balaji1312 commited on
Commit
a6dd6b8
·
1 Parent(s): 70a6ca8

Update model

Browse files
Files changed (23) hide show
  1. README.md +2 -0
  2. data/en_token_list/bpe_unigram1024/bpe.model +3 -0
  3. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/86epoch.pth +3 -0
  4. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/RESULTS.md +70 -0
  5. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/config.yaml +1247 -0
  6. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/acc.png +0 -0
  7. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/backward_time.png +0 -0
  8. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/cer.png +0 -0
  9. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/cer_ctc.png +0 -0
  10. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/clip.png +0 -0
  11. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/forward_time.png +0 -0
  12. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/gpu_max_cached_mem_GB.png +0 -0
  13. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/grad_norm.png +0 -0
  14. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/iter_time.png +0 -0
  15. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss.png +0 -0
  16. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_att.png +0 -0
  17. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_ctc.png +0 -0
  18. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_scale.png +0 -0
  19. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/optim0_lr0.png +0 -0
  20. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/optim_step_time.png +0 -0
  21. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/train_time.png +0 -0
  22. exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/wer.png +0 -0
  23. meta.yaml +8 -0
README.md ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ # balaji1312/asr_train_asr_wavlm_transformer_raw_en_bpe1024_valid.acc.best
2
+ This model was uploaded to Hugging Face by balaji1312.
data/en_token_list/bpe_unigram1024/bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31edf4d881243bcd7b93def6327a6d1d12c082bfa697934ba5ce5448894082d2
3
+ size 252821
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/86epoch.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a3f71fe9c4ca2e2951a7aed67bdf74d7aac8010076fb9b506afecf43c65db89
3
+ size 1372515526
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/RESULTS.md ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!-- Generated by scripts/utils/show_asr_result.sh -->
2
+ # RESULTS
3
+ ## Environments
4
+ - date: `Thu Jan 30 06:18:01 EST 2025`
5
+ - python version: `3.9.19 (main, May 6 2024, 19:43:03) [GCC 11.2.0]`
6
+ - espnet version: `espnet 202402`
7
+ - pytorch version: `pytorch 2.4.0`
8
+ - Git hash: `c46aa9a594ff83d52cbf61d84c5650325d1ce527`
9
+ - Commit date: `Sun Oct 13 14:39:31 2024 -0400`
10
+
11
+ ## exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024
12
+ ### WER
13
+
14
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
15
+ |---|---|---|---|---|---|---|---|---|
16
+ |decode_asr_asr_model_valid.acc.ave/test|1044|3686|34.9|41.4|23.7|20.9|86.0|60.2|
17
+ |decode_asr_asr_model_valid.acc.best/test|1044|3686|56.1|31.4|12.5|8.1|52.0|62.3|
18
+
19
+ ### CER
20
+
21
