utkarsh2299 commited on
Commit
7b749de
·
verified ·
1 Parent(s): f6f7f0b

Upload 22 files

Browse files
tamil_latest/female/model/config.yaml ADDED
@@ -0,0 +1,315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config: conf/tuning/train_fastspeech2.yaml
2
+ print_config: false
3
+ log_level: INFO
4
+ drop_last_iter: false
5
+ dry_run: false
6
+ iterator_type: sequence
7
+ valid_iterator_type: null
8
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
9
+ ngpu: 1
10
+ seed: 0
11
+ num_workers: 1
12
+ num_att_plot: 3
13
+ dist_backend: nccl
14
+ dist_init_method: env://
15
+ dist_world_size: 2
16
+ dist_rank: 0
17
+ local_rank: 0
18
+ dist_master_addr: localhost
19
+ dist_master_port: 53899
20
+ dist_launcher: null
21
+ multiprocessing_distributed: true
22
+ unused_parameters: false
23
+ sharded_ddp: false
24
+ cudnn_enabled: true
25
+ cudnn_benchmark: false
26
+ cudnn_deterministic: true
27
+ collect_stats: false
28
+ write_collected_feats: false
29
+ max_epoch: 1000
30
+ patience: null
31
+ val_scheduler_criterion:
32
+ - valid
33
+ - loss
34
+ early_stopping_criterion:
35
+ - valid
36
+ - loss
37
+ - min
38
+ best_model_criterion:
39
+ - - valid
40
+ - loss
41
+ - min
42
+ - - train
43
+ - loss
44
+ - min
45
+ keep_nbest_models: 5
46
+ nbest_averaging_interval: 0
47
+ grad_clip: 1.0
48
+ grad_clip_type: 2.0
49
+ grad_noise: false
50
+ accum_grad: 8
51
+ no_forward_run: false
52
+ resume: true
53
+ train_dtype: float32
54
+ use_amp: false
55
+ log_interval: null
56
+ use_matplotlib: true
57
+ use_tensorboard: true
58
+ create_graph_in_tensorboard: false
59
+ use_wandb: false
60
+ wandb_project: null
61
+ wandb_id: null
62
+ wandb_entity: null
63
+ wandb_name: null
64
+ wandb_model_log_interval: -1
65
+ detect_anomaly: false
66
+ use_adapter: false
67
+ adapter: lora
68
+ save_strategy: all
69
+ adapter_conf: {}
70
+ pretrain_path: null
71
+ init_param: []
72
+ ignore_init_mismatch: false
73
+ freeze_param: []
74
+ num_iters_per_epoch: 800
75
+ batch_size: 20
76
+ valid_batch_size: null
77
+ batch_bins: 3000000
78
+ valid_batch_bins: null
79
+ train_shape_file:
80
+ - exp/tts_stats_raw_char_None/train/text_shape.char
81
+ - exp/tts_stats_raw_char_None/train/speech_shape
82
+ valid_shape_file:
83
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
84
+ - exp/tts_stats_raw_char_None/valid/speech_shape
85
+ batch_type: numel
86
+ valid_batch_type: null
87
+ fold_length:
88
+ - 150
89
+ - 819200
90
+ sort_in_batch: descending
91
+ shuffle_within_batch: false
92
+ sort_batch: descending
93
+ multiple_iterator: false
94
+ chunk_length: 500
95
+ chunk_shift_ratio: 0.5
96
+ num_cache_chunks: 1024
97
+ chunk_excluded_key_prefixes: []
98
+ chunk_default_fs: null
99
+ chunk_max_abs_length: null
100
+ chunk_discard_short_samples: true
101
+ train_data_path_and_name_and_type:
102
+ - - dump/raw/tr_no_dev/text
103
+ - text
104
+ - text
105
+ - - duration_info/tr_no_dev/durations
106
+ - durations
107
+ - text_int
108
+ - - dump/raw/tr_no_dev/wav.scp
109
+ - speech
110
+ - sound
111
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
112
+ - pitch
113
+ - npy
114
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
115
+ - energy
116
+ - npy
117
+ valid_data_path_and_name_and_type:
118
+ - - dump/raw/dev/text
119
+ - text
120
+ - text
121
+ - - duration_info/dev/durations
122
+ - durations
123
+ - text_int
124
+ - - dump/raw/dev/wav.scp
125
+ - speech
126
+ - sound
127
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
128
+ - pitch
129
+ - npy
130
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
131
+ - energy
132
+ - npy
133
+ allow_variable_data_keys: false
134
+ max_cache_size: 0.0
135
+ max_cache_fd: 32
136
+ allow_multi_rates: false
137
+ valid_max_cache_size: null
138
+ exclude_weight_decay: false
139
+ exclude_weight_decay_conf: {}
140
+ optim: adam
141
+ optim_conf:
142
+ lr: 1.0
143
+ scheduler: noamlr
144
+ scheduler_conf:
145
+ model_size: 384
146
+ warmup_steps: 4000
147
+ token_list:
148
+ - <blank>
149
+ - <unk>
150
+ - <space>
151
+ - $
152
+ - ','
153
+ - .
154
+ - M
155
+ - q
156
+ - H
157
+ - k
158
+ - ख
159
+ - g
160
+ - घ
161
+ - ङ
162
+ - c
163
+ - C
164
+ - j
165
+ - J
166
+ - ञ
167
+ - ट
168
+ - ठ
169
+ - ड
170
+ - ढ
171
+ - ण
172
+ - t
173
+ - थ
174
+ - d
175
+ - ध
176
+ - n
177
+ - ऩ
178
+ - p
179
+ - P
180
+ - b
181
+ - B
182
+ - m
183
+ - y
184
+ - r
185
+ - ऱ
186
+ - l
187
+ - ള
188
+ - Z
189
+ - w
190
+ - श
191
+ - ष
192
+ - s
193
+ - h
194
+ - Y
195
+ - ऽ
196
+ - क
197
+ - K
198
+ - G
199
+ - z
200
+ - D
201
+ - T
202
+ - f
203
+ - ॠ
204
+ - ൺ
205
+ - N
206
+ - ർ
207
+ - ൽ
208
+ - ൾ
209
+ - a
210
+ - A
211
+ - i
212
+ - I
213
+ - u
214
+ - U
215
+ - R
216
+ - ऍ
217
+ - e
218
+ - E
219
+ - ऐ
220
+ - ऑ
221
+ - o
222
+ - O
223
+ - औ
224
+ - உ
225
+ - <sos/eos>
226
+ odim: null
227
+ model_conf: {}
228
+ use_preprocessor: true
229
+ token_type: char
230
+ bpemodel: null
231
+ non_linguistic_symbols: null
232
+ cleaner: null
233
+ g2p: null
234
+ feats_extract: fbank
235
+ feats_extract_conf:
236
+ n_fft: 8192
237
+ hop_length: 1024
238
+ win_length: 4096
239
+ fs: 48000
240
+ fmin: 0
241
+ fmax: 24000
242
+ n_mels: 160
243
+ normalize: global_mvn
244
+ normalize_conf:
245
+ stats_file: exp/tts_stats_raw_char_None/train/feats_stats.npz
246
+ tts: fastspeech2
247
+ tts_conf:
248
+ adim: 384
249
+ aheads: 2
250
+ elayers: 4
251
+ eunits: 1536
252
+ dlayers: 4
253
+ dunits: 1536
254
+ positionwise_layer_type: conv1d
255
+ positionwise_conv_kernel_size: 3
256
+ duration_predictor_layers: 2
257
+ duration_predictor_chans: 256
258
+ duration_predictor_kernel_size: 3
259
+ postnet_layers: 5
260
+ postnet_filts: 5
261
+ postnet_chans: 256
262
+ use_masking: true
263
+ use_scaled_pos_enc: true
264
+ encoder_normalize_before: true
265
+ decoder_normalize_before: true
266
+ reduction_factor: 1
267
+ init_type: xavier_uniform
268
+ init_enc_alpha: 1.0
269
+ init_dec_alpha: 1.0
270
+ transformer_enc_dropout_rate: 0.2
271
+ transformer_enc_positional_dropout_rate: 0.2
272
+ transformer_enc_attn_dropout_rate: 0.2
273
+ transformer_dec_dropout_rate: 0.2
274
+ transformer_dec_positional_dropout_rate: 0.2
275
+ transformer_dec_attn_dropout_rate: 0.2
276
+ pitch_predictor_layers: 5
277
+ pitch_predictor_chans: 256
278
+ pitch_predictor_kernel_size: 5
279
+ pitch_predictor_dropout: 0.5
280
+ pitch_embed_kernel_size: 1
281
+ pitch_embed_dropout: 0.0
282
+ stop_gradient_from_pitch_predictor: true
283
+ energy_predictor_layers: 2
284
+ energy_predictor_chans: 256
285
+ energy_predictor_kernel_size: 3
286
+ energy_predictor_dropout: 0.5
287
+ energy_embed_kernel_size: 1
288
+ energy_embed_dropout: 0.0
289
+ stop_gradient_from_energy_predictor: false
290
+ pitch_extract: dio
291
+ pitch_extract_conf:
292
+ fs: 48000
293
+ n_fft: 8192
294
+ hop_length: 1024
295
+ f0max: 400
296
+ f0min: 80
297
+ reduction_factor: 1
298
+ pitch_normalize: global_mvn
299
+ pitch_normalize_conf:
300
+ stats_file: exp/tts_stats_raw_char_None/train/pitch_stats.npz
301
+ energy_extract: energy
302
+ energy_extract_conf:
303
+ fs: 48000
304
+ n_fft: 8192
305
+ hop_length: 1024
306
+ win_length: 4096
307
+ reduction_factor: 1
308
+ energy_normalize: global_mvn
309
+ energy_normalize_conf:
310
+ stats_file: exp/tts_stats_raw_char_None/train/energy_stats.npz
311
+ required:
312
+ - output_dir
313
+ - token_list
314
+ version: '202402'
315
+ distributed: true
tamil_latest/female/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:998fdeb0db5bdaa330485aa81b5ec28f4cc3b6a8e49b24d94cbcd59fdbeb68a3
3
+ size 770
tamil_latest/female/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff5b28693f3b2d2ae0d9a0946eafb27b740347aa3fa184760882aae0c17920f3
3
+ size 2042
tamil_latest/female/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b41f330c8467b0bd3cd0c740a1205b23a6faec4aab36b69e4b84ff19aa548e8c
3
+ size 152128410
tamil_latest/female/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:298bf17efb880eeb8271c00b583aa5b3077d633d48f14a6cf36c5ed7239d9157
3
+ size 770
tamil_latest/female/model/text ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ shuffled_train_tamil_female_mono_06005 $uरsAgam irஉkkumbOdஉ iളamऐ tAnE warugiरadஉ appaडi unakkஉ enna uरsAgam waनdadஉ.
