ovieyra21 commited on
Commit
ba2681d
1 Parent(s): ab6d063

first commit

Browse files
README.md ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - es
4
+ license: mit
5
+ base_model: microsoft/speecht5_tts
6
+ tags:
7
+ - generated_from_trainer
8
+ datasets:
9
+ - ylacombe/google-chilean-spanish
10
+ model-index:
11
+ - name: SpeechT5 TTS mabama
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # SpeechT5 TTS mabama
19
+
20
+ This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the google-chilean-spanish dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.3995
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 1e-05
42
+ - train_batch_size: 16
43
+ - eval_batch_size: 8
44
+ - seed: 42
45
+ - gradient_accumulation_steps: 2
46
+ - total_train_batch_size: 32
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: linear
49
+ - lr_scheduler_warmup_steps: 1000
50
+ - training_steps: 1200
51
+ - mixed_precision_training: Native AMP
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:----:|:---------------:|
57
+ | 0.4516 | 25.53 | 1200 | 0.3995 |
58
+
59
+
60
+ ### Framework versions
61
+
62
+ - Transformers 4.39.0.dev0
63
+ - Pytorch 2.1.0+cu121
64
+ - Datasets 2.17.1
65
+ - Tokenizers 0.15.2
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<ctc_blank>": 80,
3
+ "<mask>": 79
4
+ }
checkpoint-1200/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<ctc_blank>": 80,
3
+ "<mask>": 79
4
+ }
checkpoint-1200/config.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/speecht5_tts",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "SpeechT5ForTextToSpeech"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 0,
10
+ "conv_bias": false,
11
+ "conv_dim": [
12
+ 512,
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512
19
+ ],
20
+ "conv_kernel": [
21
+ 10,
22
+ 3,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 2,
27
+ 2
28
+ ],
29
+ "conv_stride": [
30
+ 5,
31
+ 2,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2
37
+ ],
38
+ "decoder_attention_heads": 12,
39
+ "decoder_ffn_dim": 3072,
40
+ "decoder_layerdrop": 0.1,
41
+ "decoder_layers": 6,
42
+ "decoder_start_token_id": 2,
43
+ "encoder_attention_heads": 12,
44
+ "encoder_ffn_dim": 3072,
45
+ "encoder_layerdrop": 0.1,
46
+ "encoder_layers": 12,
47
+ "encoder_max_relative_position": 160,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_norm": "group",
51
+ "feat_proj_dropout": 0.0,
52
+ "guided_attention_loss_num_heads": 2,
53
+ "guided_attention_loss_scale": 10.0,
54
+ "guided_attention_loss_sigma": 0.4,
55
+ "hidden_act": "gelu",
56
+ "hidden_dropout": 0.1,
57
+ "hidden_size": 768,
58
+ "initializer_range": 0.02,
59
+ "is_encoder_decoder": true,
60
+ "layer_norm_eps": 1e-05,
61
+ "mask_feature_length": 10,
62
+ "mask_feature_min_masks": 0,
63
+ "mask_feature_prob": 0.0,
64
+ "mask_time_length": 10,
65
+ "mask_time_min_masks": 2,
66
+ "mask_time_prob": 0.05,
67
+ "max_length": 1876,
68
+ "max_speech_positions": 1876,
69
+ "max_text_positions": 600,
70
+ "model_type": "speecht5",
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_mel_bins": 80,
75
+ "pad_token_id": 1,
76
+ "positional_dropout": 0.1,
77
+ "reduction_factor": 2,
78
+ "scale_embedding": false,
79
+ "speaker_embedding_dim": 512,
80
+ "speech_decoder_postnet_dropout": 0.5,
81
+ "speech_decoder_postnet_kernel": 5,
82
+ "speech_decoder_postnet_layers": 5,
83
+ "speech_decoder_postnet_units": 256,
84
+ "speech_decoder_prenet_dropout": 0.5,
85
+ "speech_decoder_prenet_layers": 2,
86
+ "speech_decoder_prenet_units": 256,
87
+ "torch_dtype": "float32",