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
22
+ |---|---|---|---|---|---|---|---|---|
23
+ |decode_asr_asr_model_valid.acc.ave/test|1044|16215|54.8|17.3|28.0|25.5|70.7|60.2|
24
+ |decode_asr_asr_model_valid.acc.best/test|1044|16215|75.4|8.1|16.6|9.4|34.1|62.3|
25
+
26
+ ### TER
27
+
28
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
29
+ |---|---|---|---|---|---|---|---|---|
30
+ |decode_asr_asr_model_valid.acc.ave/test|1044|5220|45.9|29.6|24.5|21.3|75.4|60.2|
31
+ |decode_asr_asr_model_valid.acc.best/test|1044|5220|64.5|18.0|17.5|10.4|45.9|62.3|
32
+
33
+ ## exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/decode_asr_asr_model_valid.acc.best
34
+ ### WER
35
+
36
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
37
+ |---|---|---|---|---|---|---|---|---|
38
+ |org/dev|853|2372|59.8|31.2|8.9|7.2|47.3|64.0|
39
+
40
+ ### CER
41
+
42
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
43
+ |---|---|---|---|---|---|---|---|---|
44
+ |org/dev|853|9855|78.3|7.3|14.3|8.4|30.1|64.0|
45
+
46
+ ### TER
47
+
48
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
49
+ |---|---|---|---|---|---|---|---|---|
50
+ |org/dev|853|3590|68.2|16.2|15.6|6.4|38.3|64.0|
51
+
52
+ ## exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/decode_asr_asr_model_valid.acc.ave
53
+ ### WER
54
+
55
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
56
+ |---|---|---|---|---|---|---|---|---|
57
+ |org/dev|853|2372|39.9|39.2|21.0|15.8|75.9|60.3|
58
+
59
+ ### CER
60
+
61
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
62
+ |---|---|---|---|---|---|---|---|---|
63
+ |org/dev|853|9855|56.7|15.4|27.9|19.1|62.4|60.3|
64
+
65
+ ### TER
66
+
67
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
68
+ |---|---|---|---|---|---|---|---|---|
69
+ |org/dev|853|3590|51.5|25.8|22.7|13.6|62.1|60.3|
70
+
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/config.yaml ADDED
@@ -0,0 +1,1247 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config: conf/tuning/train_asr_wavlm_transformer.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/asr_train_asr_wavlm_transformer_raw_en_bpe1024
9
+ ngpu: 1
10
+ seed: 2022
11
+ num_workers: 4
12
+ num_att_plot: 0
13
+ dist_backend: nccl
14
+ dist_init_method: env://
15
+ dist_world_size: null
16
+ dist_rank: null
17
+ local_rank: 0
18
+ dist_master_addr: null
19
+ dist_master_port: null
20
+ dist_launcher: null
21
+ multiprocessing_distributed: false
22
+ unused_parameters: false
23
+ sharded_ddp: false
24
+ use_deepspeed: false
25
+ deepspeed_config: null
26
+ cudnn_enabled: true
27
+ cudnn_benchmark: false
28
+ cudnn_deterministic: false
29
+ use_tf32: false
30
+ collect_stats: false
31
+ write_collected_feats: false
32
+ max_epoch: 100
33
+ patience: null
34
+ val_scheduler_criterion:
35
+ - valid
36
+ - loss
37
+ early_stopping_criterion:
38
+ - valid
39
+ - loss
40
+ - min
41
+ best_model_criterion:
42
+ - - valid
43
+ - acc
44
+ - max
45
+ keep_nbest_models: 4
46
+ nbest_averaging_interval: 0
47
+ grad_clip: 5.0
48
+ grad_clip_type: 2.0
49
+ grad_noise: false
50
+ accum_grad: 4
51
+ no_forward_run: false
52
+ resume: true
53
+ train_dtype: float32
54
+ use_amp: true
55
+ log_interval: 400
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
