jonatasgrosman commited on
Commit
76972b8
1 Parent(s): 93244a8

jonatasgrosman/exp_w2v2r_en_vp-100k_age_teens-10_sixties-0_s232

Browse files
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ cv7_transcriptions.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ tags:
6
+ - automatic-speech-recognition
7
+ - en
8
+ datasets:
9
+ - mozilla-foundation/common_voice_7_0
10
+ ---
11
+ # exp_w2v2r_en_vp-100k_age_teens-10_sixties-0_s232
12
+
13
+ Fine-tuned [facebook/wav2vec2-large-100k-voxpopuli](https://huggingface.co/facebook/wav2vec2-large-100k-voxpopuli) for speech recognition using the train split of [Common Voice 7.0 (en)](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0).
14
+ When using this model, make sure that your speech input is sampled at 16kHz.
15
+
16
+ This model has been fine-tuned by the [HuggingSound](https://github.com/jonatasgrosman/huggingsound) tool.
17
+
config.json ADDED
@@ -0,0 +1,116 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-100k-voxpopuli",
3
+ "activation_dropout": 0.05,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForCTC"
10
+ ],
11
+ "attention_dropout": 0.05,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": true,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "mean",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_dropout": 0.0,
51
+ "feat_extract_norm": "layer",
52
+ "feat_proj_dropout": 0.05,
53
+ "feat_quantizer_dropout": 0.0,
54
+ "final_dropout": 0.05,
55
+ "gradient_checkpointing": false,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.05,
58
+ "hidden_size": 1024,
59
+ "initializer_range": 0.02,
60
+ "intermediate_size": 4096,
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.05,
63
+ "mask_channel_length": 10,
64
+ "mask_channel_min_space": 1,
65
+ "mask_channel_other": 0.0,
66
+ "mask_channel_prob": 0.0,
67
+ "mask_channel_selection": "static",
68
+ "mask_feature_length": 10,
69
+ "mask_feature_min_masks": 0,
70
+ "mask_feature_prob": 0.0,
71
+ "mask_time_length": 10,
72
+ "mask_time_min_masks": 2,
73
+ "mask_time_min_space": 1,
74
+ "mask_time_other": 0.0,
75
+ "mask_time_prob": 0.05,
76
+ "mask_time_selection": "static",
77
+ "model_type": "wav2vec2",
78
+ "num_adapter_layers": 3,
79
+ "num_attention_heads": 16,
80
+ "num_codevector_groups": 2,
81
+ "num_codevectors_per_group": 320,
82
+ "num_conv_pos_embedding_groups": 16,
83
+ "num_conv_pos_embeddings": 128,
84
+ "num_feat_extract_layers": 7,
85
+ "num_hidden_layers": 24,
86
+ "num_negatives": 100,
87
+ "output_hidden_size": 1024,
88
+ "pad_token_id": 0,
89
+ "proj_codevector_dim": 768,
90
+ "tdnn_dilation": [
91
+ 1,
92
+ 2,
93
+ 3,
94
+ 1,
95
+ 1
96
+ ],
97
+ "tdnn_dim": [
98
+ 512,
99
+ 512,
100
+ 512,
101
+ 512,
102
+ 1500
103
+ ],
104
+ "tdnn_kernel": [
105
+ 5,
106
+ 3,
107
+ 3,
108
+ 1,
109
+ 1
110
+ ],
111
+ "torch_dtype": "float32",
112
+ "transformers_version": "4.18.0",
113
+ "use_weighted_layer_sum": false,
114
+ "vocab_size": 33,
115
+ "xvector_output_dim": 512
116
+ }
cv7_evaluation.json ADDED
@@ -0,0 +1,390 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "validation": [
3
+ {
4
+ "subset_type": "all",
5
+ "subset": "all",
6
+ "cer": 0.15690520847432501,
7
+ "wer": 0.42916979244811204
8
+ },
9
+ {
10