88
+ "transformers_version": "4.39.0.dev0",
89
+ "use_cache": false,
90
+ "use_guided_attention_loss": true,
91
+ "vocab_size": 81
92
+ }
checkpoint-1200/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "eos_token_id": 2,
6
+ "max_length": 1876,
7
+ "pad_token_id": 1,
8
+ "transformers_version": "4.39.0.dev0"
9
+ }
checkpoint-1200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07465a555b92f42d646060f915c82147b9778577d2756ac9666c94563bf25d65
3
+ size 1155772233
checkpoint-1200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c31d5c2e1ede4f2e3fb805f57c4fc76f1c9db8c375b23e3049d8fecaf7190ac5
3
+ size 14244
checkpoint-1200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12c6e8db7f13bd903e719156588677a1fbe7d18cc00951615a883bcce900298c
3
+ size 1064
checkpoint-1200/special_tokens_map.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": {
5
+ "content": "<mask>",
6
+ "lstrip": true,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ },
11
+ "pad_token": "<pad>",
12
+ "unk_token": "<unk>"
13
+ }
checkpoint-1200/spm_char.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
3
+ size 238473
checkpoint-1200/tokenizer_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "79": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": true,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "80": {
44
+ "content": "<ctc_blank>",
45
+ "lstrip": false,
46
+ "normalized": true,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": false
50
+ }
51
+ },
52
+ "bos_token": "<s>",
53
+ "clean_up_tokenization_spaces": true,
54
+ "eos_token": "</s>",
55
+ "mask_token": "<mask>",
56
+ "model_max_length": 600,
57
+ "normalize": false,
58
+ "pad_token": "<pad>",
59
+ "processor_class": "SpeechT5Processor",
60
+ "sp_model_kwargs": {},
61
+ "tokenizer_class": "SpeechT5Tokenizer",
62
+ "unk_token": "<unk>"
63
+ }
checkpoint-1200/trainer_state.json ADDED
@@ -0,0 +1,365 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3995174765586853,
3
+ "best_model_checkpoint": "./speecht5_tts_mabama_es/checkpoint-1200",
4
+ "epoch": 25.53191489361702,
5
+ "eval_steps": 1200,
6
+ "global_step": 1200,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.53,
13
+ "grad_norm": 12.126326560974121,
14
+ "learning_rate": 2.2e-07,
15
+ "loss": 1.0931,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 1.06,
20
+ "grad_norm": 9.78133487701416,
21
+ "learning_rate": 4.7000000000000005e-07,
22
+ "loss": 1.0628,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 1.6,
27
+ "grad_norm": 24.999067306518555,
28
+ "learning_rate": 7.2e-07,
29
+ "loss": 0.947,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 2.13,
34
+ "grad_norm": 5.398886680603027,
35
+ "learning_rate": 9.7e-07,
36
+ "loss": 0.898,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 2.66,
41
+ "grad_norm": 12.817495346069336,
42
+ "learning_rate": 1.2200000000000002e-06,
43
+ "loss": 0.8029,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 3.19,
48
+ "grad_norm": 7.071120738983154,
49
+ "learning_rate": 1.4700000000000001e-06,
50
+ "loss": 0.7889,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 3.72,
55
+ "grad_norm": 4.3572163581848145,
56
+ "learning_rate": 1.72e-06,
57
+ "loss": 0.7719,
58
+ "step": 175
59
+ },
60
+ {
61
+ "epoch": 4.26,
62
+ "grad_norm": 5.61962366104126,
63
+ "learning_rate": 1.97e-06,
64
+ "loss": 0.764,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 4.79,
69
+ "grad_norm": 6.79171085357666,
70
+ "learning_rate": 2.2200000000000003e-06,
71
+ "loss": 0.7441,
72
+ "step": 225
73
+ },
74
+ {
75
+ "epoch": 5.32,
76
