Hugging Face's logo

smtiitm
/
FastSpeech2_HS_latest_models

tts
indictts
fs2
mfa
HS
hybrid_segmentation
fastspeech2
Model card Files
xet
 Community
3
utkarsh2299 commited on
Commit
76fb012
·
verified ·
1 Parent(s): 5dd2edf

Upload 12 files

Browse files
Files changed (12) hide show
  1. rajasthani_latest/female/model/config.yaml +320 -0
  2. rajasthani_latest/female/model/energy_stats.npz +3 -0
  3. rajasthani_latest/female/model/feats_stats.npz +3 -0
  4. rajasthani_latest/female/model/feats_type +1 -0
  5. rajasthani_latest/female/model/model.pth +3 -0
  6. rajasthani_latest/female/model/pitch_stats.npz +3 -0
  7. rajasthani_latest/male/model/config.yaml +320 -0
  8. rajasthani_latest/male/model/energy_stats.npz +3 -0
  9. rajasthani_latest/male/model/feats_stats.npz +3 -0
  10. rajasthani_latest/male/model/feats_type +1 -0
  11. rajasthani_latest/male/model/model.pth +3 -0
  12. rajasthani_latest/male/model/pitch_stats.npz +3 -0
rajasthani_latest/female/model/config.yaml ADDED
@@ -0,0 +1,320 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ category_sample_size: 10
18
+ chunk_default_fs: null
19
+ chunk_discard_short_samples: true
20
+ chunk_excluded_key_prefixes: []
21
+ chunk_length: 500
22
+ chunk_max_abs_length: null
23
+ chunk_shift_ratio: 0.5
24
+ cleaner: null
25
+ collect_stats: false
26
+ config: conf/tuning/train_fastspeech2.yaml
27
+ create_graph_in_tensorboard: false
28
+ cudnn_benchmark: false
29
+ cudnn_deterministic: true
30
+ cudnn_enabled: true
31
+ deepspeed_config: null
32
+ detect_anomaly: false
33
+ dist_backend: nccl
34
+ dist_init_method: env://
35
+ dist_launcher: null
36
+ dist_master_addr: localhost
37
+ dist_master_port: 59527
38
+ dist_rank: 0
39
+ dist_world_size: 2
40
+ distributed: true
41
+ drop_last_iter: false
42
+ dry_run: false
43
+ early_stopping_criterion:
44
+ - valid
45
+ - loss
46
+ - min
47
+ energy_extract: energy
48
+ energy_extract_conf:
49
+ fs: 48000
50
+ hop_length: 1024
51
+ n_fft: 8192
52
+ reduction_factor: 1
53
+ win_length: 4096
54
+ energy_normalize: global_mvn
55
+ energy_normalize_conf:
56
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_fem_48/energy_stats.npz
57
+ exclude_weight_decay: false
58
+ exclude_weight_decay_conf: {}
59
+ feats_extract: fbank
60
+ feats_extract_conf:
61
+ fmax: 24000
62
+ fmin: 0
63
+ fs: 48000
64
+ hop_length: 1024
65
+ n_fft: 8192
66
+ n_mels: 160
67
+ win_length: 4096
68
+ fold_length:
69
+ - 150
70
+ - 819200
71
+ freeze_param: []
72
+ g2p: null
73
+ grad_clip: 1.0
74
+ grad_clip_type: 2.0
75
+ grad_noise: false
76
+ ignore_init_mismatch: false
77
+ init_param: []
78
+ iterator_type: sequence
79
+ keep_nbest_models: 5
80
+ local_rank: 0
81
+ log_interval: null
82
+ log_level: INFO
83
+ max_cache_fd: 32
84
+ max_cache_size: 0.0
85
+ max_epoch: 1000
86
+ model_conf: {}
87
+ multi_task_dataset: false
88
+ multiple_iterator: false
89
+ multiprocessing_distributed: true
90
+ nbest_averaging_interval: 0
91
+ ngpu: 1
92
+ no_forward_run: false
93
+ non_linguistic_symbols: null
94
+ normalize: global_mvn
95
+ normalize_conf:
96
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_fem_48/feats_stats.npz
97
+ num_att_plot: 3
98
+ num_cache_chunks: 1024
99
+ num_iters_per_epoch: 800
100
+ num_workers: 1
101
+ odim: null
102
+ optim: adam
103
+ optim_conf:
104
+ lr: 1.0
105
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
106
+ patience: null
107
+ pitch_extract: dio
108
+ pitch_extract_conf:
109
+ f0max: 400
110
+ f0min: 80
111
+ fs: 48000
112
+ hop_length: 1024
113
+ n_fft: 8192
114
+ reduction_factor: 1
115
+ pitch_normalize: global_mvn
116
+ pitch_normalize_conf:
117
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_fem_48/pitch_stats.npz
118
+ pretrain_path: null
119
+ print_config: false
120
+ required:
121
+ - output_dir
122
+ - token_list
123
+ resume: true
124
+ save_strategy: all
125
+ scheduler: noamlr
126
+ scheduler_conf:
127
+ model_size: 384
128
+ warmup_steps: 4000
129
+ seed: 0
130
+ sharded_ddp: false
131
+ shuffle_within_batch: false
132
+ sort_batch: descending
133
+ sort_in_batch: descending
134
+ token_list:
135
+ - <blank>
136
+ - <unk>
137
+ - <space>
138
+ - $
139
+ - ','
140
+ - .