2
+ shuffled_train_tamil_female_mono_06006 $rAma ऐyaङgArउkku aङgirude tan paङgaളA kUrkkA aडutta paङgaളA AyAwuडan koञjik koणडஉ नiरpadஉ teriनdadஉ.
3
+ shuffled_train_tamil_female_mono_06007 $नamakkunnஉ kiडऐcca iनda Aटटa EmAttuऱada tawira wEरa waZi illऐnu muडiwஉ paणणuna aनda नari,Aडஉ kiटटa solluccu iङga नiरऐya नalla taणणi irஉkkஉ romba kuളircciyawum irஉkkஉ adAn romba saऩdOषattula kattiडEnnu solluccu.
4
+ shuffled_train_tamil_female_mono_06008 $rAmAnujar orஉ ऩAളEnum awarउkkuttam illattil amudஉ seywikka wirumbinAr.
5
+ shuffled_train_tamil_female_mono_06009 $नamacciwAyam enरஉ solwOmE enरa agattiyar नAradar टUyaटटऐ eZudinAr.
6
+ shuffled_train_tamil_female_mono_06010 $wEda riषikku yadArtta ulagऐp pUraणamAga wiटटஉc sellum tऐriyamillऐ enbadஉ idiliruनdஉ terigiरadஉ.
7
+ shuffled_train_tamil_female_mono_06011 $idஉkkஉ aनda siङgamum sammadiccadu uडanE reणडஉ mirugaङgaളum taङgaളOडa wAl reणडऐyum pakkattula iruनda koडiya payanbaडutti kaटटikkiटटஉ kugऐ pakkam pOccuङga.
8
+ shuffled_train_tamil_female_mono_06012 $pEषanङgiऱa pErla kannAbinnAnnu kAsटyUm pOडuऱa AളuङgannA enakkஉ alarji.
9
+ shuffled_train_tamil_female_mono_06013 $EdEdO kOषam seygiरArgaള sol sol makkaള enna kOषamiडugiरArgaള.
10
+ shuffled_train_tamil_female_mono_06014 $mudalAളiyum awar uttarawum ऩErttiyAnawऐ enरum,uङgaളஉkkuttAn adil sila asऔgaryam enरum solluङgaള.
11
+ shuffled_train_tamil_female_mono_06015 $ऩabuञjagan wiളakkam ऩabuञjagan enरa sol Aणdanmऐyinmऐ,ali enरa poruളgaളऐ uडऐyadஉ enरஉ agarAdigaള kUरuginरana.
12
+ shuffled_train_tamil_female_mono_06016 $paZuwUrkkArargaള pOटटa kOटटऐ awar tAणडuwadillऐ enरஉ janaङgaള solgiरArgaള.
13
+ shuffled_train_tamil_female_mono_06017 $warttaga adigarippAl irஉ नAडugaളum ippOdஉ idऐ innum irஉ Aणडugaളil pilliyan टAlarAkka.
14
+ shuffled_train_tamil_female_mono_06018 $muttana pOttAളe,mudalragam,muraणऐ,mUkkuppoडi,mUटटऐk kaटटudal,mUटटऐt tiളuppudal,mUणAwadu ragam,mUणउkkumUणu ambAram,mElsilambu,moऩdऐppuडi.
15
+ shuffled_train_tamil_female_mono_06019 $puttarऐ maरutta tiruwaളളuwar pऔttargaളin mAmisak koളgऐyऐyum wanmऐyAgak kaणडittஉ maऱउkkinऱAr.
16
+ shuffled_train_tamil_female_mono_06020 $waruषA waruषam sennऐ puttagak kAटcila Owiyap pOटटi wऐppAङga.
17
+ shuffled_train_tamil_female_mono_06021 $malEषiya tIbagaरpa pagudiyil uളളa makkaള togऐyAnadu sarAwak maटरum sabA pagudiyऐ wiडa maडaङgஉ adigamAgum.
18
+ shuffled_train_tamil_female_mono_06022 $manidargaള,sambawaङgaള,piraccinऐgaള paटरiya tam AZamAna wiruppu weऱuppuk karuttugaളऐ weളiyiडuwar.
19
+ shuffled_train_tamil_female_mono_06023 $iftஉ orஉ kAraणam paटरi iरऐwanऐ uyirऐ Eटरஉk koളളudal sahEdउkkirubऐ Agum.
20
+ shuffled_train_tamil_female_mono_06024 $idil skuwAष kOrटkaള,paളudUkkum payiरsi mऐyam maटरum piटऩes mऐyaङgaള warum.
21
+ shuffled_train_tamil_female_mono_06025 $iनda sarwIs Arambiccu koञja ऩAളlEyE नalla risalट,podஉ makkaളellAm romba sEbA fIl paणरAङga.
22
+ shuffled_train_tamil_female_mono_06026 $orஉ pattஉ warusattउkku munnAlE नAn iनda wElऐdAn seydஉ koणडiruनdEn.
23
+ shuffled_train_tamil_female_mono_06027 $idஉ kuरittஉ ञAnam kOटटaguppam pOlIs नilऐyattil pugAr koडuttAr.
24
+ shuffled_train_tamil_female_mono_06028 $rahmAn,hArIsnu pirabala isऐyamऐp pAളargaളin hiट ऩambarsla sila pAडalgaളऐ eङga sटऐlla mAttiटटu,wEरa टEsटla koडuppOm.
25
+ shuffled_train_tamil_female_mono_06029 $kiruषणऐ tuङgabattirऐ नadikkarऐ OramAy awargaളuडऐya rAjyamum peruhi waनdadஉ.
26
+ shuffled_train_tamil_female_mono_06030 $paणडi manOgar kulgarniyiडam iनdiyAwil iऩdustAni saङgIdam payinरAr srudi.
27
+ shuffled_train_tamil_female_mono_06031 $waटटiyum mudalum sAlऐyil warumbOdஉ miषkinin pudஉp paडam OऩAyum Aटटஉkkuटटiyum pOsटar pArttEn.
28
+ shuffled_train_tamil_female_mono_06032 $AnAl,sIsaganऐk kaणडஉ payaनdஉ wilagi,awargaള ऩaटpAgawE iruनdArgaള.
29
+ shuffled_train_tamil_female_mono_06033 $oruनAള akpar नagar waളam waनdArஉ,appa aङga iruनda marattஉ mEla ellAm Eरi ukkAनdஉ iruनdAङga.
30
+ shuffled_train_tamil_female_mono_06034 $awankiटटa sErऩdirउkkiऱa seydi irஉkkஉ pAruङga aनda aളawஉkkஉ tAn siनdikka muडiyum eZuda muडiyum.
31
+ shuffled_train_tamil_female_mono_06035 $ennennawO uणarwugaളin mOdalil kadirEsanum orஉ kaणam ayarनdஉ pOnAn.
32
+ shuffled_train_tamil_female_mono_06036 $enna paडikkalAm,enरஉ taङgऐ weളiनAटटil irஉkkum aणणanஉkkஉ minnaञjal anuppa udawum menboruള idஉ.