+ "subset_type": "age",
11
+ "subset": "",
12
+ "cer": 0.15553069730389169,
13
+ "wer": 0.4261148636041348
14
+ },
15
+ {
16
+ "subset_type": "age",
17
+ "subset": "thirties",
18
+ "cer": 0.1587861235283007,
19
+ "wer": 0.4339243282763663
20
+ },
21
+ {
22
+ "subset_type": "age",
23
+ "subset": "sixties",
24
+ "cer": 0.13030148185998977,
25
+ "wer": 0.3856143856143856
26
+ },
27
+ {
28
+ "subset_type": "age",
29
+ "subset": "seventies",
30
+ "cer": 0.14376053962900506,
31
+ "wer": 0.4090909090909091
32
+ },
33
+ {
34
+ "subset_type": "age",
35
+ "subset": "fifties",
36
+ "cer": 0.13370210469508906,
37
+ "wer": 0.38893280632411065
38
+ },
39
+ {
40
+ "subset_type": "age",
41
+ "subset": "twenties",
42
+ "cer": 0.1783295379416734,
43
+ "wer": 0.4683979705907645
44
+ },
45
+ {
46
+ "subset_type": "age",
47
+ "subset": "eighties",
48
+ "cer": 0.17204301075268819,
49
+ "wer": 0.46153846153846156
50
+ },
51
+ {
52
+ "subset_type": "age",
53
+ "subset": "nineties",
54
+ "cer": 0.11538461538461539,
55
+ "wer": 0.42857142857142855
56
+ },
57
+ {
58
+ "subset_type": "age",
59
+ "subset": "teens",
60
+ "cer": 0.15630703711214763,
61
+ "wer": 0.4403292181069959
62
+ },
63
+ {
64
+ "subset_type": "age",
65
+ "subset": "fourties",
66
+ "cer": 0.15195460277427492,
67
+ "wer": 0.4256410256410256
68
+ },
69
+ {
70
+ "subset_type": "gender",
71
+ "subset": "",
72
+ "cer": 0.155631526284732,
73
+ "wer": 0.42630365179231133
74
+ },
75
+ {
76
+ "subset_type": "gender",
77
+ "subset": "male",
78
+ "cer": 0.16476722412345837,
79
+ "wer": 0.44697860333881967
80
+ },
81
+ {
82
+ "subset_type": "gender",
83
+ "subset": "other",
84
+ "cer": 0.1513353115727003,
85
+ "wer": 0.41916167664670656
86
+ },
87
+ {
88
+ "subset_type": "gender",
89
+ "subset": "female",
90
+ "cer": 0.1580870795347375,
91
+ "wer": 0.43115279048490396
92
+ },
93
+ {
94
+ "subset_type": "accent",
95
+ "subset": "",
96
+ "cer": 0.15779193113956225,
97
+ "wer": 0.43014049604132976
98
+ },
99
+ {
100
+ "subset_type": "accent",
101
+ "subset": "england",
102
+ "cer": 0.15012023359670215,
103
+ "wer": 0.4285004965243297
104
+ },
105
+ {
106
+ "subset_type": "accent",
107
+ "subset": "australia",
108
+ "cer": 0.11499436302142052,
109
+ "wer": 0.3733031674208145
110
+ },
111
+ {
112
+ "subset_type": "accent",
113
+ "subset": "malaysia",
114
+ "cer": 0.2076923076923077,
115
+ "wer": 0.4722222222222222
116
+ },
117
+ {
118
+ "subset_type": "accent",
119
+ "subset": "indian",
120
+ "cer": 0.18623919308357348,
121
+ "wer": 0.48707155742633795
122
+ },
123
+ {
124
+ "subset_type": "accent",
125
+ "subset": "wales",
126
+ "cer": 0.10734463276836158,
127
+ "wer": 0.3819444444444444
128
+ },
129
+ {
130
+ "subset_type": "accent",
131
+ "subset": "philippines",
132
+ "cer": 0.19452706722189173,
133
+ "wer": 0.5190311418685121
134
+ },
135
+ {
136
+ "subset_type": "accent",
137
+ "subset": "hongkong",
138
+ "cer": 0.16390423572744015,
139
+ "wer": 0.4782608695652174
140
+ },
141
+ {
142
+ "subset_type": "accent",
143
+ "subset": "bermuda",
144
+ "cer": 0.07407407407407407,
145
+ "wer": 0.1875
146
+ },
147
+ {
148
+ "subset_type": "accent",
149
+ "subset": "ireland",
150
+ "cer": 0.18642951251646903,
151
+ "wer": 0.4763779527559055
152
+ },
153
+ {
154