141
+ - M
142
+ - q
143
+ - H
144
+ - k
145
+ - "\u0916"
146
+ - g
147
+ - "\u0918"
148
+ - "\u0919"
149
+ - c
150
+ - C
151
+ - j
152
+ - J
153
+ - "\u091E"
154
+ - "\u091F"
155
+ - "\u0920"
156
+ - "\u0921"
157
+ - "\u0922"
158
+ - "\u0923"
159
+ - t
160
+ - "\u0925"
161
+ - d
162
+ - "\u0927"
163
+ - n
164
+ - "\u0929"
165
+ - p
166
+ - P
167
+ - b
168
+ - B
169
+ - m
170
+ - y
171
+ - r
172
+ - "\u0931"
173
+ - l
174
+ - "\u0D33"
175
+ - Z
176
+ - w
177
+ - "\u0936"
178
+ - "\u0937"
179
+ - s
180
+ - h
181
+ - Y
182
+ - "\u093D"
183
+ - "\u0915"
184
+ - K
185
+ - G
186
+ - z
187
+ - D
188
+ - T
189
+ - f
190
+ - "\u0960"
191
+ - "\u0D7A"
192
+ - N
193
+ - "\u0D7C"
194
+ - "\u0D7D"
195
+ - "\u0D7E"
196
+ - a
197
+ - A
198
+ - i
199
+ - I
200
+ - u
201
+ - U
202
+ - R
203
+ - "\u090D"
204
+ - e
205
+ - E
206
+ - "\u0910"
207
+ - "\u0911"
208
+ - o
209
+ - O
210
+ - "\u0914"
211
+ - "\u0B89"
212
+ - <sos/eos>
213
+ token_type: char
214
+ train_data_path_and_name_and_type:
215
+ - - dump/raw/tr_no_dev/text
216
+ - text
217
+ - text
218
+ - - duration_info/tr_no_dev/durations
219
+ - durations
220
+ - text_int
221
+ - - dump/raw/tr_no_dev/wav.scp
222
+ - speech
223
+ - sound
224
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
225
+ - pitch
226
+ - npy
227
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
228
+ - energy
229
+ - npy
230
+ train_dtype: float32
231
+ train_shape_file:
232
+ - exp/tts_stats_raw_char_None/train/text_shape.char
233
+ - exp/tts_stats_raw_char_None/train/speech_shape
234
+ tts: fastspeech2
235
+ tts_conf:
236
+ adim: 384
237
+ aheads: 2
238
+ decoder_normalize_before: true
239
+ dlayers: 4
240
+ dunits: 1536
241
+ duration_predictor_chans: 256
242
+ duration_predictor_kernel_size: 3
243
+ duration_predictor_layers: 2
244
+ elayers: 4
245
+ encoder_normalize_before: true
246
+ energy_embed_dropout: 0.0
247
+ energy_embed_kernel_size: 1
248
+ energy_predictor_chans: 256
249
+ energy_predictor_dropout: 0.5
250
+ energy_predictor_kernel_size: 3
251
+ energy_predictor_layers: 2
252
+ eunits: 1536
253
+ init_dec_alpha: 1.0
254
+ init_enc_alpha: 1.0
255
+ init_type: xavier_uniform
256
+ pitch_embed_dropout: 0.0
257
+ pitch_embed_kernel_size: 1
258
+ pitch_predictor_chans: 256
259
+ pitch_predictor_dropout: 0.5
260
+ pitch_predictor_kernel_size: 5
261
+ pitch_predictor_layers: 5
262
+ positionwise_conv_kernel_size: 3
263
+ positionwise_layer_type: conv1d
264
+ postnet_chans: 256
265
+ postnet_filts: 5
266
+ postnet_layers: 5
267
+ reduction_factor: 1
268
+ stop_gradient_from_energy_predictor: false
269
+ stop_gradient_from_pitch_predictor: true
270
+ transformer_dec_attn_dropout_rate: 0.2
271
+ transformer_dec_dropout_rate: 0.2
272
+ transformer_dec_positional_dropout_rate: 0.2
273
+ transformer_enc_attn_dropout_rate: 0.2
274
+ transformer_enc_dropout_rate: 0.2
275
+ transformer_enc_positional_dropout_rate: 0.2
276
+ use_masking: true
277
+ use_scaled_pos_enc: true
278
+ unused_parameters: false
279
+ use_adapter: false
280
+ use_amp: false
281
+ use_deepspeed: false
282
+ use_matplotlib: true
283
+ use_preprocessor: true
284
+ use_tensorboard: true
285
+ use_tf32: false
286
+ use_wandb: false
287
+ val_scheduler_criterion:
288
+ - valid
289