33
+ shuffled_train_tamil_female_mono_06037 $uडanE tannOडa magana aनda kaZuda mEla Etti wiटटuटटஉ kUडa sErनdஉ नaडaनdAr wiwasAyi.
34
+ shuffled_train_tamil_female_mono_06038 $AणडAളin iniya pAsurattऐk kEटटஉ wiटटAyAnAl appuरam wAളऐyum wElऐyum wiटटeरiनdஉ wiटटஉ ennऐp pOl नIyum kaणणan mEl kAdal koणडஉ wiणणagara yAttirऐ kiളambi wiडuwAy unakkஉ AणडAള pAsuraङgaള teriyumA pAडuwAyA.
35
+ shuffled_train_tamil_female_mono_06039 $ediril orஉ peruञjuwar teriनdadஉ.
36
+ shuffled_train_tamil_female_mono_06040 $ellArum iरaङgiyAnadum appeणgaളil orutti ऐyO mudalऐ enरஉ kUwinAള.
37
+ shuffled_train_tamil_female_mono_06041 $adu eZuttஉ,sol,poruള,yAppஉ,saऩdam,waZakkiyal,arasiyal,amऐcciyal,kaणiyam,kAऩdaruwam,kUttiyal paटरik kUरuginरadஉ.
38
+ shuffled_train_tamil_female_mono_06042 $adஉ kuളirkkalAmgiऱadAla romba kuളir aडiccadu.
39
+ shuffled_train_tamil_female_mono_06043 $insटiडiyUटटil kannaडap piriwil नaडippஉp payiरsiyil sErनdஉ paडikkiऱagAr siwAjirAw.
40
+ shuffled_train_tamil_female_mono_06044 $षiwa paktiyil Iडubaटटஉp perumbAlum yOgattilum tiyAnattilum pUjऐyilum kAlamgaZippadAga kEളwi.
41
+ shuffled_train_tamil_female_mono_06045 $wAyppஉ kiडऐkkum pOdellAm malऐyAളa iyakkuनargaള iduguरittஉ pEsi waruwadum sattiyam.
42
+ shuffled_train_tamil_female_mono_06046 $AnAl sila miट kEp maटरum smAl kEp paङgugaള नiccayam aनda aളawஉ waളarcciyऐk kAटटakkUडum.
43
+ shuffled_train_tamil_female_mono_06047 $appOdஉ aननAडஉ puraनda mannan,enऩAडu wiളऐwinऱip pOnamऐyAl yAn uङgaളऐk kAppAटरa iyalAdஉ.
44
+ shuffled_train_tamil_female_mono_06048 $mudariऩda sटirAberip paZaङgaളin in maणam Aटरஉkkஉ mElE kamaZऩdadu.
45
+ shuffled_train_tamil_female_mono_06049 $appuरam mammAwउkku annऐkkஉ AnadE,adஉ mAdiri skin pUrA taडittaडiyA taडittu yஉ wil lஉk ali.
46
+ shuffled_train_tamil_female_mono_06050 $wAZwin magiZcci ellऐyऐk kaणडubiडittஉ wiडuwadஉ enरஉ weऱiyuडan iruनdAr.
47
+ shuffled_train_tamil_female_mono_06051 $orஉ weളiनAटटஉ taniyAr maruनdஉ kambenikkஉ aनda fArmulAwऐ नAn ऐmbadஉ laटcattउkku wilऐ pEsiटटEn saनdiyA.
48
+ shuffled_train_tamil_female_mono_06052 $appa kuZaनdऐ paruwattula iruनda pIrpAl orஉ kalla eडuttஉ awarumEla wIsuऱamAdiri ऩaडiccAn.
49
+ shuffled_train_tamil_female_mono_06053 $uञjऐ ennum ujjani नAटटilum kumAra saणmuga kArttigEya waZibAडஉ akkAlattil siरappAga नaडaनdஉ waनdadஉ.
50
+ shuffled_train_tamil_female_mono_06054 $koञja tUram pOnadஉkkஉ appuरam,orஉ pAटटi waनdஉ नI ellAm periya manusanA,sinna pऐyana नaडakka wiटटuडu नI ukkAनduटटஉ wariyEnu sonnAङga.
51
+ shuffled_train_tamil_female_mono_06055 $aनdac samayam awwisAlamAna wIdi jana sUniyamAgak kAणappaटटadஉ Or,kAkkऐ aङgE kiडऐyAdஉ.
52
+ shuffled_train_tamil_female_mono_06056 $kaZaटरik kaZittabin aणiyoणAs silambஉ enin wAളA kaणणagi silambஉ enak kUऱalamऐyum.
53
+ shuffled_train_tamil_female_mono_06057 $unnadamAna marattin uccAणi kiളऐ onरil AZwArkkaडiyAn amarनdiruppadஉ teriनdadஉ OgO wIrawऐषणawarE नAn ummuडऐya tirஉkkaणgaളஉkkஉ awwaളawஉ iniyawanAga irஉkkiरEnA नAn ummऐ siरidஉ नanरAga pArkkiरEn.
54
+ shuffled_train_tamil_female_mono_06058 $idஉkkஉ iणडarwyU oणणum tawarऱadilla summA pOy waनdஉ pAडudAn miccam.
55
+ shuffled_train_tamil_female_mono_06059 $iनdap pOटटi tamiZaga ses wIrar Anaऩdiऱkum wIrAङganऐgaള.
56
+ shuffled_train_tamil_female_mono_06060 $wasaऩdi paडiyarugE tayArAga kAttiruनda lisiyऐt tUkkik koणडஉ ऩagarऩdAn jOषwA.
57
+ shuffled_train_tamil_female_mono_06061 $iनda waruषam Askarin wiषuwal ebekटasउkkup pOटटi pOटटa ऐनdஉ paडaङgaളil iraणडஉ paडaङgaള नAङgaള kirAfiks paणणadudAn.
58
+ shuffled_train_tamil_female_mono_06062 $edirwIटटuk kUrऐgaള,kArgaള नiरkum weटटa weളi wIटटஉ wAsal,नaडऐbAdऐ eङgum uरऐनdஉ kiडakkiरadஉ.
59
+ shuffled_train_tamil_female_mono_06063 $awaटरil नAlஉ pOnkaളऐ eडஉkkac solli wiटटஉ iraणडஉ maटटumE pOdum enऱiruऩdAr.
60
+ shuffled_train_tamil_female_mono_06064 $adऐ awanஉkkஉ नIङgaള koडuttIrgaളA illawE illऐ appaडiyellAm EmAनduwiडa नAn unnऐp pOl EmAളiyA awaniडam muttirऐ mOdiram iruनdadஉ uणmऐ.
tamil_latest/male/model/config.yaml ADDED
@@ -0,0 +1,315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accum_grad: 8
2
+ adapter: lora
3
+ adapter_conf: {}
4
+ allow_multi_rates: false
5
+ allow_variable_data_keys: false
6
+ batch_bins: 3000000
7
+ batch_size: 20
8
+ batch_type: numel
9
+ best_model_criterion:
10
+ - - valid
11
+ - loss
12
+ - min
13
+ - - train
14
+ - loss
15
+ - min
16
+ bpemodel: null
17
+ chunk_default_fs: null
18
+ chunk_discard_short_samples: true
19
+ chunk_excluded_key_prefixes: []
20
+ chunk_length: 500
21
+ chunk_max_abs_length: null
22
+ chunk_shift_ratio: 0.5
23
+ cleaner: null
24
+ collect_stats: false
25
+ config: conf/tuning/train_fastspeech2.yaml
26
+ create_graph_in_tensorboard: false
27
+ cudnn_benchmark: false
28
+ cudnn_deterministic: true
29
+ cudnn_enabled: true
30
+ detect_anomaly: false
31
+ dist_backend: nccl
32
+ dist_init_method: env://
33
+ dist_launcher: null
34
+ dist_master_addr: localhost
35
+ dist_master_port: 52247
36
+ dist_rank: 0
37
+ dist_world_size: 2
38
+ distributed: true
39
+ drop_last_iter: false
40
+ dry_run: false
41
+ early_stopping_criterion:
42
+ - valid
43
+ - loss
44
+ - min
45
+ energy_extract: energy
46
+ energy_extract_conf:
47
+ fs: 48000
48
+ hop_length: 1024
49
+ n_fft: 8192
50
+ reduction_factor: 1
51
+ win_length: 4096
52
+ energy_normalize: global_mvn
53
+ energy_normalize_conf:
54
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/tamil_latest/male/model/energy_stats.npz
55
+ exclude_weight_decay: false
56
+ exclude_weight_decay_conf: {}
57
+ feats_extract: fbank
58
+ feats_extract_conf:
59
+ fmax: 24000
60
+ fmin: 0
61
+ fs: 48000
62
+ hop_length: 1024
63
+ n_fft: 8192
64
+ n_mels: 160
65
+ win_length: 4096
66
+ fold_length:
67
+ - 150
68
+ - 819200
69
+ freeze_param: []
70
+ g2p: null
71
+ grad_clip: 1.0
72
+ grad_clip_type: 2.0
73
+ grad_noise: false
74
+ ignore_init_mismatch: false
75
+ init_param: []
76
+ iterator_type: sequence
77
+ keep_nbest_models: 5
78
+ local_rank: 0
79
+ log_interval: null
80
+ log_level: INFO
81
+ max_cache_fd: 32
82
+ max_cache_size: 0.0
83
+ max_epoch: 1000
84
+ model_conf: {}
85
+ multiple_iterator: false
86
+ multiprocessing_distributed: true
87
+ nbest_averaging_interval: 0
88
+ ngpu: 1
89
+ no_forward_run: false
90
+ non_linguistic_symbols: null
91
+ normalize: global_mvn
92
+ normalize_conf:
93
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/tamil_latest/male/model/feats_stats.npz
94
+ num_att_plot: 3
95
+ num_cache_chunks: 1024
96
+ num_iters_per_epoch: 800
97
+ num_workers: 1
98
+ odim: null
99
+ optim: adam
100
+ optim_conf:
101
+ lr: 1.0
102
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
103
+ patience: null
104
+ pitch_extract: dio
105
+ pitch_extract_conf:
106
+ f0max: 350
107
+ f0min: 40
108
+ fs: 48000
109
+ hop_length: 1024
110
+ n_fft: 8192
111
+ reduction_factor: 1
112
+ pitch_normalize: global_mvn
113
+ pitch_normalize_conf:
114
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/tamil_latest/male/model/pitch_stats.npz
115
+ pretrain_path: null
116
+ print_config: false
117
+ required:
118
+ - output_dir
119
+ - token_list
120
+ resume: true
121
+ save_strategy: all
122
+ scheduler: noamlr
123
+ scheduler_conf:
124
+ model_size: 384
125
+ warmup_steps: 4000
126
+ seed: 0
127
+ sharded_ddp: false
128
+ shuffle_within_batch: false
129
+ sort_batch: descending
130
+ sort_in_batch: descending
131
+ token_list:
132
+ - <blank>
133
+ - <unk>
134
+ - <space>
135
+ - $
136
+ - ','
137
+ - .