+ "subset_type": "accent",
155
+ "subset": "newzealand",
156
+ "cer": 0.06896551724137931,
157
+ "wer": 0.275
158
+ },
159
+ {
160
+ "subset_type": "accent",
161
+ "subset": "other",
162
+ "cer": 0.1691449814126394,
163
+ "wer": 0.54
164
+ },
165
+ {
166
+ "subset_type": "accent",
167
+ "subset": "us",
168
+ "cer": 0.1360331694147141,
169
+ "wer": 0.3934370771312585
170
+ },
171
+ {
172
+ "subset_type": "accent",
173
+ "subset": "african",
174
+ "cer": 0.1800878477306003,
175
+ "wer": 0.46647230320699706
176
+ },
177
+ {
178
+ "subset_type": "accent",
179
+ "subset": "scotland",
180
+ "cer": 0.10676156583629894,
181
+ "wer": 0.32954545454545453
182
+ },
183
+ {
184
+ "subset_type": "accent",
185
+ "subset": "singapore",
186
+ "cer": 0.20594965675057209,
187
+ "wer": 0.43506493506493504
188
+ },
189
+ {
190
+ "subset_type": "accent",
191
+ "subset": "southatlandtic",
192
+ "cer": 0.16483516483516483,
193
+ "wer": 0.5757575757575758
194
+ },
195
+ {
196
+ "subset_type": "accent",
197
+ "subset": "canada",
198
+ "cer": 0.12281059063136457,
199
+ "wer": 0.38369304556354916
200
+ }
201
+ ],
202
+ "test": [
203
+ {
204
+ "subset_type": "all",
205
+ "subset": "all",
206
+ "cer": 0.1960324343832403,
207
+ "wer": 0.48523507487085593
208
+ },
209
+ {
210
+ "subset_type": "age",
211
+ "subset": "",
212
+ "cer": 0.19513983265427481,
213
+ "wer": 0.4835426476293349
214
+ },
215
+ {
216
+ "subset_type": "age",
217
+ "subset": "thirties",
218
+ "cer": 0.18660184149384734,
219
+ "wer": 0.47609756097560973
220
+ },
221
+ {
222
+ "subset_type": "age",
223
+ "subset": "seventies",
224
+ "cer": 0.162748643761302,
225
+ "wer": 0.4161073825503356
226
+ },
227
+ {
228
+ "subset_type": "age",
229
+ "subset": "fifties",
230
+ "cer": 0.24879393521709167,
231
+ "wer": 0.5578231292517006
232
+ },
233
+ {
234
+ "subset_type": "age",
235
+ "subset": "twenties",
236
+ "cer": 0.20102615028136378,
237
+ "wer": 0.5001428979708488
238
+ },
239
+ {
240
+ "subset_type": "age",
241
+ "subset": "eighties",
242
+ "cer": 0.13142857142857142,
243
+ "wer": 0.44
244
+ },
245
+ {
246
+ "subset_type": "age",
247
+ "subset": "fourties",
248
+ "cer": 0.19296762741949117,
249
+ "wer": 0.47274436090225563
250
+ },
251
+ {
252
+ "subset_type": "age",
253
+ "subset": "teens",
254
+ "cer": 0.2143495620046481,
255
+ "wer": 0.5187202178352621
256
+ },
257
+ {
258
+ "subset_type": "age",
259
+ "subset": "sixties",
260
+ "cer": 0.22173215717722533,
261
+ "wer": 0.43832599118942733
262
+ },
263
+ {
264
+ "subset_type": "gender",
265
+ "subset": "",
266
+ "cer": 0.1950776003267668,
267
+ "wer": 0.48345030529301286
268
+ },
269
+ {
270
+ "subset_type": "gender",
271
+ "subset": "male",
272
+ "cer": 0.20361625144186687,
273
+ "wer": 0.498960498960499
274
+ },
275
+ {
276
+ "subset_type": "gender",
277
+ "subset": "other",
278
+ "cer": 0.22453222453222454,
279
+ "wer": 0.48249027237354086
280
+ },
281
+ {
282
+ "subset_type": "gender",
283
+ "subset": "female",
284
+ "cer": 0.19046625638885992,
285
+ "wer": 0.4826048171275647
286
+ },
287
+ {
288
+ "subset_type": "accent",
289
+ "subset": "",
290
+ "cer": 0.19672541140770627,
291
+ "wer": 0.48648395779655407
292
+ },
293
+ {
294
+ "subset_type": "accent",
295
+ "subset": "england",