+ - loss
290
+ valid_batch_bins: null
291
+ valid_batch_size: null
292
+ valid_batch_type: null
293
+ valid_data_path_and_name_and_type:
294
+ - - dump/raw/dev/text
295
+ - text
296
+ - text
297
+ - - duration_info/dev/durations
298
+ - durations
299
+ - text_int
300
+ - - dump/raw/dev/wav.scp
301
+ - speech
302
+ - sound
303
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
304
+ - pitch
305
+ - npy
306
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
307
+ - energy
308
+ - npy
309
+ valid_iterator_type: null
310
+ valid_max_cache_size: null
311
+ valid_shape_file:
312
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
313
+ - exp/tts_stats_raw_char_None/valid/speech_shape
314
+ version: '202412'
315
+ wandb_entity: null
316
+ wandb_id: null
317
+ wandb_model_log_interval: -1
318
+ wandb_name: null
319
+ wandb_project: null
320
+ write_collected_feats: false
rajasthani_latest/female/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:895b4003e6e8753ef9fe5f4b94c06cf7ec05cdbcd5c5350ee5639f6fc43376f4
3
+ size 770
rajasthani_latest/female/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f64fa57b21949f4b7297535767786f34aa2aa0efc467ba435a325ae629d891ea
3
+ size 2042
rajasthani_latest/female/model/feats_type ADDED
@@ -0,0 +1 @@
 
 
1
+ raw
rajasthani_latest/female/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:911da13deed6042414f3b197e41196434e9eeae42fc386e15322a3dc70c0dcad
3
+ size 152129434
rajasthani_latest/female/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0cf4809edd26d6d578a4038cda5318201b9f6f55f96948dcc8f90e9d009a18f
3
+ size 770
rajasthani_latest/male/model/config.yaml ADDED
@@ -0,0 +1,320 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ category_sample_size: 10
18
+ chunk_default_fs: null
19
+ chunk_discard_short_samples: true
20
+ chunk_excluded_key_prefixes: []
21
+ chunk_length: 500
22
+ chunk_max_abs_length: null
23
+ chunk_shift_ratio: 0.5
24
+ cleaner: null
25
+ collect_stats: false
26
+ config: conf/tuning/train_fastspeech2.yaml
27
+ create_graph_in_tensorboard: false
28
+ cudnn_benchmark: false
29
+ cudnn_deterministic: true
30
+ cudnn_enabled: true
31
+ deepspeed_config: null
32
+ detect_anomaly: false
33
+ dist_backend: nccl
34
+ dist_init_method: env://
35
+ dist_launcher: null
36
+ dist_master_addr: localhost
37
+ dist_master_port: 43245
38
+ dist_rank: 0
39
+ dist_world_size: 2
40
+ distributed: true
41
+ drop_last_iter: false
42
+ dry_run: false
43
+ early_stopping_criterion:
44
+ - valid
45
+ - loss
46
+ - min
47
+ energy_extract: energy
48
+ energy_extract_conf:
49
+ fs: 48000
50
+ hop_length: 1024
51
+ n_fft: 8192
52
+ reduction_factor: 1
53
+ win_length: 4096
54
+ energy_normalize: global_mvn
55
+ energy_normalize_conf:
56
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_male_mono/energy_stats.npz
57
+ exclude_weight_decay: false
58
+ exclude_weight_decay_conf: {}
59
+ feats_extract: fbank
60
+ feats_extract_conf:
61
+ fmax: 24000
62
+ fmin: 0
63
+ fs: 48000
64
+ hop_length: 1024
65
+ n_fft: 8192
66
+ n_mels: 160
67
+ win_length: 4096
68
+ fold_length:
69
+ - 150
70
+ - 819200
71
+ freeze_param: []
72
+ g2p: null
73
+ grad_clip: 1.0
74
+ grad_clip_type: 2.0
75
+ grad_noise: false
76
+ ignore_init_mismatch: false
