MCPcopy Create free account

hub / github.com/cnlinxi/style-token_tacotron2 / functions

Functions388 in github.com/cnlinxi/style-token_tacotron2

⨍ Functions 388◇ Types & classes 41
↓ 1 callersFunction_add_lab
(txtlines, wav_dir_path)
utils/mtts.py:21
↓ 1 callersFunction_adjust
Make sure that segment word is smaller than prosody word
utils/mandarin_frontend.py:14
↓ 1 callersFunction_adjust_time_resolution
Adjust time resolution between audio and local condition
test_wavenet_feeder.py:37
↓ 1 callersMethod_adjust_time_resolution
Adjust time resolution between audio and local condition
wavenet_vocoder/feeder.py:355
↓ 1 callersFunction_arpabet_to_sequence
(text)
tacotron/utils/text.py:70
↓ 1 callersFunction_asfloat
(x)
wavenet_vocoder/util.py:159
↓ 1 callersFunction_asint
(x)
wavenet_vocoder/util.py:152
↓ 1 callersMethod_batch_size_checks
(self, batch_size, error_message)
tacotron/models/Architecture_wrappers.py:144
↓ 1 callersFunction_build_entry
takes the ith line that begin an interval and returns a dictionary of values
utils/textgrid.py:116
↓ 1 callersMethod_check_conditions
(self)
wavenet_vocoder/feeder.py:340
↓ 1 callersFunction_close_logfile
()
infolog.py:32
↓ 1 callersFunction_compute_attention
Computes the attention and alignments for a given attention_mechanism.
tacotron/models/attention.py:10
↓ 1 callersMethod_compute_weights
Generate weights by combining the direction of weight vector with it's norm
wavenet_vocoder/models/modules.py:95
↓ 1 callersMethod_data_dep_init
Data dependent initialization (Done by Calling a feedforward pass at step 0 of training)
wavenet_vocoder/models/modules.py:108
↓ 1 callersFunction_ensure_divisible
(length, divisible_by=256, lower=True)
test_wavenet_feeder.py:78
↓ 1 callersFunction_ensure_divisible
(length, divisible_by=256, lower=True)
wavenet_vocoder/feeder.py:403
↓ 1 callersMethod_exponential_learning_rate_decay
(self, init_lr, global_step, decay_rate=0.5, decay_steps=300000)
wavenet_vocoder/models/wavenet.py:485
↓ 1 callersFunction_find_tiers
(interval_lines, tier_lines, tiers)
utils/textgrid.py:72
↓ 1 callersMethod_get_linearized_weight
(self, in_channels)
wavenet_vocoder/models/modules.py:234
↓ 1 callersMethod_get_next_example
Get a single example (input, output, len_output) from disk
wavenet_vocoder/feeder.py:228
↓ 1 callersMethod_get_next_example
Gets a single example (input, mel_target, token_target, linear_target, mel_length) from_ disk
tacotron/feeder.py:179
↓ 1 callersMethod_get_output_lengths
(self, stop_tokens)
tacotron/synthesizer.py:276
↓ 1 callersFunction_get_pronunciation
(s)
tacotron/utils/cmudict.py:57
↓ 1 callersFunction_get_str_val
returns the last item in quotes from a string
utils/textgrid.py:138
↓ 1 callersMethod_get_test_groups
(self)
wavenet_vocoder/feeder.py:163
↓ 1 callersMethod_get_test_groups
(self)
tacotron/feeder.py:121
↓ 1 callersMethod_init_norm
Set the norm of the weight vector
wavenet_vocoder/models/modules.py:102
↓ 1 callersMethod_learning_rate_decay
(self, init_lr, global_step)
tacotron/models/tacotron.py:520
↓ 1 callersFunction_limit_time
Limit time resolution to save GPU memory.
test_wavenet_feeder.py:10
↓ 1 callersMethod_limit_time
Limit time resolution to save GPU memory.
wavenet_vocoder/feeder.py:345
↓ 1 callersFunction_linear_to_mel
(spectogram, hparams)
datasets/audio.py:168
↓ 1 callersFunction_location_sensitive_score
Impelements Bahdanau-style (cumulative) scoring function. This attention is described in: J. K. Chorowski, D. Bahdanau, D. Serdyuk, K. Cho, and Y.