+ - frontend.upstream
75
+ num_iters_per_epoch: null
76
+ batch_size: 20
77
+ valid_batch_size: null
78
+ batch_bins: 1200000
79
+ valid_batch_bins: null
80
+ category_sample_size: 10
81
+ train_shape_file:
82
+ - exp/asr_stats_raw_en_bpe1024/train/speech_shape
83
+ - exp/asr_stats_raw_en_bpe1024/train/text_shape.bpe
84
+ valid_shape_file:
85
+ - exp/asr_stats_raw_en_bpe1024/valid/speech_shape
86
+ - exp/asr_stats_raw_en_bpe1024/valid/text_shape.bpe
87
+ batch_type: numel
88
+ valid_batch_type: null
89
+ fold_length:
90
+ - 80000
91
+ - 150
92
+ sort_in_batch: descending
93
+ shuffle_within_batch: false
94
+ sort_batch: descending
95
+ multiple_iterator: false
96
+ chunk_length: 500
97
+ chunk_shift_ratio: 0.5
98
+ num_cache_chunks: 1024
99
+ chunk_excluded_key_prefixes: []
100
+ chunk_default_fs: null
101
+ chunk_max_abs_length: null
102
+ chunk_discard_short_samples: true
103
+ train_data_path_and_name_and_type:
104
+ - - dump/raw/train/wav.scp
105
+ - speech
106
+ - sound
107
+ - - dump/raw/train/text
108
+ - text
109
+ - text
110
+ valid_data_path_and_name_and_type:
111
+ - - dump/raw/dev/wav.scp
112
+ - speech
113
+ - sound
114
+ - - dump/raw/dev/text
115
+ - text
116
+ - text
117
+ multi_task_dataset: false
118
+ allow_variable_data_keys: false
119
+ max_cache_size: 0.0
120
+ max_cache_fd: 32
121
+ allow_multi_rates: false
122
+ valid_max_cache_size: null
123
+ exclude_weight_decay: false
124
+ exclude_weight_decay_conf: {}
125
+ optim: adam
126
+ optim_conf:
127
+ lr: 0.002
128
+ weight_decay: 1.0e-06
129
+ scheduler: warmuplr
130
+ scheduler_conf:
131
+ warmup_steps: 15000
132
+ token_list:
133
+ - <blank>
134
+ - <unk>
135
+ - .
136
+ - ▁I
137
+ - ▁AND
138
+ - ''''
139
+ - ▁A
140
+ - ▁YOU
141
+ - S
142
+ - ▁IT
143
+ - T
144
+ - ▁TO
145
+ - ▁THE
146
+ - ▁LIKE
147
+ - ▁THAT
148
+ - ▁NO
149
+ - ▁BECAUSE
150
+ - ▁ONE
151
+ - ▁THEN
152
+ - ▁DON
153
+ - ▁TEETH
154
+ - ▁TWO
155
+ - ▁FIVE
156
+ - ▁KNOW
157
+ - ▁MY
158
+ - ▁SO
159
+ - ▁YOUR
160
+ - ▁IS
161
+ - ▁THEM
162
+ - ▁DO
163
+ - ▁SIX
164
+ - ▁THREE
165
+ - ▁G
166
+ - ▁U
167
+ - ▁TEN
168
+ - ▁FOUR
169
+ - ▁GET
170
+ - ▁O
171
+ - ▁K
172
+ - ▁B
173
+ - ▁L
174
+ - ▁N
175
+ - ▁S
176
+ - ▁E
177
+ - ▁M
178
+ - ▁BRUSH
179
+ - ▁THIS
180
+ - ▁T
181
+ - ▁CAN
182
+ - ▁SEVEN
183
+ - ▁EIGHT
184
+ - ▁C
185
+ - ▁HAVE
186
+ - ▁PUT
187
+ - ▁MAKE
188
+ - ▁W
189
+ - ▁J
190
+ - ▁F
191
+ - ▁IN
192
+ - ▁P
193
+ - ▁NINE
194
+ - ▁Y
195
+ - ▁D
196
+ - ▁V
197
+ - ▁OKAY
198
+ - ▁Q
199
+ - ▁Z
200
+ - ▁ZERO
201
+ - ▁IF
202
+ - ▁H
203
+ - ▁WHAT
204
+ - ▁COUNT
205
+ - ING
206
+ - ▁R
207
+ - ▁X
208
+ - ▁OF
209
+ - ▁HOW
210
+ - ▁
211
+ - ▁WANT
212
+ - ▁COLOR
213
+ - ▁JUST
214
+ - ▁WITH
215
+ - ▁ON
216
+ - N
217
+ - ▁AN
218
+ - ▁MIX
219
+ - ▁COLORS
220
+ - ▁THEY
221
+ - ▁YEAH
222
+ - ▁YES
223
+ - ▁UP
224
+ - ▁BLUE
225
+ - ▁BY
226
+ - ▁GO
227
+ - M
228
+ - ▁THERE
229
+ - ▁ALL
230
+ - ▁OR
231
+ - ▁CLEAN
232
+ - ED
233
+ - ▁SEE
234
+ - ▁BUT
235
+ - ▁USE
236
+ - ▁FOR
237
+ - ▁BE
238
+ - ▁TOOTHPASTE
239
+ - ▁WAS
240
+ - ▁UM
241
+ - ▁LETTER
242
+ - ▁NEED
243
+ - ▁HE
244
+ - ▁WILL
245
+ - ▁PLUS
246
+ - ▁DOG
247
+ - ▁RED
248
+ - RE
249
+ - ▁PURPLE
250
+ - ▁NOT
251
+ - ▁CAVITIES
252
+ - ▁OH
253
+ - ▁ARE
254
+ - ▁THINK
255
+ - ▁WHY
256
+ - ▁SHE
257
+ - ▁DID