77
+ init_param: []
78
+ iterator_type: sequence
79
+ keep_nbest_models: 5
80
+ local_rank: 0
81
+ log_interval: null
82
+ log_level: INFO
83
+ max_cache_fd: 32
84
+ max_cache_size: 0.0
85
+ max_epoch: 1000
86
+ model_conf: {}
87
+ multi_task_dataset: false
88
+ multiple_iterator: false
89
+ multiprocessing_distributed: true
90
+ nbest_averaging_interval: 0
91
+ ngpu: 1
92
+ no_forward_run: false
93
+ non_linguistic_symbols: null
94
+ normalize: global_mvn
95
+ normalize_conf:
96
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_male_mono/feats_stats.npz
97
+ num_att_plot: 3
98
+ num_cache_chunks: 1024
99
+ num_iters_per_epoch: 800
100
+ num_workers: 1
101
+ odim: null
102
+ optim: adam
103
+ optim_conf:
104
+ lr: 1.0
105
+ output_dir: exp/tts_train_fastspeech2_raw_char_None
106
+ patience: null
107
+ pitch_extract: dio
108
+ pitch_extract_conf:
109
+ f0max: 350
110
+ f0min: 40
111
+ fs: 48000
112
+ hop_length: 1024
113
+ n_fft: 8192
114
+ reduction_factor: 1
115
+ pitch_normalize: global_mvn
116
+ pitch_normalize_conf:
117
+ stats_file: /home/speech/Suji/FS2/Fastspeech2_HS/Models_ph3/Raj_male_mono/pitch_stats.npz
118
+ pretrain_path: null
119
+ print_config: false
120
+ required:
121
+ - output_dir
122
+ - token_list
123
+ resume: true
124
+ save_strategy: all
125
+ scheduler: noamlr
126
+ scheduler_conf:
127
+ model_size: 384
128
+ warmup_steps: 4000
129
+ seed: 0
130
+ sharded_ddp: false
131
+ shuffle_within_batch: false
132
+ sort_batch: descending
133
+ sort_in_batch: descending
134
+ token_list:
135
+ - <blank>
136
+ - <unk>
137
+ - <space>
138
+ - $
139
+ - ','
140
+ - .
141
+ - M
142
+ - q
143
+ - H
144
+ - k
145
+ - "\u0916"
146
+ - g
147
+ - "\u0918"
148
+ - "\u0919"
149
+ - c
150
+ - C
151
+ - j
152
+ - J
153
+ - "\u091E"
154
+ - "\u091F"
155
+ - "\u0920"
156
+ - "\u0921"
157
+ - "\u0922"
158
+ - "\u0923"
159
+ - t
160
+ - "\u0925"
161
+ - d
162
+ - "\u0927"
163
+ - n
164
+ - "\u0929"
165
+ - p
166
+ - P
167
+ - b
168
+ - B
169
+ - m
170
+ - y
171
+ - r
172
+ - "\u0931"
173
+ - l
174
+ - "\u0D33"
175
+ - Z
176
+ - w
177
+ - "\u0936"
178
+ - "\u0937"
179
+ - s
180
+ - h
181
+ - Y
182
+ - "\u093D"
183
+ - "\u0915"
184
+ - K
185
+ - G
186
+ - z
187
+ - D
188
+ - T
189
+ - f
190
+ - "\u0960"
191
+ - "\u0D7A"
192
+ - N
193
+ - "\u0D7C"
194
+ - "\u0D7D"
195
+ - "\u0D7E"
196
+ - a
197
+ - A
198
+ - i
199
+ - I
200
+ - u
201
+ - U
202
+ - R
203
+ - "\u090D"
204
+ - e
205
+ - E
206
+ - "\u0910"
207
+ - "\u0911"
208
+ - o
209
+ - O
210
+ - "\u0914"
211
+ - "\u0B89"
212
+ - <sos/eos>
213
+ token_type: char
214
+ train_data_path_and_name_and_type:
215
+ - - dump/raw/tr_no_dev/text
216
+ - text
217
+ - text
218
+ - - duration_info/tr_no_dev/durations
219
+ - durations
220
+ - text_int
221
+ - - dump/raw/tr_no_dev/wav.scp
222
+ - speech
223
+ - sound
224
+ - - exp/tts_stats_raw_char_None/train/collect_feats/pitch.scp
225
+ - pitch
226
+ - npy
227
+ - - exp/tts_stats_raw_char_None/train/collect_feats/energy.scp
228
+ - energy
229
+ - npy
230
+ train_dtype: float32
231
+ train_shape_file:
232
+ - exp/tts_stats_raw_char_None/train/text_shape.char
233
+ - exp/tts_stats_raw_char_None/train/speech_shape
234
+ tts: fastspeech2
235
+ tts_conf:
236
+ adim: 384
237
+ aheads: 2
