Functions — github.com/ASLP-lab/SenSE

CodeHub MCPcopy Index your code

hub / github.com/ASLP-lab/SenSE / functions

Functions360 in github.com/ASLP-lab/SenSE

⨍ Functions 360◇ Types & classes 78

( self, query: Tensor, key: Tensor, value: Tensor,

src/sense/model/encoder/conformer/attention.py:72

(self, inputs: Tensor, mask: Optional[Tensor] = None)

src/sense/model/encoder/conformer/attention.py:143

(self, inputs: Tensor)

src/sense/model/encoder/conformer/convolution.py:64

(self, inputs: Tensor)

src/sense/model/encoder/conformer/convolution.py:104

(self, inputs: Tensor)

src/sense/model/encoder/conformer/convolution.py:148

(self, inputs: Tensor, input_lengths: Tensor)

src/sense/model/encoder/conformer/convolution.py:176

(self, inputs: Tensor, input_lengths: Tensor)

src/sense/model/encoder/conformer/convolution.py:212

(self, inputs: Tensor)

src/sense/model/encoder/conformer/encoder.py:108

Forward propagate a `inputs` for encoder training. Args: inputs (torch.FloatTensor): A input sequence passed to encoder

src/sense/model/encoder/conformer/encoder.py:181

(self, inputs: Tensor)

src/sense/model/encoder/conformer/activation.py:27

(self, inputs: Tensor)

src/sense/model/encoder/conformer/activation.py:40

Args: x : Input tensor B X T X C Returns: torch.Tensor: Encoded tensor B X T X C

src/sense/model/encoder/conformer/embedding.py:58

(self, inputs: Tensor)

src/sense/model/encoder/conformer/feed_forward.py:56

src/sense/eval/ecapa_tdnn.py:37

src/sense/eval/ecapa_tdnn.py:66

src/sense/eval/ecapa_tdnn.py:80

src/sense/eval/ecapa_tdnn.py:117

src/sense/eval/ecapa_tdnn.py:146

src/sense/eval/ecapa_tdnn.py:297

(x: torch.Tensor)

src/sense/model/modules2.py:138

Functiongelu_accurate

src/sense/model/modules2.py:130

Functionget_bigvgan_mel_spectrogram

( waveform, n_fft=1024, n_mel_channels=100, target_sample_rate=24000, hop_length=256,

src/sense/model/modules.py:31

Functionget_bigvgan_qwen_mel_spectrogram

Compute normalized mel spectrogram (in dB scale) from waveform using PyTorch. Returns: spec (Tensor): [B, num_mels, T'] normalized me

src/sense/model/modules.py:75

Functionget_conformer_mel_spectrogram

( waveform, n_fft=400, n_mel_channels=80, target_sample_rate=24000, hop_length=160, wi

src/sense/model/modules.py:183

Methodget_frame_len

src/sense/model/dataset.py:210

Functionget_librispeech_test

(metalst, gen_wav_dir, gpus, librispeech_test_clean_path, eval_ground_truth=False)

src/sense/eval/utils_eval.py:315

Functionget_seed_tts_test

(metalst, gen_wav_dir, gpus)

src/sense/eval/utils_eval.py:280

Functionget_vocos_mel_spectrogram

( waveform, n_fft=1024, n_mel_channels=100, target_sample_rate=24000, hop_length=256,

src/sense/model/modules.py:155

Functionget_whisper_mel_spectrogram

( waveform, n_fft=None, n_mel_channels=None, target_sample_rate=None, hop_length=None,

src/sense/model/modules.py:211

Functioninfer_batch_process

( ref_audio, noisy_ref_audio, ref_text, gen_text_batches, model_obj, vocoder, no_r

src/sense/infer/utils_infer.py:450

Functioninit_bert_params

Initialize the weights specific to the BERT Model. This overrides the default initializations depending on the specified arguments. 1

src/sense/model/modules2.py:166

src/sense/model/trainer_llm.py:140

src/sense/model/trainer_cfm.py:146

Functionlist_str_to_idx

( text: list[str] | list[list[str]], vocab_char_map: dict[str, int], # {char: idx} padding_value=

src/sense/model/utils.py:114

Functionlist_str_to_tensor

(text: list[str], padding_value=-1)

src/sense/model/utils.py:107

(cls, model_path)

src/sense/model/encoder/whisper/whisper_encoder.py:11

Methodlong_term_modeling

( self, source: torch.Tensor, padding_mask: Optional[torch.Tensor] = None, mas

src/sense/model/encoder/wavlm/wavlm.py:380

Methodmake_conv

src/sense/model/encoder/wavlm/wavlm.py:450

Functionmaybe_masked_mean

(t: float["b n d"], mask: bool["b n"] = None)

src/sense/model/utils.py:68

Functionmix_audio_with_snr

(clean_audio, noise_audio, snr_db_range=(-10, 10))

src/sense/model/dataset.py:37

Methodpad_to_multiple_of_320

src/sense/model/dataset.py:217

Functionprocess_audio_directory

src/sense/train/datasets/prepare_emilia_v2.py:21

Functionprocess_audio_file

Process a single audio file to create noisy version

src/sense/train/noisy_datasets/prepare_libritts_noisy.py:96

Functionprocess_audio_file

Process a single audio file by checking its existence and extracting duration.

src/sense/train/datasets/prepare_csv_wavs.py:67

Functionremove_silence_for_generated_wav

src/sense/infer/utils_infer.py:635

src/sense/model/trainer_cfm.py:456

Functionrun_asr_wer

src/sense/eval/utils_eval.py:404

Functionrun_sim

src/sense/eval/utils_eval.py:487

Functionrun_speechbertscore

src/sense/eval/utils_eval.py:549

Functionrun_wesim

src/sense/eval/utils_eval.py:528

Functionsave_spectrogram

(spectrogram, path)

src/sense/infer/utils_infer.py:650

Functionseed_everything

src/sense/model/utils.py:17

Functionsignal_handler

(signum, frame)

src/sense/train/datasets/prepare_csv_wavs.py:49

Functiontest_audio_functions

Test audio processing functions

src/sense/train/noisy_datasets/test_libritts_noisy.py:41

Functiontest_directory_structure

Test directory structure copying

src/sense/train/noisy_datasets/test_libritts_noisy.py:67

Functiontest_load_noise_list

Test loading noise list

src/sense/train/noisy_datasets/test_libritts_noisy.py:24

(self, train_dataset: Dataset, num_workers=16, resumable_with_seed: int = None)

src/sense/model/trainer_cfm.py:264

Functiontranscribe

(ref_audio, language=None)

src/sense/infer/utils_infer.py:177

Methodupdate_dropout

Update dropout probability of encoder

src/sense/model/encoder/conformer/encoder2.py:176

Methodupdate_dropout

Update dropout probability of encoder

src/sense/model/encoder/conformer/encoder.py:175

← previous301–360 of 360, ranked by callers