258
+ - ▁HAT
259
+ - Y
260
+ - ▁PAINT
261
+ - ▁BRUSHING
262
+ - ▁BOX
263
+ - ▁TOOTHBRUSH
264
+ - ▁SICK
265
+ - ▁OUT
266
+ - ▁ME
267
+ - ▁JUG
268
+ - ▁DOES
269
+ - ▁FLU
270
+ - ▁MAKES
271
+ - ▁WIG
272
+ - ▁SH
273
+ - ▁MAN
274
+ - ▁WE
275
+ - ▁MORE
276
+ - OULD
277
+ - ▁PLAY
278
+ - ▁SOME
279
+ - ▁JIBO
280
+ - ▁GREEN
281
+ - ▁VAN
282
+ - ▁NUMBER
283
+ - ▁YELLOW
284
+ - ▁REALLY
285
+ - D
286
+ - ▁WHITE
287
+ - ▁PINK
288
+ - ▁WATER
289
+ - ▁QUIZ
290
+ - ▁NOW
291
+ - ▁UH
292
+ - ▁DIFFERENT
293
+ - ▁RIGHT
294
+ - IND
295
+ - ▁SAY
296
+ - ▁TREE
297
+ - LL
298
+ - CH
299
+ - ▁HELP
300
+ - ▁HUNDRED
301
+ - ▁LOOK
302
+ - ▁COULD
303
+ - ▁COUNTING
304
+ - ▁WAY
305
+ - ▁MAYBE
306
+ - ▁EASY
307
+ - ▁WOULD
308
+ - ▁BLACK
309
+ - ▁TAKE
310
+ - ▁HER
311
+ - ▁LI
312
+ - E
313
+ - TTLE
314
+ - F
315
+ - ▁AL
316
+ - ▁THING
317
+ - ▁ELSE
318
+ - ▁WELL
319
+ - LY
320
+ - ▁TOGETHER
321
+ - ▁WHEN
322
+ - ▁SIDE
323
+ - ▁CAVITY
324
+ - ▁FIRST
325
+ - ▁DOWN
326
+ - ▁DAY
327
+ - ▁OTHER
328
+ - ▁HERE
329
+ - ▁CUBES
330
+ - ▁COUNTED
331
+ - ▁EVERY
332
+ - ▁SA
333
+ - ▁TELL
334
+ - ▁DAD
335
+ - ▁ORANGE
336
+ - ▁SAME
337
+ - ▁SOMETIMES
338
+ - ▁MANY
339
+ - OTHER
340
+ - ID
341
+ - ▁WON
342
+ - ▁BIT
343
+ - ▁HI
344
+ - ▁TOO
345
+ - ▁TIME
346
+ - UH
347
+ - ▁WAIT
348
+ - ▁NOTHING
349
+ - ▁FALL
350
+ - ▁NAME
351
+ - ▁LOT
352
+ - ▁THAN
353
+ - ▁EH
354
+ - ▁MEAN
355
+ - ▁NEW
356
+ - W
357
+ - H
358
+ - ▁TOOTH
359
+ - ER
360
+ - ▁FLOSS
361
+ - ▁START
362
+ - ▁BROWN
363
+ - ▁STACK
364
+ - ▁NOPE
365
+ - ▁GOOD
366
+ - A
367
+ - L
368
+ - ▁LET
369
+ - ▁WHI
370
+ - O
371
+ - ▁ALREADY
372
+ - ▁INAUDIBLE
373
+ - ▁MOUTH
374
+ - ▁EAT
375
+ - ▁HAS
376
+ - ▁DONE
377
+ - ▁THOSE
378
+ - ▁BETTER
379
+ - ▁FUN
380
+ - ▁GERMS
381
+ - TO
382
+ - ▁UMM
383
+ - CK
384
+ - SO
385
+ - EVEN
386
+ - ▁WASH
387
+ - ▁ACTUALLY
388
+ - ▁DRINK
389
+ - ▁FRIEND
390
+ - ▁REMEMBER
391
+ - ▁SUGAR
392
+ - ▁SOMETHING
393
+ - ▁HARD
394
+ - ▁COME
395
+ - ▁PAINTING
396
+ - ▁SPI
397
+ - ▁AT
398
+ - I
399
+ - TER
400
+ - ▁MUCH
401
+ - ▁GUESS
402
+ - ▁HIM
403
+ - ▁HA
404
+ - IGHT
405
+ - Z
406
+ - ▁FRO
407
+ - ▁IMPORTANT
408
+ - ▁AGAIN
409
+ - ▁STUFF
410
+ - ▁BACK
411
+ - ▁BUGS
412
+ - ▁NIGHT
413
+ - ▁ADD
414
+ - G
415
+ - ▁EA
416
+ - HIS
417
+ - K
418
+ - EVER
419
+ - ▁TH
420
+ - ▁DARK
421
+ - ▁FORGOT
422
+ - ▁MOM
423
+ - BODY
424
+ - ▁UHHUH
425
+ - ▁BAD
426
+ - ▁TURN
427
+ - ▁ANY
428
+ - AH
429
+ - EL
430
+ - U
431
+ - AKING
432
+ - VERY
433
+ - ▁GONNA
434
+ - ▁FOUAH
435
+ - ▁SURE
436
+ - ▁PULL
437
+ - ▁LONG
438
+ - ▁KEEP
439
+ - ES
440
+ - P
441
+ - ▁WAYS
442
+ - TING
443
+ - ALLY
444
+ - VE
445
+ - ONE
446
+ - ▁QUESTION
447
+ - ▁PAPER
448
+ - ▁STU
449
+ - YTHING
450
+ - ▁SHOW
451
+ - ▁CALLED
452
+ - ▁LOVE
453
+ - ▁MM
454
+ - ▁TRY
455
+ - ▁BYE
456
+ - ▁TOP
457
+ - LD
458
+ - ▁MMM
459
+ - ▁PE
460
+ - ▁NUMBERS
461
+ - BLE
462
+ - PLE
463
+ - ▁CUBE
464
+ - OUT
465
+ - R
466
+ - ▁BOTTOM
467
+ - ▁FAVORITE
468
+ - ▁SPANISH
469
+ - ▁TONGUE
470
+ - ▁SCHOOL
471
+ - ▁TWENTY
472
+ - ▁MHM
473
+ - ▁FRONT
474
+ - ▁STAY
475
+ - ▁SPELL
476
+ - ▁TEEF
477
+ - ▁LAST
478
+ - ▁GUM
479
+ - ▁HOLD
480
+ - TY
481
+ - ▁GROUPS
482
+ - ▁OFF
483
+ - ▁EQUALS
484
+ - ▁FINGERS
485
+ - ▁QUI
486
+ - RAB
487
+ - ▁MEANS
488
+ - AW
489
+ - ▁UHH
490
+ - IT
491
+ - WEE
492
+ - ▁CH
493