138
+ - M
139
+ - q
140
+ - H
141
+ - k
142
+ - "\u0916"
143
+ - g
144
+ - "\u0918"
145
+ - "\u0919"
146
+ - c
147
+ - C
148
+ - j
149
+ - J
150
+ - "\u091E"
151
+ - "\u091F"
152
+ - "\u0920"
153
+ - "\u0921"
154
+ - "\u0922"
155
+ - "\u0923"
156
+ - t
157
+ - "\u0925"
158
+ - d
159
+ - "\u0927"
160
+ - n
161
+ - "\u0929"
162
+ - p
163
+ - P
164
+ - b
165
+ - B
166
+ - m
167
+ - y
168
+ - r
169
+ - "\u0931"
170
+ - l
171
+ - "\u0D33"
172
+ - Z
173
+ - w
174
+ - "\u0936"
175
+ - "\u0937"
176
+ - s
177
+ - h
178
+ - Y
179
+ - "\u093D"
180
+ - "\u0915"
181
+ - K
182
+ - G
183
+ - z
184
+ - D
185
+ - T
186
+ - f
187
+ - "\u0960"
188
+ - "\u0D7A"
189
+ - N
190
+ - "\u0D7C"
191
+ - "\u0D7D"
192
+ - "\u0D7E"
193
+ - a
194
+ - A
195
+ - i
196
+ - I
197
+ - u
198
+ - U
199
+ - R
200
+ - "\u090D"
201
+ - e
202
+ - E
203
+ - "\u0910"
204
+ - "\u0911"
205
+ - o
206
+ - O
207
+ - "\u0914"
208
+ - "\u0B89"
209
+ - <sos/eos>
210
+ token_type: char
211
+ train_data_path_and_name_and_type:
212
+ - - dump/raw/tr_no_dev/text
213
+ - text
214
+ - text
215
+ - - duration_info/tr_no_dev/durations
216
+ - durations
217
+ - text_int
218
+ - - dump/raw/tr_no_dev/wav.scp
219
+ - speech
220
+ - sound
221
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
222
+ - pitch
223
+ - npy
224
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
225
+ - energy
226
+ - npy
227
+ train_dtype: float32
228
+ train_shape_file:
229
+ - exp/tts_stats_raw_char_None/train/text_shape.char
230
+ - exp/tts_stats_raw_char_None/train/speech_shape
231
+ tts: fastspeech2
232
+ tts_conf:
233
+ adim: 384
234
+ aheads: 2
235
+ decoder_normalize_before: true
236
+ dlayers: 4
237
+ dunits: 1536
238
+ duration_predictor_chans: 256
239
+ duration_predictor_kernel_size: 3
240
+ duration_predictor_layers: 2
241
+ elayers: 4
242
+ encoder_normalize_before: true
243
+ energy_embed_dropout: 0.0
244
+ energy_embed_kernel_size: 1
245
+ energy_predictor_chans: 256
246
+ energy_predictor_dropout: 0.5
247
+ energy_predictor_kernel_size: 3
248
+ energy_predictor_layers: 2
249
+ eunits: 1536
250
+ init_dec_alpha: 1.0
251
+ init_enc_alpha: 1.0
252
+ init_type: xavier_uniform
253
+ pitch_embed_dropout: 0.0
254
+ pitch_embed_kernel_size: 1
255
+ pitch_predictor_chans: 256
256
+ pitch_predictor_dropout: 0.5
257
+ pitch_predictor_kernel_size: 5
258
+ pitch_predictor_layers: 5
259
+ positionwise_conv_kernel_size: 3
260
+ positionwise_layer_type: conv1d
261
+ postnet_chans: 256
262
+ postnet_filts: 5
263
+ postnet_layers: 5
264
+ reduction_factor: 1
265
+ stop_gradient_from_energy_predictor: false
266
+ stop_gradient_from_pitch_predictor: true
267
+ transformer_dec_attn_dropout_rate: 0.2
268
+ transformer_dec_dropout_rate: 0.2
269
+ transformer_dec_positional_dropout_rate: 0.2
270
+ transformer_enc_attn_dropout_rate: 0.2
271
+ transformer_enc_dropout_rate: 0.2
272
+ transformer_enc_positional_dropout_rate: 0.2
273
+ use_masking: true
274
+ use_scaled_pos_enc: true
275
+ unused_parameters: false
276
+ use_adapter: false
277
+ use_amp: false
278
+ use_matplotlib: true
279
+ use_preprocessor: true
280
+ use_tensorboard: true
281
+ use_wandb: false
282
+ val_scheduler_criterion:
283
+ - valid
284
+ - loss
285
+ valid_batch_bins: null
286
+ valid_batch_size: null
287
+ valid_batch_type: null
288
+ valid_data_path_and_name_and_type:
289
+ - - dump/raw/dev/text
290
+ - text
291
+ - text
292
+ - - duration_info/dev/durations
293
+ - durations
294
+ - text_int
295
+ - - dump/raw/dev/wav.scp
296
+ - speech
297
+ - sound
298
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
299
+ - pitch
300
+ - npy
301
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
302
+ - energy
303
+ - npy
304
+ valid_iterator_type: null
305
+ valid_max_cache_size: null
306
+ valid_shape_file:
307
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
308
+ - exp/tts_stats_raw_char_None/valid/speech_shape
309
+ version: '202402'
310
+ wandb_entity: null
311
+ wandb_id: null
312
+ wandb_model_log_interval: -1
313
+ wandb_name: null
314
+ wandb_project: null
315
+ write_collected_feats: false
tamil_latest/male/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d34b352ef0ac0a0d56ac613e9cbc8edcbd0a2ff6fca02a1ba8e7e8fdd8d44d59
3
+ size 770
tamil_latest/male/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c530bcc3478562c3c84697b460664054757793faf111bbf0bace96af8d772da5
3
+ size 2042
tamil_latest/male/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d8a4d64a04bf2d5b19644460240f6634eeb2056d2dd7e5796f475ae587b246c
3
+ size 152128410
tamil_latest/male/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b16e095de0bdcec6f650e0c860012f893bcc4a8ecd4c47bb4255883f09338f7
3
+ size 770
tamil_latest/male/model/text ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ shuffled_train_tamil_male_mono_06543 $motta tEsattऐyum iणऐkka wEणडum enरஉ miguनda yOsanऐyuडan aनdak kAriyaङgaളऐc seydiruनdAr.
2
+ shuffled_train_tamil_male_mono_06544 $uडanE ellA maनdirigaളum wऐraङgaളऐ piरakka ArambiccAङga.
3
+ shuffled_train_tamil_male_mono_06545 $adumAdiriya orஉ wArattஉkkஉ appuऱamA aनda pऐyana aङga kUटटiटटஉ pOnAङga.
4
+ shuffled_train_tamil_male_mono_06546 $ten Aprikka,spAniष moZigaളilum iनda kArटटUn kadऐ weളiwara Arambittadஉ.