+ "grad_norm": 6.525346755981445,
77
+ "learning_rate": 2.47e-06,
78
+ "loss": 0.711,
79
+ "step": 250
80
+ },
81
+ {
82
+ "epoch": 5.85,
83
+ "grad_norm": 1.7001874446868896,
84
+ "learning_rate": 2.7200000000000002e-06,
85
+ "loss": 0.7059,
86
+ "step": 275
87
+ },
88
+ {
89
+ "epoch": 6.38,
90
+ "grad_norm": 4.468942642211914,
91
+ "learning_rate": 2.97e-06,
92
+ "loss": 0.7021,
93
+ "step": 300
94
+ },
95
+ {
96
+ "epoch": 6.91,
97
+ "grad_norm": 7.033669948577881,
98
+ "learning_rate": 3.2200000000000005e-06,
99
+ "loss": 0.6266,
100
+ "step": 325
101
+ },
102
+ {
103
+ "epoch": 7.45,
104
+ "grad_norm": 3.8984785079956055,
105
+ "learning_rate": 3.4700000000000002e-06,
106
+ "loss": 0.6133,
107
+ "step": 350
108
+ },
109
+ {
110
+ "epoch": 7.98,
111
+ "grad_norm": 3.643061876296997,
112
+ "learning_rate": 3.7200000000000004e-06,
113
+ "loss": 0.58,
114
+ "step": 375
115
+ },
116
+ {
117
+ "epoch": 8.51,
118
+ "grad_norm": 6.500895023345947,
119
+ "learning_rate": 3.97e-06,
120
+ "loss": 0.5389,
121
+ "step": 400
122
+ },
123
+ {
124
+ "epoch": 9.04,
125
+ "grad_norm": 3.851013422012329,
126
+ "learning_rate": 4.22e-06,
127
+ "loss": 0.5578,
128
+ "step": 425
129
+ },
130
+ {
131
+ "epoch": 9.57,
132
+ "grad_norm": 5.976474285125732,
133
+ "learning_rate": 4.47e-06,
134
+ "loss": 0.5266,
135
+ "step": 450
136
+ },
137
+ {
138
+ "epoch": 10.11,
139
+ "grad_norm": 4.052149772644043,
140
+ "learning_rate": 4.7200000000000005e-06,
141
+ "loss": 0.5102,
142
+ "step": 475
143
+ },
144
+ {
145
+ "epoch": 10.64,
146
+ "grad_norm": 6.552683353424072,
147
+ "learning_rate": 4.970000000000001e-06,
148
+ "loss": 0.5382,
149
+ "step": 500
150
+ },
151
+ {
152
+ "epoch": 11.17,
153
+ "grad_norm": 2.714641571044922,
154
+ "learning_rate": 5.220000000000001e-06,
155
+ "loss": 0.5167,
156
+ "step": 525
157
+ },
158
+ {
159
+ "epoch": 11.7,
160
+ "grad_norm": 3.426229953765869,
161
+ "learning_rate": 5.470000000000001e-06,
162
+ "loss": 0.5106,
163
+ "step": 550
164
+ },
165
+ {
166
+ "epoch": 12.23,
167
+ "grad_norm": 5.78665018081665,
168
+ "learning_rate": 5.72e-06,
169
+ "loss": 0.5155,
170
+ "step": 575
171
+ },
172
+ {
173
+ "epoch": 12.77,
174
+ "grad_norm": 1.958343744277954,
175
+ "learning_rate": 5.9700000000000004e-06,
176
+ "loss": 0.4872,
177
+ "step": 600
178
+ },
179
+ {
180
+ "epoch": 13.3,
181
+ "grad_norm": 2.1352875232696533,
182
+ "learning_rate": 6.220000000000001e-06,
183
+ "loss": 0.4866,
184
+ "step": 625
185
+ },
186
+ {
187
+ "epoch": 13.83,
188
+ "grad_norm": 3.6809630393981934,
189
+ "learning_rate": 6.470000000000001e-06,
190
+ "loss": 0.5014,
191
+ "step": 650
192
+ },
193
+ {
194
+ "epoch": 14.36,
195
+ "grad_norm": 2.484553098678589,
196
+ "learning_rate": 6.720000000000001e-06,
197
+ "loss": 0.4892,
198
+ "step": 675
199
+ },
200
+ {
201
+ "epoch": 14.89,
202
+ "grad_norm": 3.656264305114746,
203
+ "learning_rate": 6.97e-06,
204
+ "loss": 0.4712,
205
+ "step": 700
206
+ },
207
+ {
208
+ "epoch": 15.43,
209
+ "grad_norm": 3.709062337875366,
210
+ "learning_rate": 7.22e-06,
211
+ "loss": 0.4843,
212
+ "step": 725
213
+ },
214
+ {
215
+ "epoch": 15.96,
216
+ "grad_norm": 2.776505708694458,
217
+ "learning_rate": 7.4700000000000005e-06,
218
+ "loss": 0.4743,
219
+ "step": 750
220
+ },
221
+ {
222
+ "epoch": 16.49,
223
+ "grad_norm": 3.913029909133911,
224
+ "learning_rate": 7.72e-06,
225
+ "loss": 0.4841,
226
+ "step": 775
227
+ },
228
+ {
229
+ "epoch": 17.02,
230
+ "grad_norm": 3.667405366897583,
231
+ "learning_rate": 7.970000000000002e-06,
232
+ "loss": 0.4853,
233
+ "step": 800
234
+ },