+ - ▁AM
494
+ - ▁SI
495
+ - RY
496
+ - SIX
497
+ - ▁WI
498
+ - ▁BEAUTIFUL
499
+ - ▁DENTIST
500
+ - ▁HEALTHY
501
+ - ▁HURT
502
+ - ▁ZEWO
503
+ - ▁KNEW
504
+ - ▁MATH
505
+ - ▁BOY
506
+ - ▁HOLE
507
+ - ▁DIRTY
508
+ - ▁YET
509
+ - ▁EX
510
+ - ▁STARTED
511
+ - ▁LIGHT
512
+ - ▁THESE
513
+ - ▁CU
514
+ - B
515
+ - ▁THINGS
516
+ - ▁GRA
517
+ - ▁WHO
518
+ - ▁TWOS
519
+ - ▁CIRCLE
520
+ - ▁YO
521
+ - ▁FINGER
522
+ - ▁BA
523
+ - CE
524
+ - OTH
525
+ - X
526
+ - IR
527
+ - MOST
528
+ - ▁LEARN
529
+ - FIVE
530
+ - CI
531
+ - ▁ANSWER
532
+ - ▁EASIER
533
+ - ▁LAUGHS
534
+ - ▁MORNING
535
+ - ▁MOUTHWASH
536
+ - ▁PICTURE
537
+ - ▁RINSE
538
+ - ▁FORGET
539
+ - ▁SISTER
540
+ - ▁THOUGH
541
+ - ▁TALKING
542
+ - ▁GROW
543
+ - ▁WHERE
544
+ - ▁MINUTES
545
+ - ▁SUP
546
+ - ▁WISH
547
+ - ▁OUR
548
+ - ▁STI
549
+ - ▁FLOSSING
550
+ - SIC
551
+ - EPT
552
+ - ▁BIG
553
+ - PER
554
+ - ▁AH
555
+ - TH
556
+ - TEN
557
+ - EN
558
+ - ▁FAI
559
+ - ▁ONES
560
+ - ▁EQUAL
561
+ - ▁SP
562
+ - KAY
563
+ - SIDE
564
+ - WAYS
565
+ - ▁AROUND
566
+ - ▁PRETTY
567
+ - ▁RAINBOW
568
+ - ▁VIOLET
569
+ - ▁LEFT
570
+ - ▁GIRL
571
+ - ▁SENSE
572
+ - ▁SOUND
573
+ - ▁EYES
574
+ - ▁EVERYTHING
575
+ - ▁GUY
576
+ - ▁SHINY
577
+ - ▁ELEVEN
578
+ - ▁READY
579
+ - ▁STICK
580
+ - ▁FROG
581
+ - ▁FOOD
582
+ - ▁KEY
583
+ - DE
584
+ - ▁PL
585
+ - ▁PART
586
+ - OVE
587
+ - ▁PR
588
+ - ▁ROT
589
+ - ▁TEE
590
+ - ▁WERE
591
+ - VER
592
+ - ▁DIS
593
+ - ▁HEY
594
+ - USH
595
+ - OH
596
+ - IN
597
+ - ISH
598
+ - OVER
599
+ - EEN
600
+ - ▁MIND
601
+ - ▁AB
602
+ - SE
603
+ - SH
604
+ - DENTAL
605
+ - OOL
606
+ - ET
607
+ - AR
608
+ - ICK
609
+ - NA
610
+ - ENT
611
+ - ▁BU
612
+ - AT
613
+ - UNTI
614
+ - OW
615
+ - OK
616
+ - ▁EL
617
+ - ▁MA
618
+ - ▁QU
619
+ - ▁WOR
620
+ - ▁SIN
621
+ - AKE
622
+ - AND
623
+ - ▁PRETEND
624
+ - ▁BUS
625
+ - ▁PLA
626
+ - ▁CALL
627
+ - ▁ONETWOTHREEFOUR
628
+ - ▁CLASS
629
+ - ▁CONNECT
630
+ - ▁DISCOVER
631
+ - ▁HOUSE
632
+ - ▁RABBIT
633
+ - ▁SQUEEZE
634
+ - ▁THOUSAND
635
+ - ▁ROBOT
636
+ - ▁SCRUB
637
+ - ▁SMELL
638
+ - EXT
639
+ - ▁BROTHER
640
+ - ▁PILE
641
+ - ▁BOTTLE
642
+ - ▁PAINTBRUSH
643
+ - IMA
644
+ - ▁CROCODILE
645
+ - ▁JUMP
646
+ - ▁CANNOT
647
+ - ▁TWICE
648
+ - ▁STOP
649
+ - UNCH
650
+ - ▁SKIN
651
+ - ▁TUR
652
+ - ▁MOVING
653
+ - IES
654
+ - ▁FAST
655
+ - ▁PRETENDING
656
+ - EEP
657
+ - ▁SHAKING
658
+ - ▁MAY
659
+ - ▁FAKE
660
+ - ▁AWAY
661
+ - ▁DI
662
+ - ▁HAPP
663
+ - ▁DUH
664
+ - OO
665
+ - ▁JUH
666
+ - LE
667
+ - ▁HUH
668
+ - ▁BUH
669
+ - BOOK
670
+ - WENT
671
+ - ▁CA
672
+ - OSE
673
+ - EM
674
+ - IC
675
+ - AG
676
+ - ▁LETTERS
677
+ - IS
678
+ - EW
679
+ - ONG
680
+ - V
681
+ - AL
682
+ - PAY
683
+ - REE
684
+ - EE
685
+ - ▁TIMES
686
+ - ▁SPIN
687
+ - UR
688
+ - CU
689
+ - GER
690
+ - ▁TR
691
+ - ▁AW
692
+ - UGH
693
+ - UT
694
+ - ▁BL
695
+ - ▁SL
696
+ - ▁FORT
697
+ - ▁GE
698
+ - EA
699
+ - ▁TA
700
+ - GU
701
+ - ▁FINISH
702
+ - ▁UN
703
+ - READ
704
+ - THER
705
+ - DAY
706
+ - ▁BLA
707
+ - ▁ARTIST
708
+ - ▁BACKWARDS
709
+ - ▁DOCTOR
710
+ - ▁DREAMS
711
+ - ▁EXPLA
712
+ - ▁MIDDLE
713
+ - ▁MOUSE
714
+ - ▁PROB
715
+ - ▁RINSING
716
+ - ▁STRAIGHT
717
+ - ▁SUNFLOWER
718
+ - ▁TOOTHPICK
719
+ - ▁TWELVE
720
+ - ▁VULTURE
721
+ - ▁CONFUS
722
+ - TION
723
+ - ▁HOME
724
+ - ▁OPEN
725
+ - ▁SORRY
726
+ - ▁BORING
727
+ - ▁MINE
728