5
+ shuffled_train_tamil_male_mono_06547 $adஉkkஉ kuऱऐञju kEkkAdE,kEs pOटटஉ pArtturalAm wakkIla sAr,kEs illऐppA.
6
+ shuffled_train_tamil_male_mono_06548 $eppaडik kAy ऩagarttinAl Aङgila kAlaniyऐ uruwAkka muडiyum enरa wittऐ teriनdawar rAbiളs.
7
+ shuffled_train_tamil_male_mono_06549 $waralAरஉ enbadan aडippaडऐyil tAn manidanउkkuള waनdஉ sErनduളളa kalऐgaള,meyyiyal enरஉ toडaङgip palawaटऱऐyum नAm pEsamuडiyum.
8
+ shuffled_train_tamil_male_mono_06550 $oruwar waङgiyil kaडan wAङgi suyadoZil seywadஉ periya wiषayamalla.
9
+ shuffled_train_tamil_male_mono_06551 $Or abalऐt tAyin kaडऐsi Asऐ enरa kadऐ soनdak kaणawanAl waञjikkappaटटa Or abalऐyin kadऐ.
10
+ shuffled_train_tamil_male_mono_06552 $sila waruडaङgaള senरabin eङgaളஉkkஉ aनdap pEyp payam koञjam pOyiruनdadஉ.
11
+ shuffled_train_tamil_male_mono_06553 $minsArattऐc sEmikka wEणडum enbadऐ palwEरஉ wiളambaraङgaള mUlam नAളdOरum makkaളஉkkஉ arasஉ aरiwuरuttugiरadஉ.
12
+ shuffled_train_tamil_male_mono_06554 $lAl puषp नamakkuള irஉkkum pOrAटटattin,uऩdudalin uणmऐyAna weളippAडudAn ilakkiyamAy malara muडiyum.
13
+ shuffled_train_tamil_male_mono_06555 $AnAl ulagap pirasitti peटरa wisuwaनAdar kOyilAl wisuwaनAdar kaटटaडam enरஉ adaरkஉp peyarillऐ.
14
+ shuffled_train_tamil_male_mono_06556 $uडanE tannOडa maganऐyum kUडa Ettikkiटटu oणणA payaणam seञjAru awarஉ.
15
+ shuffled_train_tamil_male_mono_06557 $appadAn नamakkஉ orஉ pirasnऐ waनdA,adऐ orஉ mUणAm manuषan kiटटa wiडaणumnu purञjउkkiटटEn.
16
+ shuffled_train_tamil_male_mono_06558 $kallऐ Urஉkkuള sellum pOdஉ awwUrऐs sErनdawargaള eङgaള Uril kudirऐ mEl Eरic sellak kUडAdஉ.
17
+ shuffled_train_tamil_male_mono_06559 $awan awaളiडam नeruङgi नinरஉ mUnरஉ नAnkஉ नimiषam pEsinAn.
18
+ shuffled_train_tamil_male_mono_06560 $uणawஉ नiरuttum kAlam नAnkAm नAളilO ऐनdAm नAളilO iwऐ uणawஉ uटkoളളuwadऐ नiरutti wiडuginरana.
19
+ shuffled_train_tamil_male_mono_06561 $Eडugaളil eZudiya eणणaटरa kaटटurऐgaളum,mAऩAडugaളil muZaङgiya Aywurऐgaളum awwऐ awargaളin wiZumiya pulamऐ नalattऐp paऱऐsAटऱina.
20
+ shuffled_train_tamil_male_mono_06562 $mOडiyin timir awarऐ tOरkaडikkum enkiरAr mAनila kAङgiras podஉc seyalAളar tIbak pabAriyA.
21
+ shuffled_train_tamil_male_mono_06563 $mEरkஉ mugamAy amऐनda murugan Alayattiऱkuള नuZऐनdadum,mAडattup piളളऐyArऐ waZibaटटஉ,murugap perumAnin aZagiya tirउkkOlattऐk kAणalAm.
22
+ shuffled_train_tamil_male_mono_06564 $waडakkE kanyAgupjattil harषawarttanar waruषaनdOरum नaडattum uरsawattऐp paटरit teriyumallawA नarasimmA.
23
+ shuffled_train_tamil_male_mono_06565 $sugaनdi tEwagiya waccA,enakkஉ uङgaളऐp patti onnum teriyAdE,नAङga solरOm.
24
+ shuffled_train_tamil_male_mono_06566 $sammar Ap enरa iनdap paडattil iwarஉkkஉ jOडiyAga नaडippadஉ sigaऩdar kEr.
25
+ shuffled_train_tamil_male_mono_06567 $नAn pAडal eZudiya mudal paडam sऔbAkkiyadi enरAr paZaऩiccAmi.
26
+ shuffled_train_tamil_male_mono_06568 $wO,amerikkAwO,ऐrOppiya yUniyanO talऐyiटटAl awargaളऐyum samAളikka wEणडum.
27
+ shuffled_train_tamil_male_mono_06569 $kiषan saऩdarin moZi peyarppuk kadऐyऐ uളളaडakkiya नAn sAga mAटटEn enbadum kuरippiडat taguनda orஉ नUl Agum.
28
+ shuffled_train_tamil_male_mono_06570 $teriyumA teriyawillऐ ammaणi enakkஉ maनdiramum teriyAdஉ jOsiyamum teriyAdஉ unnऐ नEटरஉk kAlऐyil pArttadஉ mudal unnuडऐya ञAbagamAgawE iruनdadஉ.
29
+ shuffled_train_tamil_male_mono_06571 $orumuरऐ iनda waZiyil wEऱeAru paडattiरkAga kIZE pAyऩdabEAdu sटiyariङ sakkaram uडऐनdஉ enadஉ kAlgaളऐp padam pArttஉ wiटटadஉ.
30
+ shuffled_train_tamil_male_mono_06572 $appOdஉ siwagAmiyum aनda iडattஉkkஉ waनdஉ sErawE putta pikषஉ awaളऐ Eरiटटஉp pArttuwiटटஉ siwagAmiyiडam नIङgaള ellA wiwaraङgaളऐyum solla wEणडum.
31
+ shuffled_train_tamil_male_mono_06573 $awwAरE,enbadum munsuZiyAm AydattOडu lagaram sErऩdiणऐऩda karaनda eZuttஉ enbar apperiyAr lagarattin urasoli adu enbadAyiऱ.
32
+ shuffled_train_tamil_male_mono_06574 $iwwAरஉ hiटlarum jerman paडऐgaളum sOwiyat turuppugaളAl muटऱugऐyiडap paटटஉ tappikkum ubAyam inरi maणigaള kaडaनdana.
33
+ shuffled_train_tamil_male_mono_06575 $orஉ नAള karuppusAmi tAnam koडuttउkkiटटu irஉkkuरa iडattஉkkஉ waनdஉ,unnOडa ऩilatta uङga appA eppawO eङga appA kiटटa wittuटटAru,adउkkAna sAटci iनda pऐla irउkkunu sonnAn.
34
+ shuffled_train_tamil_male_mono_06576 $jeyicciटटu waनdA innum नiरऐya sAdikkaणumnu adஉkkAga eppaडi uZऐkkaणumnu kattuk kuडuppOm enरAr.
35
+ shuffled_train_tamil_male_mono_06577 $atri,pirugu,kutsar,wasटटar,kऔdamar,tAsyabar,Aङgirasar enbawargaളE ammuniwargaള.
36
+ shuffled_train_tamil_male_mono_06578 $tamiZ enbadan pinnum iwwAरஉ agaras sAriyऐ peटऱum peऱAdum,walloटऱu mikkum warum.
37
+ shuffled_train_tamil_male_mono_06579 $टijiटटal wIडiyO ऩAडAkkaള,pOnरadஉ.
38
+ shuffled_train_tamil_male_mono_06580 $नaनdini kambIra pAwattuडan waनdiyattEwanऐp pArttஉ ऐyA नIr iनdak kOटटऐyiliruनdஉ uyirOडஉ tappic sella eणणugiरIr allawA enरஉ kEटटAള.
39
+ shuffled_train_tamil_male_mono_06581 $appaडi sariyA madippiडuऱawaङgaളउkku miga periya parisஉ kAttirउkku,नAളऐkkஉ araणmanऐkkஉ waनdஉ wAङgikkOङgannu sonnArஉ.
40
+ shuffled_train_tamil_male_mono_06582 $iruള mAളigऐ kAणAmaरpOna waनdiyattEwan enna AnAn enbadऐ ippOdஉ नAm kawanikkalAm.
41
+ shuffled_train_tamil_male_mono_06583 $uram piरaनda kadऐ teriyumA uङgaളஉkkஉ,teriyAdஉ enbadubOl mUwarumE iषimurAwऐp pArttArgaള.
42
+ shuffled_train_tamil_male_mono_06584 $idஉ kAlऐyil नAlarऐ maणikkஉ wAnattil tOnऱuwadAl idanऐ wiडiweളളi ena aZऐkkinऱArgaള.