238
+ decoder_normalize_before: true
239
+ dlayers: 4
240
+ dunits: 1536
241
+ duration_predictor_chans: 256
242
+ duration_predictor_kernel_size: 3
243
+ duration_predictor_layers: 2
244
+ elayers: 4
245
+ encoder_normalize_before: true
246
+ energy_embed_dropout: 0.0
247
+ energy_embed_kernel_size: 1
248
+ energy_predictor_chans: 256
249
+ energy_predictor_dropout: 0.5
250
+ energy_predictor_kernel_size: 3
251
+ energy_predictor_layers: 2
252
+ eunits: 1536
253
+ init_dec_alpha: 1.0
254
+ init_enc_alpha: 1.0
255
+ init_type: xavier_uniform
256
+ pitch_embed_dropout: 0.0
257
+ pitch_embed_kernel_size: 1
258
+ pitch_predictor_chans: 256
259
+ pitch_predictor_dropout: 0.5
260
+ pitch_predictor_kernel_size: 5
261
+ pitch_predictor_layers: 5
262
+ positionwise_conv_kernel_size: 3
263
+ positionwise_layer_type: conv1d
264
+ postnet_chans: 256
265
+ postnet_filts: 5
266
+ postnet_layers: 5
267
+ reduction_factor: 1
268
+ stop_gradient_from_energy_predictor: false
269
+ stop_gradient_from_pitch_predictor: true
270
+ transformer_dec_attn_dropout_rate: 0.2
271
+ transformer_dec_dropout_rate: 0.2
272
+ transformer_dec_positional_dropout_rate: 0.2
273
+ transformer_enc_attn_dropout_rate: 0.2
274
+ transformer_enc_dropout_rate: 0.2
275
+ transformer_enc_positional_dropout_rate: 0.2
276
+ use_masking: true
277
+ use_scaled_pos_enc: true
278
+ unused_parameters: false
279
+ use_adapter: false
280
+ use_amp: false
281
+ use_deepspeed: false
282
+ use_matplotlib: true
283
+ use_preprocessor: true
284
+ use_tensorboard: true
285
+ use_tf32: false
286
+ use_wandb: false
287
+ val_scheduler_criterion:
288
+ - valid
289
+ - loss
290
+ valid_batch_bins: null
291
+ valid_batch_size: null
292
+ valid_batch_type: null
293
+ valid_data_path_and_name_and_type:
294
+ - - dump/raw/dev/text
295
+ - text
296
+ - text
297
+ - - duration_info/dev/durations
298
+ - durations
299
+ - text_int
300
+ - - dump/raw/dev/wav.scp
301
+ - speech
302
+ - sound
303
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/pitch.scp
304
+ - pitch
305
+ - npy
306
+ - - exp/tts_stats_raw_char_None/valid/collect_feats/energy.scp
307
+ - energy
308
+ - npy
309
+ valid_iterator_type: null
310
+ valid_max_cache_size: null
311
+ valid_shape_file:
312
+ - exp/tts_stats_raw_char_None/valid/text_shape.char
313
+ - exp/tts_stats_raw_char_None/valid/speech_shape
314
+ version: '202412'
315
+ wandb_entity: null
316
+ wandb_id: null
317
+ wandb_model_log_interval: -1
318
+ wandb_name: null
319
+ wandb_project: null
320
+ write_collected_feats: false
rajasthani_latest/male/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9fd1763526c68808542ff8b30c985275b81cf362cacf9fde798de0fabe52d32
3
+ size 770
rajasthani_latest/male/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99a70109ff9f7a71255917dcd7dc384c75065dcba9ee97f3619248f50a7382da
3
+ size 2042
rajasthani_latest/male/model/feats_type ADDED
@@ -0,0 +1 @@
 
 
1
+ raw
rajasthani_latest/male/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f06ac6bcb9e7caa21ad6e82928f64bc94d03be64cc15cdf3d97722769cfc33cd
3
+ size 152129434
rajasthani_latest/male/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fd9a0fc2cbbf1be75ee36c77478d56f5097dfb5acd43189466d70f4cdb33ac9
3
+ size 770