+ - ▁ENOUGH
729
+ - ▁HELLO
730
+ - ▁BORED
731
+ - RITE
732
+ - ▁TOWER
733
+ - ▁BUIL
734
+ - ▁ODD
735
+ - ▁UNP
736
+ - ▁APPLY
737
+ - ▁ANYMORE
738
+ - ▁FOUW
739
+ - APE
740
+ - OUNT
741
+ - ▁FIFT
742
+ - ▁ZEBRA
743
+ - ▁LION
744
+ - ▁BLAH
745
+ - ▁BLOCK
746
+ - ▁COP
747
+ - ▁HMM
748
+ - ▁ASK
749
+ - ▁BAB
750
+ - ▁DARKER
751
+ - ▁HEAR
752
+ - ▁CHO
753
+ - ▁CLOSE
754
+ - ▁JACK
755
+ - ▁FULL
756
+ - ▁CUP
757
+ - ▁WHE
758
+ - ▁IDEA
759
+ - ▁PIRATES
760
+ - ▁SPE
761
+ - ▁HEAD
762
+ - ▁GIVE
763
+ - ▁END
764
+ - DER
765
+ - ▁HAND
766
+ - ▁BOXES
767
+ - ▁BEST
768
+ - ▁LEARNING
769
+ - ▁MESS
770
+ - ▁MOST
771
+ - ▁FLA
772
+ - LIT
773
+ - ▁AC
774
+ - ▁AHW
775
+ - ▁FUH
776
+ - ▁LU
777
+ - ▁SSS
778
+ - OWN
779
+ - ▁PUH
780
+ - ▁PW
781
+ - POS
782
+ - ▁CIRCLES
783
+ - ENS
784
+ - LK
785
+ - ▁PLAYING
786
+ - AIL
787
+ - AP
788
+ - PIT
789
+ - NG
790
+ - ▁LETTU
791
+ - IK
792
+ - DDING
793
+ - HH
794
+ - PPER
795
+ - ▁GW
796
+ - ABL
797
+ - OL
798
+ - ▁KID
799
+ - DING
800
+ - ▁KA
801
+ - ERS
802
+ - ▁FI
803
+ - LIP
804
+ - ▁SE
805
+ - ▁TREES
806
+ - UN
807
+ - ▁RO
808
+ - ATE
809
+ - ND
810
+ - ▁FO
811
+ - ICE
812
+ - IF
813
+ - HW
814
+ - AY
815
+ - ▁BIGGE
816
+ - UST
817
+ - ▁DE
818
+ - ▁KI
819
+ - ▁LOS
820
+ - ▁THA
821
+ - ▁PAN
822
+ - IL
823
+ - MB
824
+ - ▁BOO
825
+ - SPE
826
+ - 'NO'
827
+ - ACK
828
+ - ▁FIN
829
+ - C
830
+ - ▁GROUP
831
+ - ▁GERM
832
+ - EAD
833
+ - ▁SOMETIME
834
+ - LZ
835
+ - IVE
836
+ - UP
837
+ - TWO
838
+ - HIRT
839
+ - HRO
840
+ - JELLYFISH
841
+ - ▁PAR
842
+ - PART
843
+ - IBO
844
+ - WHAT
845
+ - KEY
846
+ - FOUR
847
+ - AME
848
+ - ANGE
849
+ - EC
850
+ - TIME
851
+ - ▁REAL
852
+ - ELEPHANT
853
+ - ▁BATHROOM
854
+ - ▁BIVY
855
+ - ▁BRACES
856
+ - ▁FLOWER
857
+ - ▁GARFIELD
858
+ - ▁GARGLE
859
+ - ▁KOALA
860
+ - ▁PROBLEMS
861
+ - ▁SEVENEIGHTNINE
862
+ - ���STINKY
863
+ - ▁SWORD
864
+ - ▁UPPERCASE
865
+ - EMBER
866
+ - FUL
867
+ - ▁SEPARAT
868
+ - ▁BEFORE
869
+ - ▁BROKE
870
+ - ▁LOUD
871
+ - ▁MONSTER
872
+ - ▁MOUF
873
+ - ▁POOP
874
+ - ▁SHINNY
875
+ - ▁DRAW
876
+ - ▁MAILBOX
877
+ - ▁HUNGRY
878
+ - ▁BREAK
879
+ - ▁SARA
880
+ - ▁JOB
881
+ - ▁WATCH
882
+ - ▁SPARKL
883
+ - ▁SHORT
884
+ - ▁WEEK
885
+ - ▁BIRD
886
+ - ▁MOMMY
887
+ - ▁LOOSE
888
+ - ▁GREAT
889
+ - ▁PRETTIER
890
+ - ▁SMIL
891
+ - ▁FACE
892
+ - ▁HAV
893
+ - ▁PIECE
894
+ - ▁FUNNY
895
+ - ▁UNDER
896
+ - ▁SLOWER
897
+ - ACT
898
+ - ▁PLEA
899
+ - ▁VEHA
900
+ - ▁PEAR
901
+ - ▁FEEL
902
+ - ▁SPIDER
903
+ - ▁WORSE
904
+ - ▁SWI
905
+ - ▁AYE
906
+ - UNU
907
+ - ▁EVER
908
+ - ▁HOPE
909
+ - ▁SIGN
910
+ - AK
911
+ - UIZ
912
+ - ▁SOFT
913
+ - ▁POP
914
+ - ▁TEEH
915
+ - ▁DEH
916
+ - IBLE
917
+ - ▁SIDEWAYS
918
+ - ROT
919
+ - ▁ORDER
920
+ - ▁FINISHED
921
+ - ▁JELLYFISH
922
+ - ▁FELL
923
+ - KEU
924
+ - ▁IMPO
925
+ - HEAD
926
+ - UM
927
+ - ▁PRESS
928
+ - ▁SECONDS
929
+ - ▁LEA
930
+ - ▁MOLD
931
+ - LLUH
932
+ - ▁READ
933
+ - ▁ONETWO
934
+ - ▁LINE
935
+ - FE
936
+ - ▁FOH
937
+ - ▁HOT
938
+ - ▁FOU
939
+ - ▁MOH
940
+ - ▁DEN
941
+ - ▁WIN
942
+ - ▁NINETY
943
+ - IRTY
944
+ - ▁TWEE
945
+ - OUR
946
+ - IRED
947
+ - TLE
948
+ - ▁HEH
949
+ - ▁JU
950
+ - PASTE
951
+ - ▁FEVER
952
+ - ▁WR
953
+ - ▁PAI
954
+ - MINT
955
+ - TEEN
956
+ - ▁WASHING
957
+ - ▁BI
958
+ - ▁NAH
959
+ - DY
960
+ - ▁RA
961
+ - ▁DA
962
+ - AHW
963
+ - ▁YUH
964
+ - ULL