43
+ shuffled_train_tamil_male_mono_06585 $idऐppOlawE kuZaनdऐyin AZनda tanmऐyAnadஉ EZஉ muऱऐyanऱu,AnAl EZஉ eZubadஉ muरऐ mElE tOnरugiरadஉ.
44
+ shuffled_train_tamil_male_mono_06586 $adஉkkஉ appuऱamA नari kiटटa solluccu,ऩariyArE नamakkஉ sulabamA innऐkkஉ iरऐ kiडऐccuडuccu,AnA enakkஉ innum sOmbEऱittanamA irஉkkஉ adanAla नAn koञjam Oywஉ eडuttuटटஉ warरEn.
45
+ shuffled_train_tamil_male_mono_06587 $sigiccऐyil kuरऐyillऐ enरஉ tIrppஉk kUरiya mAनilak kamiषan,srIwillipputtUr नugarwOr नIdimanरam aളitta uttarawऐyum rattஉ seydadஉ.
46
+ shuffled_train_tamil_male_mono_06588 $टEsaणड,pEsaट hawuणड,pulडAk नAygaളin kAdugaളil muडi irஉkkAdஉ.
47
+ shuffled_train_tamil_male_mono_06589 $AdalAl tAn नAm iwwaളawஉ kAlamEnum taङgi wAZनdஉ waनdirஉkkiरOm enbadஉ awwiञञAniyin muडiwஉ.
48
+ shuffled_train_tamil_male_mono_06590 $adஉ romba mukkiyam sAr,ippO orஉ hIrO suwattउkkiटटa sErla uटkArऩdirउkkAr.
49
+ shuffled_train_tamil_male_mono_06591 $kOडi rUbAy aളawஉkkஉ नidi mOsaडi seydஉ,ऐ.
50
+ shuffled_train_tamil_male_mono_06592 $appIjE surEऩdirA kurUp enkiरa नiरuwanattin motta wiyAbArattn madippஉ sumAr Ayirattil ऐनUरஉ kOडikkum mEl.
51
+ shuffled_train_tamil_male_mono_06593 $aनda aളawaटरa Anaनdattil iളawarasi tanakkஉ udawi puriनda sarpparAjanऐ aरawE maरaनdஉ wiटटAള.
52
+ shuffled_train_tamil_male_mono_06594 $aZagiya mayil,orஉ kAटटula orஉ mayil wAZऩdugiटटu waनduccஉ.
53
+ shuffled_train_tamil_male_mono_06595 $AnwIksigI enak kऔडilyar tam poruള नUlil idanऐk kuऱippAr.
54
+ shuffled_train_tamil_male_mono_06596 $enरum,enरum,enरum sollappaडum mEरkattiya soरkaളऐ नagar,enरa tamiZc sollin tiribugaളE.
55
+ shuffled_train_tamil_male_mono_06597 $ippaडippaटटa sUZनilऐyil ayOtti नagaril koളളऐyargaള aटटagAsam adigarittadஉ.
56
+ shuffled_train_tamil_male_mono_06598 $kawiञargaളin kadiyAna karuणऐ waളളal wAZga paणडida watsalarAgiya parAनdaga sakkarawarttiyin tirஉp pErar नIडUZi wAZga enरஉ wAZttinArgaള.
57
+ shuffled_train_tamil_male_mono_06599 $mUrttiyuडan sErनdஉ pOga muडiyAdஉ enbadAlEyE pas payaणattऐt tawirttஉ,नaडaनdE waruwAn rawi.
58
+ shuffled_train_tamil_male_mono_06600 $AnAl mUwarum maनdabutti uडऐyawargaളAga iruनdanar.
59
+ shuffled_train_tamil_male_mono_06601 $jIragam,kaडugஉ,tuwaram paruppஉ,maञjaള poडi,uppஉ,wiളakkeणणey,ellAwaटरऐyum onरAgak kalakkawum.
60
+ shuffled_train_tamil_male_mono_06602 $orஉ wAram warऐ Arpiyas karऐyil amarनdஉ tEmbikkoणडu,uणawinरi maणणAlum puZudiyAlum maऱऐkkappaटटuk kiडaनdAn.
61
+ shuffled_train_tamil_male_mono_06603 $pakkattil waनdஉ,नAn koञjam Asiramattup pakkam pOyiटटஉ warEn.
62
+ shuffled_train_tamil_male_mono_06604 $en poऱumऐkkum ellऐyuणडu rajani,adऐ maऱakkAdE,enरAn.
63
+ shuffled_train_tamil_male_mono_06605 $kalgi pattirigऐ toडaङgiya pudidil sila नaणbargaള toडargadऐ eZudumbaडi kEटटArgaള.
64
+ shuffled_train_tamil_male_mono_06606 $aञjAdE paडattऐ aञjAmal iyakkiya टऐrakटar miषkin aडuttஉ orஉ Akषan paडattऐ iyakkawuളളAr.
65
+ shuffled_train_tamil_male_mono_06607 $anbuളളa टAkटar mAmAwஉkkஉ,edऐyum wiञञAna rIdiyAga ArAyum mElनAटटஉ sAstiraङgaളऐ yellAm karऐ kaणडawar नIङgaള.
66
+ shuffled_train_tamil_male_mono_06608 $awan tinamum kAटटஉkkஉ pakkattula irஉkkuरa malऐkkஉ Aडஉ mEykka pOwAn.
telugu_latest/female/model/config.yaml ADDED
@@ -0,0 +1,315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accum_grad: 8
2
+ adapter: lora
3
+ adapter_conf: {}
4
+ allow_multi_rates: false
5
+ allow_variable_data_keys: false
6
+ batch_bins: 3000000
7
+ batch_size: 20
8
+ batch_type: numel
9
+ best_model_criterion:
10
+ - - valid
11
+ - loss
12
+ - min
13
+ - - train
14
+ - loss
15
+ - min
16
+ bpemodel: null
17
+ chunk_default_fs: null
18
+ chunk_discard_short_samples: true
19
+ chunk_excluded_key_prefixes: []
20
+ chunk_length: 500
21
+ chunk_max_abs_length: null
22
+ chunk_shift_ratio: 0.5
23
+ cleaner: null
24
+ collect_stats: false
25
+ config: conf/tuning/train_fastspeech2.yaml
26
+ create_graph_in_tensorboard: false
27
+ cudnn_benchmark: false
28
+ cudnn_deterministic: true
29
+ cudnn_enabled: true
30
+ detect_anomaly: false
31
+ dist_backend: nccl
32
+ dist_init_method: env://
33
+ dist_launcher: null
34
+ dist_master_addr: localhost
35
+ dist_master_port: 51589
36
+ dist_rank: 0
37
+ dist_world_size: 2
38
+ distributed: true
39
+ drop_last_iter: false
40
+ dry_run: false
41
+ early_stopping_criterion:
42
+ - valid
43
+ - loss
44
+ - min
45
+ energy_extract: energy
46
+ energy_extract_conf:
47
+ fs: 48000
48
+ hop_length: 1024
49
+ n_fft: 8192
50
+ reduction_factor: 1
51
+ win_length: 4096
52
+ energy_normalize: global_mvn
53
+ energy_normalize_conf:
54
+ stats_file: /home/speech/Fastspeech2_HS/telugu_latest/female/model/energy_stats.npz
55
+ exclude_weight_decay: false
56
+ exclude_weight_decay_conf: {}
57
+ feats_extract: fbank
58
+ feats_extract_conf:
59
+ fmax: 24000
60
+ fmin: 0
61
+ fs: 48000
62
+ hop_length: 1024
63
+ n_fft: 8192
64
+ n_mels: 160
65
+ win_length: 4096
66
+ fold_length:
67
+ - 150
68
+ - 819200
69
+ freeze_param: []
70
+ g2p: null
71
+ grad_clip: 1.0
72
+ grad_clip_type: 2.0
73
+ grad_noise: false
74
+ ignore_init_mismatch: false
75
+ init_param: []
76
+ iterator_type: sequence
77
+ keep_nbest_models: 5
78
+ local_rank: 0
79
+ log_interval: null
80
+ log_level: INFO
81
+ max_cache_fd: 32
82
+ max_cache_size: 0.0
83
+ max_epoch: 1000
84
+ model_conf: {}
85
+ multiple_iterator: false
86
+ multiprocessing_distributed: true
87
+ nbest_averaging_interval: 0
88
+ ngpu: 1
89
+ no_forward_run: false
90
+ non_linguistic_symbols: null
91
+ normalize: global_mvn
92
+ normalize_conf:
93
+ stats_file: /home/speech/Fastspeech2_HS/telugu_latest/female/model/feats_stats.npz
94
+ num_att_plot: 3
95
+ num_cache_chunks: 1024
96
+ num_iters_per_epoch: 800
97
+ num_workers: 1
98
+ odim: null
99
+ optim: adam
100
+ optim_conf:
101
+ lr: 1.0
102
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
103
+ patience: null
104
+ pitch_extract: dio
105
+ pitch_extract_conf:
106
+ f0max: 400
107
+ f0min: 80
108
+ fs: 48000
109
+ hop_length: 1024
110
+ n_fft: 8192
111
+ reduction_factor: 1
112
+ pitch_normalize: global_mvn
113
+ pitch_normalize_conf:
114
+ stats_file: /home/speech/Fastspeech2_HS/telugu_latest/female/model/pitch_stats.npz
115
+ pretrain_path: null
116
+ print_config: false
117
+ required:
118
+ - output_dir
119
+ - token_list
120
+ resume: true
121
+ save_strategy: all
122
+ scheduler: noamlr
123
+ scheduler_conf:
124
+ model_size: 384
125
+ warmup_steps: 4000
126
+ seed: 0
127
+ sharded_ddp: false
128
+ shuffle_within_batch: false
129
+ sort_batch: descending
130
+ sort_in_batch: descending
131
+ token_list:
132
+ - <blank>
133
+ - <unk>
134
+ - <space>
135
+ - $
136
+ - ','
137
+ - .