tacotron/models/attention.py:38
↓ 1 callersFunction_mel_to_linear
(mel_spectrogram, hparams)
datasets/audio.py:174
↓ 1 callersFunction_mfa_align
(txtlines, wav_dir_path, output_path, acoustic_model_path)
utils/mtts.py:117
↓ 1 callersMethod_noam_learning_rate_decay
(self, init_lr, global_step, warmup_steps=4000.0)
wavenet_vocoder/models/wavenet.py:480
↓ 1 callersMethod_pad_input
(self, x, length)
tacotron/feeder.py:244
↓ 1 callersMethod_pad_input
(self, x, length)
tacotron/synthesizer.py:265
↓ 1 callersFunction_pad_inputs
(x, maxlen, _pad=0)
wavenet_vocoder/synthesizer.py:97
↓ 1 callersMethod_pad_target
(self, t, length)
tacotron/feeder.py:247
↓ 1 callersMethod_pad_target
(self, t, length)
tacotron/synthesizer.py:273
↓ 1 callersMethod_pad_token_target
(self, t, length)
tacotron/feeder.py:250
↓ 1 callersFunction_pre_pinyin_setting
fix pinyin error
utils/txt2pinyin.py:48
↓ 1 callersFunction_pre_pinyin_setting
fix pinyin error
utils/mtts.py:14
↓ 1 callersMethod_prepare_global_conditions
(self, global_condition, g_features)
wavenet_vocoder/feeder.py:333
↓ 1 callersMethod_prepare_inputs
(self, inputs, maxlen)
wavenet_vocoder/feeder.py:287
↓ 1 callersMethod_prepare_inputs
(self, inputs)
tacotron/feeder.py:230
↓ 1 callersMethod_prepare_inputs
(self, inputs)
tacotron/synthesizer.py:261
↓ 1 callersMethod_prepare_local_conditions
(self, local_condition, c_features)
wavenet_vocoder/feeder.py:311
↓ 1 callersMethod_prepare_targets
(self, targets, maxlen)
wavenet_vocoder/feeder.py:300
↓ 1 callersMethod_prepare_token_targets
(self, targets, alignment)
tacotron/feeder.py:239
↓ 1 callersFunction_round_down
(x, multiple)
wavenet_vocoder/feeder.py:399
↓ 1 callersMethod_round_down
(self, x, multiple)
tacotron/feeder.py:257
↓ 1 callersMethod_round_up
(self, x, multiple)
tacotron/synthesizer.py:257
↓ 1 callersFunction_round_up_tf
(x, multiple)
tacotron/models/modules.py:597
↓ 1 callersFunction_set_logger
(output_path)
utils/mtts.py:204
↓ 1 callersMethod_set_sons
设置son的father为自己，统计sons的个数，根据son更新自己的txt
utils/labcnp.py:30
↓ 1 callersFunction_sfs2label
(txtlines, output_path)
utils/mtts.py:175
↓ 1 callersFunction_should_keep_symbol
(s)
tacotron/utils/text.py:74
↓ 1 callersFunction_standard_sfs
Change csv_list like "0 0.21 sil phones" to standard format like "2100000 s"
utils/mtts.py:97
↓ 1 callersFunction_teacher_forcing_ratio_decay
(init_tfr, global_step, hparams)
tacotron/models/helpers.py:135
↓ 1 callersFunction_textgrid2sfs
(txtlines, output_path)
utils/mtts.py:139
↓ 1 callersFunction_txt_preprocess
(txtfile, output_path)
utils/mtts.py:68
↓ 1 callersMethoda
a1 | 前一音节/字的声调 a2 | 当前音节/字的声调 a3 | 后一音节/字的声调
utils/labcnp.py:109
↓ 1 callersFunctionadd_embedding_stats
(summary_writer, embedding_names, paths_to_meta, checkpoint_path)
tacotron/train.py:23
↓ 1 callersFunctionadd_eval_stats
(summary_writer, step, linear_loss, before_loss, after_loss, stop_token_loss, loss)
tacotron/train.py:63
↓ 1 callersFunctionadd_head_middle_tail_silence
Add a 'sil' tree node at the head or middle or tail of tree according to the phs_type
utils/labformat.py:97
↓ 1 callersMethodadd_optimizer
Adds optimizer to the graph. Supposes that initialize function has already been called.