965
+ - ▁WL
966
+ - UHTY
967
+ - ▁SHO
968
+ - ▁CUH
969
+ - ASTE
970
+ - OOD
971
+ - ▁LAM
972
+ - ▁CI
973
+ - OLD
974
+ - UNN
975
+ - NUH
976
+ - OCK
977
+ - US
978
+ - ▁SM
979
+ - MPLE
980
+ - ▁HIT
981
+ - ▁THRO
982
+ - ▁DEU
983
+ - HOLE
984
+ - ▁THINKING
985
+ - UBB
986
+ - ▁FU
987
+ - ▁PI
988
+ - ▁SMO
989
+ - ▁VO
990
+ - AN
991
+ - UG
992
+ - ▁HM
993
+ - UE
994
+ - GLE
995
+ - ▁MOV
996
+ - LI
997
+ - ▁BLU
998
+ - PORT
999
+ - ▁WED
1000
+ - ▁TRI
1001
+ - ▁CHE
1002
+ - CA
1003
+ - ▁SC
1004
+ - ▁STO
1005
+ - ▁BED
1006
+ - ▁TELLS
1007
+ - ▁MI
1008
+ - OR
1009
+ - TTER
1010
+ - NES
1011
+ - OUGH
1012
+ - ▁AR
1013
+ - ROW
1014
+ - UA
1015
+ - AB
1016
+ - IG
1017
+ - OF
1018
+ - MAN
1019
+ - RK
1020
+ - OUN
1021
+ - ROUGH
1022
+ - LUH
1023
+ - DENT
1024
+ - ▁PIE
1025
+ - LAP
1026
+ - KUH
1027
+ - OT
1028
+ - RSE
1029
+ - ▁LA
1030
+ - ▁PAST
1031
+ - ▁ANOTH
1032
+ - OP
1033
+ - EP
1034
+ - ▁LATE
1035
+ - AM
1036
+ - LU
1037
+ - ▁WOO
1038
+ - HUH
1039
+ - ▁CER
1040
+ - OU
1041
+ - IPP
1042
+ - ▁CO
1043
+ - EH
1044
+ - TE
1045
+ - WHERE
1046
+ - ASH
1047
+ - PPY
1048
+ - WAY
1049
+ - RO
1050
+ - SHE
1051
+ - OST
1052
+ - AIN
1053
+ - ▁SECOND
1054
+ - ▁PIRATE
1055
+ - ▁MINUTE
1056
+ - ABET
1057
+ - ▁DIFFEREN
1058
+ - BE
1059
+ - IGH
1060
+ - ▁COO
1061
+ - ▁WHA
1062
+ - BRUSH
1063
+ - TRA
1064
+ - ▁PRES
1065
+ - ▁TRYIN
1066
+ - ▁GIV
1067
+ - OPE
1068
+ - SHIN
1069
+ - STRO
1070
+ - SIGN
1071
+ - ▁PLU
1072
+ - ZEBRA
1073
+ - LION
1074
+ - CROCODILE
1075
+ - LATE
1076
+ - UF
1077
+ - EQUALS
1078
+ - COME
1079
+ - UBE
1080
+ - J
1081
+ - TOGETHER
1082
+ - MAYBE
1083
+ - BOX
1084
+ - CLEAN
1085
+ - THEY
1086
+ - JIBO
1087
+ - EASY
1088
+ - MOUTH
1089
+ - ▁TALK
1090
+ - ▁SKI
1091
+ - WHY
1092
+ - WICE
1093
+ - CAUSE
1094
+ - UMP
1095
+ - TRIC
1096
+ - ▁CLOS
1097
+ - ▁SEVE
1098
+ - ▁DIRT
1099
+ - ▁NUMB
1100
+ - YOU
1101
+ - PRI
1102
+ - ▁JIB
1103
+ - ETTIER
1104
+ - FFERENT
1105
+ - ERCASE
1106
+ - ROOM
1107
+ - ▁DIFF
1108
+ - ▁JELLY
1109
+ - ▁SEVENEIGHT
1110
+ - ORGE
1111
+ - ▁YELL
1112
+ - DRA
1113
+ - ▁SLOW
1114
+ - ▁MON
1115
+ - ▁BUG
1116
+ - YPE
1117
+ - ▁BRU
1118
+ - COL
1119
+ - PUS
1120
+ - WO
1121
+ - INET
1122
+ - NGRY
1123
+ - BRUSHING
1124
+ - ▁CUB
1125
+ - OCTO
1126
+ - HIC
1127
+ - UDE
1128
+ - RUB
1129
+ - MOR
1130
+ - LOCK
1131
+ - ▁BR
1132
+ - YOUR
1133
+ - ▁STR
1134
+ - ▁KNE
1135
+ - ▁CRO
1136
+ - ▁BO
1137
+ - UALLY
1138
+ - ▁TOOTHB
1139
+ - ▁ANYMO
1140
+ - UKU
1141
+ - ▁GUE
1142
+ - MA
1143
+ - ENTY
1144
+ - PHA
1145
+ - ▁QUE
1146
+ - PF
1147
+ - KE
1148
+ - NOW
1149
+ - ▁LAS
1150
+ - ▁SHIN
1151
+ - ARN
1152
+ - GE
1153
+ - ▁MAIL
1154
+ - RUSHING
1155
+ - Q
1156
+ - <sos/eos>
1157
+ init: null
1158
+ input_size: null
1159
+ ctc_conf:
1160
+ dropout_rate: 0.0
1161
+ ctc_type: builtin
1162
+ reduce: true
1163
+ ignore_nan_grad: null
1164
+ zero_infinity: true
1165
+ brctc_risk_strategy: exp
1166
+ brctc_group_strategy: end
1167
+ brctc_risk_factor: 0.0
1168
+ joint_net_conf: null
1169
+ use_preprocessor: true
1170
+ use_lang_prompt: false
1171
+ use_nlp_prompt: false
1172
+ token_type: bpe
1173
+ bpemodel: data/en_token_list/bpe_unigram1024/bpe.model
1174
+ non_linguistic_symbols: null
1175
+ cleaner: null
1176
+ g2p: null
1177
+ speech_volume_normalize: null
1178
+ rir_scp: null