296
+ "cer": 0.1707678624369997,
297
+ "wer": 0.44771057094403616
298
+ },
299
+ {
300
+ "subset_type": "accent",
301
+ "subset": "malaysia",
302
+ "cer": 0.15677491601343785,
303
+ "wer": 0.4267515923566879
304
+ },
305
+ {
306
+ "subset_type": "accent",
307
+ "subset": "australia",
308
+ "cer": 0.15578158458244112,
309
+ "wer": 0.4140127388535032
310
+ },
311
+ {
312
+ "subset_type": "accent",
313
+ "subset": "indian",
314
+ "cer": 0.21252747252747253,
315
+ "wer": 0.5098963242224317
316
+ },
317
+ {
318
+ "subset_type": "accent",
319
+ "subset": "wales",
320
+ "cer": 0.1917808219178082,
321
+ "wer": 0.4166666666666667
322
+ },
323
+ {
324
+ "subset_type": "accent",
325
+ "subset": "philippines",
326
+ "cer": 0.14046822742474915,
327
+ "wer": 0.36363636363636365
328
+ },
329
+ {
330
+ "subset_type": "accent",
331
+ "subset": "hongkong",
332
+ "cer": 0.24855491329479767,
333
+ "wer": 0.5583333333333333
334
+ },
335
+ {
336
+ "subset_type": "accent",
337
+ "subset": "bermuda",
338
+ "cer": 0.24044943820224718,
339
+ "wer": 0.6071428571428571
340
+ },
341
+ {
342
+ "subset_type": "accent",
343
+ "subset": "ireland",
344
+ "cer": 0.17053701015965167,
345
+ "wer": 0.4177215189873418
346
+ },
347
+ {
348
+ "subset_type": "accent",
349
+ "subset": "newzealand",
350
+ "cer": 0.1280323450134771,
351
+ "wer": 0.375
352
+ },
353
+ {
354
+ "subset_type": "accent",
355
+ "subset": "other",
356
+ "cer": 0.250390015600624,
357
+ "wer": 0.5482625482625483
358
+ },
359
+ {
360
+ "subset_type": "accent",
361
+ "subset": "us",
362
+ "cer": 0.1815231501514496,
363
+ "wer": 0.46523001681186
364
+ },
365
+ {
366
+ "subset_type": "accent",
367
+ "subset": "african",
368
+ "cer": 0.2226680040120361,
369
+ "wer": 0.49710982658959535
370
+ },
371
+ {
372
+ "subset_type": "accent",
373
+ "subset": "scotland",
374
+ "cer": 0.1753155680224404,
375
+ "wer": 0.46099290780141844
376
+ },
377
+ {
378
+ "subset_type": "accent",
379
+ "subset": "singapore",
380
+ "cer": 0.24064171122994651,
381
+ "wer": 0.5151515151515151
382
+ },
383
+ {
384
+ "subset_type": "accent",
385
+ "subset": "canada",
386
+ "cer": 0.2048162859980139,
387
+ "wer": 0.46629213483146065
388
+ }
389
+ ]
390
+ }
cv7_transcriptions.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c50f27406bd9a7b156c25ab6795a14518527bc800a009103db3704e42ae05fa
3
+ size 120843077
preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "processor_class": "Wav2Vec2Processor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f03a48c4a49ac93c46c0eaac64765ca5fc6b294ba14b6bae6413054f31edcbb2
3
+ size 1262034033
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "do_lower_case": false, "word_delimiter_token": "|", "replace_word_delimiter_char": " ", "special_tokens_map_file": null, "name_or_path": "/tmp/tmplzmdoo04", "tokenizer_class": "Wav2Vec2CTCTokenizer", "processor_class": "Wav2Vec2Processor"}
vocab.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"<pad>": 0, "<s>": 1, "</s>": 2, "<unk>": 3, "|": 4, "'": 5, "-": 6, "a": 7, "b": 8, "c": 9, "d": 10, "e": 11, "f": 12, "g": 13, "h": 14, "i": 15, "j": 16, "k": 17, "l": 18, "m": 19, "n": 20, "o": 21, "p": 22, "q": 23, "r": 24, "s": 25, "t": 26, "u": 27, "v": 28, "w": 29, "x": 30, "y": 31, "z": 32}