MCPcopy Create free account

hub / github.com/IIP-Sogang/olkavs-avspeech / functions

Functions234 in github.com/IIP-Sogang/olkavs-avspeech

⨍ Functions 234◇ Types & classes 60
Methodforward
Forward propagate a `inputs` for encoder training. Args: inputs (torch.FloatTensor): A input sequence passed to encoder
avsr/models/conformer/encoder.py:188
Methodforward
(self, inputs: Tensor)
avsr/models/conformer/activation.py:27
Methodforward
(self, inputs: Tensor)
avsr/models/conformer/activation.py:40
Methodforward
(self, length: int)
avsr/models/conformer/embedding.py:41
Methodforward
(self, inputs: Tensor)
avsr/models/conformer/feed_forward.py:56
Methodforward
Forward propagate a `inputs` and `targets` pair for training. Args: inputs (torch.FloatTensor): A input sequence passed
avsr/models/conformer/model.py:148
Functionget_batch_from_paths
(config, vocab, tr_video_paths, tr_audio_paths, tr_transcripts)
dataset/dataset.py:334
Methodget_lr
(self)
avsr/scheduler/abstract.py:8
Methodget_lr
(self)
avsr/scheduler/schedulers.py:39
Functiongrp2char
(sentence)
vocabulary/utils.py:69
Methodlabel_to_string
Converts label to string (number => Hangeul) Args: labels (numpy.ndarray): number label Returns: sentence
avsr/utils/korean_funcs.py:171
Methodlabel_to_string
(self, labels)
vocabulary/utils.py:151
Methodlabel_to_string
Converts label to string (number => Hangeul) Args: labels (numpy.ndarray): number label Returns: sentence
vocabulary/utils.py:170
Methodmetric
(self, s1: str, s2: str)
avsr/utils/metric.py:127
Methodmetric
Computes the Word Error Rate, defined as the edit distance between the two provided sentences after tokenizing to words. A
avsr/utils/metric.py:151
Methodmetric
Computes the Word Error Rate, defined as the edit distance between the two provided sentences after tokenizing to words. A
avsr/utils/metric.py:258
Functionord_labeling
(df)
avsr/utils/korean_funcs.py:29
Functionord_labeling
(df)
vocabulary/utils.py:29
Methodrecognize
Recognize input speech. This method consists of the forward of the encoder and the decode() of the decoder. Args: inputs
avsr/models/conformer/model.py:201
Methodreset
(self)
avsr/utils/metric.py:35
Methodreset
(self)
avsr/utils/metric.py:61
Methodscore
(self, text:str, features:torch.Tensor, embedder:torch.nn.Module, vocab2id
avsr/models/decoder.py:72
Functionseq_length_
(p)
dataset/dataset.py:397
Methodset_decoder
Setter for decoder
avsr/models/conformer/model.py:105
Methodset_encoder
Setter for encoder
avsr/models/conformer/model.py:101
Methodshuffle
Shuffle dataset
dataset/dataset.py:321
Functionsingle_infer
(config, vocab, search, metrics, audio_transform, tr_video_paths, tr_audio_paths, tr_korean_transcripts, tr_tr
inference.py:44
Methodstep
(self, *args, **kwargs)
avsr/scheduler/abstract.py:11
Methodstep
(self, on='step', loss=None, step=None, *args, **kwargs)
avsr/scheduler/schedulers.py:42
Functiontarget_length_
(p)
dataset/dataset.py:400
Functiontrain
(rank, world_size, config, vocab, trainset, validset, port)
train.py:85
Methodupdate_dropout
Update dropout probability of encoder
avsr/models/conformer/decoder.py:88
Methodupdate_dropout
Update dropout probability of encoder
avsr/models/conformer/encoder.py:182
Methodupdate_dropout
Update dropout probability of model
avsr/models/conformer/model.py:115
← previous201–234 of 234, ranked by callers