1179
+ rir_apply_prob: 1.0
1180
+ noise_scp: null
1181
+ noise_apply_prob: 1.0
1182
+ noise_db_range: '13_15'
1183
+ short_noise_thres: 0.5
1184
+ aux_ctc_tasks: []
1185
+ frontend: s3prl
1186
+ frontend_conf:
1187
+ frontend_conf:
1188
+ upstream: wavlm_large
1189
+ download_dir: ./hub
1190
+ multilayer_feature: true
1191
+ fs: 16k
1192
+ specaug: specaug
1193
+ specaug_conf:
1194
+ apply_time_warp: true
1195
+ time_warp_window: 5
1196
+ time_warp_mode: bicubic
1197
+ apply_freq_mask: true
1198
+ freq_mask_width_range:
1199
+ - 0
1200
+ - 27
1201
+ num_freq_mask: 2
1202
+ apply_time_mask: true
1203
+ time_mask_width_ratio_range:
1204
+ - 0.0
1205
+ - 0.05
1206
+ num_time_mask: 5
1207
+ normalize: utterance_mvn
1208
+ normalize_conf: {}
1209
+ model: espnet
1210
+ model_conf:
1211
+ ctc_weight: 0.3
1212
+ lsm_weight: 0.1
1213
+ length_normalized_loss: false
1214
+ extract_feats_in_collect_stats: false
1215
+ preencoder: linear
1216
+ preencoder_conf:
1217
+ input_size: 1024
1218
+ output_size: 80
1219
+ encoder: transformer
1220
+ encoder_conf:
1221
+ output_size: 256
1222
+ attention_heads: 4
1223
+ linear_units: 1024
1224
+ num_blocks: 18
1225
+ dropout_rate: 0.1
1226
+ positional_dropout_rate: 0.1
1227
+ attention_dropout_rate: 0.1
1228
+ input_layer: conv2d2
1229
+ normalize_before: true
1230
+ postencoder: null
1231
+ postencoder_conf: {}
1232
+ decoder: transformer
1233
+ decoder_conf:
1234
+ attention_heads: 4
1235
+ linear_units: 2048
1236
+ num_blocks: 6
1237
+ dropout_rate: 0.1
1238
+ positional_dropout_rate: 0.1
1239
+ self_attention_dropout_rate: 0.1
1240
+ src_attention_dropout_rate: 0.1
1241
+ preprocessor: default
1242
+ preprocessor_conf: {}
1243
+ required:
1244
+ - output_dir
1245
+ - token_list
1246
+ version: '202402'
1247
+ distributed: false
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/acc.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/backward_time.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/cer.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/cer_ctc.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/clip.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/forward_time.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/gpu_max_cached_mem_GB.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/grad_norm.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/iter_time.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_att.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_ctc.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/loss_scale.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/optim0_lr0.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/optim_step_time.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/train_time.png ADDED
exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/images/wer.png ADDED
meta.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ espnet: '202402'
2
+ files:
3
+ asr_model_file: exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/86epoch.pth
4
+ python: "3.9.19 (main, May 6 2024, 19:43:03) \n[GCC 11.2.0]"
5
+ timestamp: 1738236401.482846
6
+ torch: 2.4.0
7
+ yaml_files:
8
+ asr_train_config: exp/asr_train_asr_wavlm_transformer_raw_en_bpe1024/config.yaml