235
+ {
236
+ "epoch": 17.55,
237
+ "grad_norm": 2.917257070541382,
238
+ "learning_rate": 8.220000000000001e-06,
239
+ "loss": 0.4773,
240
+ "step": 825
241
+ },
242
+ {
243
+ "epoch": 18.09,
244
+ "grad_norm": 3.051287889480591,
245
+ "learning_rate": 8.47e-06,
246
+ "loss": 0.4772,
247
+ "step": 850
248
+ },
249
+ {
250
+ "epoch": 18.62,
251
+ "grad_norm": 3.4974617958068848,
252
+ "learning_rate": 8.720000000000001e-06,
253
+ "loss": 0.4599,
254
+ "step": 875
255
+ },
256
+ {
257
+ "epoch": 19.15,
258
+ "grad_norm": 2.259305238723755,
259
+ "learning_rate": 8.97e-06,
260
+ "loss": 0.4694,
261
+ "step": 900
262
+ },
263
+ {
264
+ "epoch": 19.68,
265
+ "grad_norm": 3.4457366466522217,
266
+ "learning_rate": 9.220000000000002e-06,
267
+ "loss": 0.4616,
268
+ "step": 925
269
+ },
270
+ {
271
+ "epoch": 20.21,
272
+ "grad_norm": 7.221055030822754,
273
+ "learning_rate": 9.47e-06,
274
+ "loss": 0.4613,
275
+ "step": 950
276
+ },
277
+ {
278
+ "epoch": 20.74,
279
+ "grad_norm": 9.327741622924805,
280
+ "learning_rate": 9.72e-06,
281
+ "loss": 0.4634,
282
+ "step": 975
283
+ },
284
+ {
285
+ "epoch": 21.28,
286
+ "grad_norm": 3.205131769180298,
287
+ "learning_rate": 9.970000000000001e-06,
288
+ "loss": 0.4691,
289
+ "step": 1000
290
+ },
291
+ {
292
+ "epoch": 21.81,
293
+ "grad_norm": 5.403283596038818,
294
+ "learning_rate": 8.900000000000001e-06,
295
+ "loss": 0.46,
296
+ "step": 1025
297
+ },
298
+ {
299
+ "epoch": 22.34,
300
+ "grad_norm": 2.766523838043213,
301
+ "learning_rate": 7.650000000000001e-06,
302
+ "loss": 0.4462,
303
+ "step": 1050
304
+ },
305
+ {
306
+ "epoch": 22.87,
307
+ "grad_norm": 3.033956289291382,
308
+ "learning_rate": 6.4000000000000006e-06,
309
+ "loss": 0.4491,
310
+ "step": 1075
311
+ },
312
+ {
313
+ "epoch": 23.4,
314
+ "grad_norm": 3.4086856842041016,
315
+ "learning_rate": 5.150000000000001e-06,
316
+ "loss": 0.4452,
317
+ "step": 1100
318
+ },
319
+ {
320
+ "epoch": 23.94,
321
+ "grad_norm": 2.144127607345581,
322
+ "learning_rate": 3.900000000000001e-06,
323
+ "loss": 0.4508,
324
+ "step": 1125
325
+ },
326
+ {
327
+ "epoch": 24.47,
328
+ "grad_norm": 2.2593584060668945,
329
+ "learning_rate": 2.6500000000000005e-06,
330
+ "loss": 0.4432,
331
+ "step": 1150
332
+ },
333
+ {
334
+ "epoch": 25.0,
335
+ "grad_norm": 3.3258821964263916,
336
+ "learning_rate": 1.4000000000000001e-06,
337
+ "loss": 0.4526,
338
+ "step": 1175
339
+ },
340
+ {
341
+ "epoch": 25.53,
342
+ "grad_norm": 1.783236026763916,
343
+ "learning_rate": 1.5000000000000002e-07,
344
+ "loss": 0.4516,
345
+ "step": 1200
346
+ },
347
+ {
348
+ "epoch": 25.53,
349
+ "eval_loss": 0.3995174765586853,
350
+ "eval_runtime": 4.5311,
351
+ "eval_samples_per_second": 36.857,
352
+ "eval_steps_per_second": 4.635,
353
+ "step": 1200
354
+ }
355
+ ],
356
+ "logging_steps": 25,
357
+ "max_steps": 1200,
358
+ "num_input_tokens_seen": 0,
359
+ "num_train_epochs": 26,
360
+ "save_steps": 1200,
361
+ "total_flos": 3336935371995456.0,
362
+ "train_batch_size": 16,
363
+ "trial_name": null,
364
+ "trial_params": null
365
+ }
checkpoint-1200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2348a7d121ebb0f003e5327a860d2b586573a1dfa4f4efd4966d03b80bdffb6b
3
+ size 5048
config.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/speecht5_tts",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "SpeechT5ForTextToSpeech"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 0,
10
+ "conv_bias": false,
11
+ "conv_dim": [
12
+ 512,
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512
19
+ ],
20
+ "conv_kernel": [
21
