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
4ad7769
·
verified ·
1 Parent(s): dac8fd5

Upload 11 files

Browse files
Files changed (11) hide show
  1. hindi_latest/female/model/config.yaml +315 -0
  2. hindi_latest/female/model/energy_stats.npz +3 -0
  3. hindi_latest/female/model/feats_stats.npz +3 -0
  4. hindi_latest/female/model/feats_type +1 -0
  5. hindi_latest/female/model/model.pth +3 -0
  6. hindi_latest/female/model/pitch_stats.npz +3 -0
  7. hindi_latest/male/model/config.yaml +315 -0
  8. hindi_latest/male/model/energy_stats.npz +3 -0
  9. hindi_latest/male/model/feats_stats.npz +3 -0
  10. hindi_latest/male/model/model.pth +3 -0
  11. hindi_latest/male/model/pitch_stats.npz +3 -0
hindi_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: 46493
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/hindi_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_latest models/Fastspeech2_HS/hindi_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_latest models/Fastspeech2_HS/hindi_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
hindi_latest/female/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b44ea2ed74bc1e7af45af7648ab670dba4a3c730348ccb47577f6a39edcd0fcd
3
+ size 770
hindi_latest/female/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c4dbc43b70f4a380b84a6ea2c333ebcbc6648fcd08b43a54f14eb4bcc48baef
3
+ size 2042
hindi_latest/female/model/feats_type ADDED
@@ -0,0 +1 @@
 
 
1
+ raw
hindi_latest/female/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:915f27d8dc30199fdcff8a69c33bb412639b0a8cac0c16508efe10a85e14c055
3
+ size 152128410
hindi_latest/female/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b27b409f5f96d40d22ab0ff6b229153ce9e7c1bb77901b474db52d7c32fd8be
3
+ size 770
hindi_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: 43833
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/hindi_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/hindi_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/hindi_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
hindi_latest/male/model/energy_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c4978d68634bc22b158ee2f763fb9d3e22fafc339e38526a16475645abb81da
3
+ size 770
hindi_latest/male/model/feats_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71f32e049987675bc9827b3d4bf42ce40cf205a173a13417be4a85406cb1a16d
3
+ size 2042
hindi_latest/male/model/model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d01835e3969800080f4e006850c21ab4186edd5d451c4f7d6be71edb47ffacf0
3
+ size 152128410
hindi_latest/male/model/pitch_stats.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83c6cace2471652aa85ee6ebc9997c3cd81393782e2cc58e33a2461e4cbda8fd
3
+ size 770