNoEmbeddingsV3_0

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomNoEmbeddingsV3_0(model_path, device_name, half_precision, **kwargs)

Bases: AcousticTFBackendCustomV3_0

A custom 3.0 classifier whose file exposes no embeddings output.

classmethod supports_encoding()
Return type:

bool

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendCustomV3_0(model_path, device_name, half_precision, **kwargs)

Bases: CustomTFLiteBackend, VersionedAcousticBackendProtocol

A custom 3.0 classifier: signature outputs ‘predictions’ and ‘embeddings’.

classmethod model_version()
Return type:

Literal['2.4', '3.0']

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: _AcousticTFBackendV3_0

classmethod encoding_out_idx()
Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_out_idx()
Return type:

int

class birdnet.acoustic.models.v3_0.tf.AcousticTFBackendFP32V3_0(model_path, device_name, half_precision, **kwargs)

Bases: _AcousticTFBackendV3_0

classmethod encoding_out_idx()
Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_out_idx()
Return type:

int

class birdnet.acoustic.models.v3_0.tf.AcousticTFDownloaderV3_0

Bases: _AcousticSingleFileDownloaderV3_0

birdnet.acoustic.models.v3_0.pt module

class birdnet.acoustic.models.v3_0.pt.AcousticPTBackendFP32V3_0(model_path, device_name, half_precision, **kwargs)

Bases: TorchBackend, VersionedAcousticBackendProtocol

classmethod applies_sigmoid()

Whether the sigmoid is applied inside the model graph.

A model that applies the sigmoid inside its graph must not get the pipeline’s sigmoid on top: that squashes every score into [0.5, 0.73]. The prediction worker recovers the logits from such a model’s scores when a non-default sigmoid sensitivity asks for the pipeline sigmoid after all.

Return type:

bool

classmethod encoding_out_idx()
Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_out_idx()
Return type:

int

classmethod probe_input_size_samples()
Return type:

int

classmethod supports_encoding()
Return type:

bool

class birdnet.acoustic.models.v3_0.pt.AcousticPTDownloaderV3_0

Bases: _AcousticSingleFileDownloaderV3_0

birdnet.acoustic.models.v3_0.onnx module

class birdnet.acoustic.models.v3_0.onnx.AcousticOnnxBackendFP16V3_0(model_path, device_name, half_precision, **kwargs)

Bases: OnnxBackend, VersionedAcousticBackendProtocol

classmethod applies_sigmoid()

Whether the sigmoid is applied inside the model graph.

A model that applies the sigmoid inside its graph must not get the pipeline’s sigmoid on top: that squashes every score into [0.5, 0.73]. The prediction worker recovers the logits from such a model’s scores when a non-default sigmoid sensitivity asks for the pipeline sigmoid after all.

Return type:

bool

classmethod encoding_out_idx()
Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_out_idx()
Return type:

int

classmethod probe_input_size_samples()
Return type:

int

classmethod supports_encoding()
Return type:

bool

class birdnet.acoustic.models.v3_0.onnx.AcousticOnnxBackendFP32V3_0(model_path, device_name, half_precision, **kwargs)

Bases: OnnxBackend, VersionedAcousticBackendProtocol

classmethod applies_sigmoid()

Whether the sigmoid is applied inside the model graph.

A model that applies the sigmoid inside its graph must not get the pipeline’s sigmoid on top: that squashes every score into [0.5, 0.73]. The prediction worker recovers the logits from such a model’s scores when a non-default sigmoid sensitivity asks for the pipeline sigmoid after all.

Return type:

bool

classmethod encoding_out_idx()
Return type:

int | None

classmethod precision()
Return type:

Literal['int8', 'fp16', 'fp32']

classmethod prediction_out_idx()
Return type:

int

classmethod probe_input_size_samples()
Return type:

int

classmethod supports_encoding()
Return type:

bool

class birdnet.acoustic.models.v3_0.onnx.AcousticOnnxDownloaderV3_0

Bases: _AcousticSingleFileDownloaderV3_0

Module contents