+ 10,
22
+ 3,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 2,
27
+ 2
28
+ ],
29
+ "conv_stride": [
30
+ 5,
31
+ 2,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2
37
+ ],
38
+ "decoder_attention_heads": 12,
39
+ "decoder_ffn_dim": 3072,
40
+ "decoder_layerdrop": 0.1,
41
+ "decoder_layers": 6,
42
+ "decoder_start_token_id": 2,
43
+ "encoder_attention_heads": 12,
44
+ "encoder_ffn_dim": 3072,
45
+ "encoder_layerdrop": 0.1,
46
+ "encoder_layers": 12,
47
+ "encoder_max_relative_position": 160,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_norm": "group",
51
+ "feat_proj_dropout": 0.0,
52
+ "guided_attention_loss_num_heads": 2,
53
+ "guided_attention_loss_scale": 10.0,
54
+ "guided_attention_loss_sigma": 0.4,
55
+ "hidden_act": "gelu",
56
+ "hidden_dropout": 0.1,
57
+ "hidden_size": 768,
58
+ "initializer_range": 0.02,
59
+ "is_encoder_decoder": true,
60
+ "layer_norm_eps": 1e-05,
61
+ "mask_feature_length": 10,
62
+ "mask_feature_min_masks": 0,
63
+ "mask_feature_prob": 0.0,
64
+ "mask_time_length": 10,
65
+ "mask_time_min_masks": 2,
66
+ "mask_time_prob": 0.05,
67
+ "max_length": 1876,
68
+ "max_speech_positions": 1876,
69
+ "max_text_positions": 600,
70
+ "model_type": "speecht5",
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_mel_bins": 80,
75
+ "pad_token_id": 1,
76
+ "positional_dropout": 0.1,
77
+ "reduction_factor": 2,
78
+ "scale_embedding": false,
79
+ "speaker_embedding_dim": 512,
80
+ "speech_decoder_postnet_dropout": 0.5,
81
+ "speech_decoder_postnet_kernel": 5,
82
+ "speech_decoder_postnet_layers": 5,
83
+ "speech_decoder_postnet_units": 256,
84
+ "speech_decoder_prenet_dropout": 0.5,
85
+ "speech_decoder_prenet_layers": 2,
86
+ "speech_decoder_prenet_units": 256,
87
+ "torch_dtype": "float32",
88
+ "transformers_version": "4.39.0.dev0",
89
+ "use_cache": false,
90
+ "use_guided_attention_loss": true,
91
+ "vocab_size": 81
92
+ }
generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "eos_token_id": 2,
6
+ "max_length": 1876,
7
+ "pad_token_id": 1,
8
+ "transformers_version": "4.39.0.dev0"
9
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1389aa9355d6fe37832cd91011d8c734dd3038ed3f20a4edb21bf74ecac186fc
3
+ size 577789320
runs/Feb25_12-26-41_fde21f8503a9/events.out.tfevents.1708864066.fde21f8503a9.849.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:884b25fe81c36a40748f4473682bc1abc568f9153c8552854c41f2b1699d9d16
3
+ size 16959
special_tokens_map.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": {
5
+ "content": "<mask>",
6
+ "lstrip": true,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ },
11
+ "pad_token": "<pad>",
12
+ "unk_token": "<unk>"
13
+ }
spm_char.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
3
+ size 238473
tokenizer_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "79": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": true,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "80": {
44
+ "content": "<ctc_blank>",
45
+ "lstrip": false,
46
+ "normalized": true,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": false
50
+ }
51
+ },
52
+ "bos_token": "<s>",
53
+ "clean_up_tokenization_spaces": true,
54
+ "eos_token": "</s>",
55
+ "mask_token": "<mask>",
56
+ "model_max_length": 600,
57
+ "normalize": false,
58
+ "pad_token": "<pad>",
59
+ "processor_class": "SpeechT5Processor",
60
+ "sp_model_kwargs": {},
61
+ "tokenizer_class": "SpeechT5Tokenizer",
62
+ "unk_token": "<unk>"
63
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2348a7d121ebb0f003e5327a860d2b586573a1dfa4f4efd4966d03b80bdffb6b
3
+ size 5048