wavenet_vocoder/models/wavenet.py:434
↓ 1 callersMethodadd_optimizer
Adds optimizer. Sets "gradients" and "optimize" fields. add_loss must have been called. Args: global_step: int32 scalar Tensor rep
tacotron/models/tacotron.py:452
↓ 1 callersFunctionadd_test_stats
(summary_writer, step, eval_loss)
wavenet_vocoder/train.py:35
↓ 1 callersFunctionadd_train_stats
(model)
wavenet_vocoder/train.py:23
↓ 1 callersFunctionadd_train_stats
(model, hparams)
tacotron/train.py:38
↓ 1 callersMethodb
b1 | 当前音节/字到语句开始字的距离 b2 | 当前音节/字到语句结束字的距离 b3 | 当前音节/字在词中的位置（正序） b4 | 当前音节/字在词中的位置（倒序） b5 | 当前音节/字在韵
utils/labcnp.py:124
↓ 1 callersFunctionbiaobei_cleaner
(input_path, output_path)
utils/biaobei_cleaner.py:19
↓ 1 callersMethodc
c1 | 前一个词的词性 c2 | 当前词的词性 c3 | 后一个词的词性 c4 | 前一个词的音节数目 c5 | 当前词中的音节数目 c6 | 后一个词的音节数目
utils/labcnp.py:167
↓ 1 callersFunctionchange2absd
(phone, csv_list)
utils/mtts.py:100
↓ 1 callersFunctioncheck_time_alignment
(hparams, batch, local_condition)
test_wavenet_feeder.py:71
↓ 1 callersFunctionconvert_four_number
此时传入的num已经是倒叙
utils/number2chinese.py:46
↓ 1 callersMethodd
d1 | 前一个韵律词的音节数目 d2 | 当前韵律词的音节数目 d3 | 后一个韵律词的音节数目 d4 | 当前韵律词在韵律短语的位置（正序） d5 | 当前韵律词在韵律短语的位置（倒序）
utils/labcnp.py:189
↓ 1 callersFunctiondiscretized_mix_logistic_loss
Discretized mix of logistic distributions loss. Note that it is assumed that input is scaled to [-1, 1] Args: y_hat: Tensor [batch_size, channel
wavenet_vocoder/models/mixture.py:18
↓ 1 callersMethode
e1 | 前一韵律短语的音节数目 e2 | 当前韵律短语的音节数目 e3 | 后一韵律短语的音节数目 e4 | 前一韵律短语的韵律词个数 e5 | 当前韵律短语的韵律词个数 e6
utils/labcnp.py:212
↓ 1 callersFunctioneval_step
Evaluate model during training. Supposes that model variables are averaged.
wavenet_vocoder/train.py:65
↓ 1 callersFunctionexpand_abbreviations
(text)
tacotron/utils/cleaners.py:49
↓ 1 callersFunctionexpand_numbers
(text)
tacotron/utils/cleaners.py:55
↓ 1 callersMethodf
f1 | 语句的语调类型 f2 | 语句的音节数目 f3 | 语句的词数目 f4 | 语句的韵律词数目 f5 | 语句的韵律短语数目
utils/labcnp.py:242
↓ 1 callersFunctiongaussian_maximum_likelihood_estimation_loss
compute the gaussian MLE loss
wavenet_vocoder/models/gaussian.py:5
↓ 1 callersFunctiongenerate_label
(txtfile, wav_dir_path, output_path, acoustic_model_path)
utils/mtts.py:225
↓ 1 callersFunctionget_first_node_of_tree
root_node's rhythm is #4
utils/labformat.py:92
↓ 1 callersFunctionget_groups
(args, hparams, meta, local_condition)
test_wavenet_feeder.py:21
↓ 1 callersMethodget_mask
(self, input_lengths, maxlen=None)
wavenet_vocoder/models/wavenet.py:497
↓ 1 callersFunctionget_sentences
(args)
synthesize.py:25
↓ 1 callersFunctionget_sentences
(args)
utils/interval_synthesis.py:33
↓ 1 callersFunctionget_some_inversed_samples
(training_data_path='training_data',output_inversed_path='tmp_inverse_wav_out',n_samples=5)
utils/check_slience_trim.py:13
↓ 1 callersFunctionis_raw
(s)
wavenet_vocoder/util.py:21
↓ 1 callersFunctionmain
()
train.py:92
↓ 1 callersFunctionmain
()
wavenet_preprocess.py:36
↓ 1 callersFunctionmain
()
test_wavenet_feeder.py:100
↓ 1 callersFunctionmain
()
preprocess.py:117
↓ 1 callersFunctionmain
()
synthesize.py:47
↓ 1 callersFunctionmain
(input_path,output_path,sample_rate=0.5)
utils/check_min_samples.py:10
↓ 1 callersFunctionmain
()
utils/interval_synthesis.py:55
← previousnext →101–200 of 388, ranked by callers