Forward audio and text into the CLAP Parameters ---------- audio: torch.Tensor (batch_size, audio_length) the tim

r"""Get AudioSet 632 classes ID to label mapping.

Output dictionary from out.txt log file

r"""Load trained universal source separation model. Args: configs (Dict) checkpoint_path (str): path of the checkpoint to load

Get tar path from txt path

Get a dataset for wdsdataloader.

Initialize a Linear or Convolutional layer.

Initialize a GRU layer.

r"""Get linear warm up scheduler for LambdaLR. Args: step (int): global step warm_up_steps (int): steps for warm up reduc

returns list of pretrained models Returns a tuple (model_name, pretrain_tag) by default or 'name:tag' if as_str == True

return all pretrain tags for the specified model architecture

r"""Load pretrained pretrained audio neural networks (PANNs). Args: model_type: str, e.g., "Cnn14" checkpoint_path, str, e.g., "C

lock modules Args: unlocked_groups (int): leave last n layer groups unlocked (default: 0)

Call in an exception handler to ignore any exception, isssue a warning, and continue.

Loudness normalize a signal. Normalize an input signal to a user loudness in dB LKFS. Params ------- data : torch.Tensor

r"""Save a checkpoint every N steps.

Pad framewise_output to the same length as input frames. The pad value is the same as the value of the last frame. Args: framewise_outpu

Run a single prediction on the model

Preprocess a single sample for wdsdataloader.

r"""Remove silent frames.

r"""Repeat audio to length.

Load the model into memory to make running multiple predictions efficient

r"""called on every device.

Spectrogram to waveform. Args: input: (batch_size, segment_samples, channels_num) spectrogram: (batch_size, channels_num,

r"""Get train loader.

r"""Forward a mini-batch data to model, calculate loss function, and train for one step. A mini-batch data is evenly distributed to multiple

Waveform to spectrogram. Args: input: (batch_size, segment_samples, channels_num) Outputs: output: (batch_size,

Return a dictionary of the batch, with keys as the names of the fields.