138
+ - M
139
+ - q
140
+ - H
141
+ - k
142
+ - "\u0916"
143
+ - g
144
+ - "\u0918"
145
+ - "\u0919"
146
+ - c
147
+ - C
148
+ - j
149
+ - J
150
+ - "\u091E"
151
+ - "\u091F"
152
+ - "\u0920"
153
+ - "\u0921"
154
+ - "\u0922"
155
+ - "\u0923"
156
+ - t
157
+ - "\u0925"
158
+ - d
159
+ - "\u0927"
160
+ - n
161
+ - "\u0929"
162
+ - p
163
+ - P
164
+ - b
165
+ - B
166
+ - m
167
+ - y
168
+ - r
169
+ - "\u0931"
170
+ - l
171
+ - "\u0D33"
172
+ - Z
173
+ - w
174
+ - "\u0936"
175
+ - "\u0937"
176
+ - s
177
+ - h
178
+ - Y
179
+ - "\u093D"
180
+ - "\u0915"
181
+ - K
182
+ - G
183
+ - z
184
+ - D
185
+ - T
186
+ - f
187
+ - "\u0960"
188
+ - "\u0D7A"
189
+ - N
190
+ - "\u0D7C"
191
+ - "\u0D7D"
192
+ - "\u0D7E"
193
+ - a
194
+ - A
195
+ - i
196
+ - I
197
+ - u
198
+ - U
199
+ - R
200
+ - "\u090D"
201
+ - e
202
+ - E
203
+ - "\u0910"
204
+ - "\u0911"
205
+ - o
206
+ - O
207
+ - "\u0914"
208
+ - "\u0B89"
209
+ - <sos/eos>
210
+ token_type: char
211
+ train_data_path_and_name_and_type:
212
+ - - dump/raw/tr_no_dev/text
213
+ - text
214
+ - text
215
+ - - duration_info/tr_no_dev/durations
216
+ - durations
217
+ - text_int
218
+ - - dump/raw/tr_no_dev/wav.scp
219
+ - speech
220
+ - sound
221
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
222
+ - pitch
223
+ - npy
224
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
225
+ - energy
226
+ - npy
227
+ train_dtype: float32
228
+ train_shape_file:
229
+ - exp/tts_stats_raw_char_None/train/text_shape.char
230
+ - exp/tts_stats_raw_char_None/train/speech_shape
231
+ tts: fastspeech2
232
+ tts_conf:
233
+ adim: 384
234
+ aheads: 2
235
+ decoder_normalize_before: true
236
+ dlayers: 4
237
+ dunits: 1536
238
+ duration_predictor_chans: 256
239
+ duration_predictor_kernel_size: 3
240
+ duration_predictor_layers: 2
241
+ elayers: 4
242
+ encoder_normalize_before: true
243
+ energy_embed_dropout: 0.0
244
+ energy_embed_kernel_size: 1
245
+ energy_predictor_chans: 256
246
+ energy_predictor_dropout: 0.5
247
+ energy_predictor_kernel_size: 3
248
+ energy_predictor_layers: 2
249
+ eunits: 1536
250
+ init_dec_alpha: 1.0
251
+ init_enc_alpha: 1.0
252
+ init_type: xavier_uniform
253
+ pitch_embed_dropout: 0.0
254
+ pitch_embed_kernel_size: 1
255
+ pitch_predictor_chans: 256
256
+ pitch_predictor_dropout: 0.5
257
+ pitch_predictor_kernel_size: 5
258
+ pitch_predictor_layers: 5
259
+ positionwise_conv_kernel_size: 3
260
+ positionwise_layer_type: conv1d
261
+ postnet_chans: 256
262
+ postnet_filts: 5
263
+ postnet_layers: 5
264
+ reduction_factor: 1
265
+ stop_gradient_from_energy_predictor: false
266
+ stop_gradient_from_pitch_predictor: true
267
+ transformer_dec_attn_dropout_rate: 0.2
268
+ transformer_dec_dropout_rate: 0.2
269
+ transformer_dec_positional_dropout_rate: 0.2
270
+ transformer_enc_attn_dropout_rate: 0.2
271
+ transformer_enc_dropout_rate: 0.2
272
+ transformer_enc_positional_dropout_rate: 0.2
273
+ use_masking: true
274
+ use_scaled_pos_enc: true
275
+ unused_parameters: false
276
+ use_adapter: false
277
+ use_amp: false
278
+ use_matplotlib: true
279
+ use_preprocessor: true
280
+ use_tensorboard: true
281
+ use_wandb: false
282
+ val_scheduler_criterion:
283
+ - valid
284
+ - loss
285
+ valid_batch_bins: null
286
+ valid_batch_size: null
287
+ valid_batch_type: null
288
+ valid_data_path_and_name_and_type:
289
+ - - dump/raw/dev/text
290
+ - text
291
+ - text
292
+ - - duration_info/dev/durations
293
+ - durations
294
+ - text_int
295
+ - - dump/raw/dev/wav.scp
296
+ - speech
297
+ - sound
298
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
299
+ - pitch
300
+ - npy
301
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
302
+ - energy
303
+ - npy
304
+ valid_iterator_type: null
305
+ valid_max_cache_size: null
306
+ valid_shape_file:
307
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
308
+ - exp/tts_stats_raw_char_None/valid/speech_shape
309
+ version: '202402'
310
+ wandb_entity: null
311
+ wandb_id: null
312
+ wandb_model_log_interval: -1
313
+ wandb_name: null
314
+ wandb_project: null
315
+ write_collected_feats: false
telugu_latest/female/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e3a86ad6b1dc7850d96a201bb7c247a556d0e9a0d055cfdd2ef73c4b290254f
3
+ size 770
telugu_latest/female/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3275dec355cd1a2ac93c6e04752dc955032df1acaec77e48ec9a1d231d6bc1c2
3
+ size 2042
telugu_latest/female/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ad0bc1fe155c7985a02f1e012990b9448d86dea0e7f33f814df9c122cd70a0d
3
+ size 152128410
telugu_latest/female/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e09952bd06e003ae18181d70be820cf3dc00bc4d1d7c1113b062d352e667833e
3
+ size 770
telugu_latest/male/model/config.yaml ADDED
@@ -0,0 +1,315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accum_grad: 8
2
+ adapter: lora
3
+ adapter_conf: {}
4
+ allow_multi_rates: false
5
+ allow_variable_data_keys: false
6
+ batch_bins: 3000000
7
+ batch_size: 20
8
+ batch_type: numel
9
+ best_model_criterion:
10
+ - - valid
11
+ - loss
12
+ - min
13
+ - - train
14
+ - loss
15
+ - min
16
+ bpemodel: null
17
+ chunk_default_fs: null
18
+ chunk_discard_short_samples: true
19
+ chunk_excluded_key_prefixes: []
20
+ chunk_length: 500
21
+ chunk_max_abs_length: null
22
+ chunk_shift_ratio: 0.5
23
+ cleaner: null
24
+ collect_stats: false
25
+ config: conf/tuning/train_fastspeech2.yaml
26
+ create_graph_in_tensorboard: false
27
+ cudnn_benchmark: false
28
+ cudnn_deterministic: true
29
+ cudnn_enabled: true
30
+ detect_anomaly: false
31
+ dist_backend: nccl
32
+ dist_init_method: env://
33
+ dist_launcher: null
34
+ dist_master_addr: localhost
35
+ dist_master_port: 51885
36
+ dist_rank: 0
37
+ dist_world_size: 2
38
+ distributed: true
39
+ drop_last_iter: false
40
+ dry_run: false
41
+ early_stopping_criterion:
42
+ - valid
43
+ - loss
44
+ - min
45
+ energy_extract: energy
46
+ energy_extract_conf:
47
+ fs: 48000
48
+ hop_length: 1024
49
+ n_fft: 8192
50
+ reduction_factor: 1
51
+ win_length: 4096
52
+ energy_normalize: global_mvn
53
+ energy_normalize_conf:
54
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/telugu_latest/male/model/energy_stats.npz
55
+ exclude_weight_decay: false
56
+ exclude_weight_decay_conf: {}
57
+ feats_extract: fbank
58
+ feats_extract_conf:
59
+ fmax: 24000
60
+ fmin: 0
61
+ fs: 48000
62
+ hop_length: 1024
63
+ n_fft: 8192
64
+ n_mels: 160
65
+ win_length: 4096
66
+ fold_length:
67
+ - 150
68
+ - 819200
69
+ freeze_param: []
70
+ g2p: null
71
+ grad_clip: 1.0
72
+ grad_clip_type: 2.0
73
+ grad_noise: false
74
+ ignore_init_mismatch: false
75
+ init_param: []
76
+ iterator_type: sequence
77
+ keep_nbest_models: 5
78
+ local_rank: 0
79
+ log_interval: null
80
+ log_level: INFO
81
+ max_cache_fd: 32
82
+ max_cache_size: 0.0
83
+ max_epoch: 1000
84
+ model_conf: {}
85
+ multiple_iterator: false
86
+ multiprocessing_distributed: true
87
+ nbest_averaging_interval: 0
88
+ ngpu: 1
89
+ no_forward_run: false
90
+ non_linguistic_symbols: null
91
+ normalize: global_mvn
92
+ normalize_conf:
93
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/telugu_latest/male/model/feats_stats.npz
94
+ num_att_plot: 3
95
+ num_cache_chunks: 1024
96
+ num_iters_per_epoch: 800
97
+ num_workers: 1
98
+ odim: null
99
+ optim: adam
100
+ optim_conf:
101
+ lr: 1.0
102
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
103
+ patience: null
104
+ pitch_extract: dio
105
+ pitch_extract_conf:
106
+ f0max: 350
107
+ f0min: 40
108
+ fs: 48000
109
+ hop_length: 1024
110
+ n_fft: 8192
111
+ reduction_factor: 1
112
+ pitch_normalize: global_mvn
113
+ pitch_normalize_conf:
114
+ stats_file: /home/speech/Fastspeech2_latest models/Fastspeech2_HS/telugu_latest/male/model/pitch_stats.npz
115
+ pretrain_path: null
116
+ print_config: false
117
+ required:
118
+ - output_dir
119
+ - token_list
120
+ resume: true
121
+ save_strategy: all
122
+ scheduler: noamlr
123
+ scheduler_conf:
124
+ model_size: 384
125
+ warmup_steps: 4000
126
+ seed: 0
127
+ sharded_ddp: false
128
+ shuffle_within_batch: false
129
+ sort_batch: descending
130
+ sort_in_batch: descending
131
+ token_list:
132
+ - <blank>
133
+ - <unk>
134
+ - <space>
135
+ - $
136
+ - ','
137
+ - .
138
+ - M
139
+ - q
140
+ - H
141
+ - k
142
+ - "\u0916"
143
+ - g
144
+ - "\u0918"
145
+ - "\u0919"
146
+ - c
147
+ - C
148
+ - j
149
+ - J
150
+ - "\u091E"
151
+ - "\u091F"
152
+ - "\u0920"
153
+ - "\u0921"
154
+ - "\u0922"
155
+ - "\u0923"
156
+ - t
157
+ - "\u0925"
158
+ - d
159
+ - "\u0927"
160
+ - n
161
+ - "\u0929"
162
+ - p
163
+ - P
164
+ - b
165
+ - B
166
+ - m
167
+ - y
168
+ - r
169
+ - "\u0931"
170
+ - l
171
+ - "\u0D33"
172
+ - Z
173
+ - w
174
+ - "\u0936"
175
+ - "\u0937"
176
+ - s
177
+ - h
178
+ - Y
179
+ - "\u093D"
180
+ - "\u0915"
181
+ - K
182
+ - G
183
+ - z
184
+ - D
185
+ - T
186
+ - f
187
+ - "\u0960"
188
+ - "\u0D7A"
189
+ - N
190
+ - "\u0D7C"
191
+ - "\u0D7D"
192
+ - "\u0D7E"
193
+ - a
194
+ - A
195
+ - i
196
+ - I
197
+ - u
198
+ - U
199
+ - R
200
+ - "\u090D"
201
+ - e
202
+ - E
203
+ - "\u0910"
204
+ - "\u0911"
205
+ - o
206
+ - O
207
+ - "\u0914"
208
+ - "\u0B89"
209
+ - <sos/eos>
210
+ token_type: char
211
+ train_data_path_and_name_and_type:
212
+ - - dump/raw/tr_no_dev/text
213
+ - text
214
+ - text
215
+ - - duration_info/tr_no_dev/durations
216
+ - durations
217
+ - text_int
218
+ - - dump/raw/tr_no_dev/wav.scp
219
+ - speech
220
+ - sound
221
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
222
+ - pitch
223
+ - npy
224
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
225
+ - energy
226
+ - npy
227
+ train_dtype: float32
228
+ train_shape_file:
229
+ - exp/tts_stats_raw_char_None/train/text_shape.char
230
+ - exp/tts_stats_raw_char_None/train/speech_shape
231
+ tts: fastspeech2
232
+ tts_conf:
233
+ adim: 384
234
+ aheads: 2
235
+ decoder_normalize_before: true
236
+ dlayers: 4
237
+ dunits: 1536
238
+ duration_predictor_chans: 256
239
+ duration_predictor_kernel_size: 3
240
+ duration_predictor_layers: 2
241
+ elayers: 4
242
+ encoder_normalize_before: true
243
+ energy_embed_dropout: 0.0
244
+ energy_embed_kernel_size: 1
245
+ energy_predictor_chans: 256
246
+ energy_predictor_dropout: 0.5
247
+ energy_predictor_kernel_size: 3
248
+ energy_predictor_layers: 2
249
+ eunits: 1536
250
+ init_dec_alpha: 1.0
251
+ init_enc_alpha: 1.0
252
+ init_type: xavier_uniform
253
+ pitch_embed_dropout: 0.0
254
+ pitch_embed_kernel_size: 1
255
+ pitch_predictor_chans: 256
256
+ pitch_predictor_dropout: 0.5
257
+ pitch_predictor_kernel_size: 5
258
+ pitch_predictor_layers: 5
259
+ positionwise_conv_kernel_size: 3
260
+ positionwise_layer_type: conv1d
261
+ postnet_chans: 256
262
+ postnet_filts: 5
263
+ postnet_layers: 5
264
+ reduction_factor: 1
265
+ stop_gradient_from_energy_predictor: false
266
+ stop_gradient_from_pitch_predictor: true
267
+ transformer_dec_attn_dropout_rate: 0.2
268
+ transformer_dec_dropout_rate: 0.2
269
+ transformer_dec_positional_dropout_rate: 0.2
270
+ transformer_enc_attn_dropout_rate: 0.2
271
+ transformer_enc_dropout_rate: 0.2
272
+ transformer_enc_positional_dropout_rate: 0.2
273
+ use_masking: true
274
+ use_scaled_pos_enc: true
275
+ unused_parameters: false
276
+ use_adapter: false
277
+ use_amp: false
278
+ use_matplotlib: true
279
+ use_preprocessor: true
280
+ use_tensorboard: true
281
+ use_wandb: false
282
+ val_scheduler_criterion:
283
+ - valid
284
+ - loss
285
+ valid_batch_bins: null
286
+ valid_batch_size: null
287
+ valid_batch_type: null
288
+ valid_data_path_and_name_and_type:
289
+ - - dump/raw/dev/text
290
+ - text
291
+ - text
292
+ - - duration_info/dev/durations
293
+ - durations
294
+ - text_int
295
+ - - dump/raw/dev/wav.scp
296
+ - speech
297
+ - sound
298
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
299
+ - pitch
300
+ - npy
301
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
302
+ - energy
303
+ - npy
304
+ valid_iterator_type: null
305
+ valid_max_cache_size: null
306
+ valid_shape_file:
307
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
308
+ - exp/tts_stats_raw_char_None/valid/speech_shape
309
+ version: '202402'
310
+ wandb_entity: null
311
+ wandb_id: null
312
+ wandb_model_log_interval: -1
313
+ wandb_name: null
314
+ wandb_project: null
315
+ write_collected_feats: false
telugu_latest/male/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:836d86ad9256687abe3b29c5d216e09513b9e5644208c2c80cbe4e2bf3bff8a4
3
+ size 770
telugu_latest/male/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a19dda037c9bbbdb2d4e89c01c73b94f407cddbacc1876047c34598e2eccdd33
3
+ size 2042
telugu_latest/male/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e112166efb214d6ccc0684abc561aa94544cb652f974732f6a833034462d4708
3
+ size 152128410
telugu_latest/male/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8597f126532136ba5cfb699e0d53bb9f1e813518c78a9d27cb9ee7dbb4edd41e
3
+ size 770