artitsu commited on
Commit
0ac24d9
1 Parent(s): 39e7f9d

Upload 2 files

Browse files
thai-central/config.yaml ADDED
@@ -0,0 +1,235 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config: conf/train_asr_hubert.yaml
2
+ print_config: false
3
+ log_level: INFO
4
+ dry_run: false
5
+ iterator_type: sequence
6
+ output_dir: exp/asr_hubert_pt
7
+ ngpu: 1
8
+ seed: 0
9
+ num_workers: 1
10
+ num_att_plot: 3
11
+ dist_backend: nccl
12
+ dist_init_method: env://
13
+ dist_world_size: 2
14
+ dist_rank: 0
15
+ local_rank: 0
16
+ dist_master_addr: localhost
17
+ dist_master_port: null
18
+ dist_launcher: null
19
+ multiprocessing_distributed: true
20
+ unused_parameters: true
21
+ sharded_ddp: false
22
+ cudnn_enabled: true
23
+ cudnn_benchmark: false
24
+ cudnn_deterministic: true
25
+ collect_stats: false
26
+ write_collected_feats: false
27
+ max_epoch: 20
28
+ patience: null
29
+ val_scheduler_criterion:
30
+ - valid
31
+ - loss
32
+ early_stopping_criterion:
33
+ - valid
34
+ - loss
35
+ - min
36
+ best_model_criterion:
37
+ - - valid
38
+ - acc
39
+ - max
40
+ keep_nbest_models: 10
41
+ nbest_averaging_interval: 0
42
+ grad_clip: 5.0
43
+ grad_clip_type: 2.0
44
+ grad_noise: false
45
+ accum_grad: 2
46
+ no_forward_run: false
47
+ resume: true
48
+ train_dtype: float32
49
+ use_amp: false
50
+ log_interval: null
51
+ use_matplotlib: true
52
+ use_tensorboard: true
53
+ use_wandb: false
54
+ wandb_project: null
55
+ wandb_id: null
56
+ wandb_entity: null
57
+ wandb_name: null
58
+ wandb_model_log_interval: -1
59
+ detect_anomaly: false
60
+ pretrain_path: null
61
+ init_param: []
62
+ ignore_init_mismatch: true
63
+ freeze_param: []
64
+ num_iters_per_epoch: 10000
65
+ batch_size: 20
66
+ valid_batch_size: null
67
+ batch_bins: 9000000
68
+ valid_batch_bins: null
69
+ train_shape_file:
70
+ - exp/asr_stats_raw_char_sp/train/speech_shape
71
+ - exp/asr_stats_raw_char_sp/train/text_shape.char
72
+ valid_shape_file:
73
+ - exp/asr_stats_raw_char_sp/valid/speech_shape
74
+ - exp/asr_stats_raw_char_sp/valid/text_shape.char
75
+ batch_type: numel
76
+ valid_batch_type: null
77
+ fold_length:
78
+ - 80000
79
+ - 150
80
+ sort_in_batch: descending
81
+ sort_batch: descending
82
+ multiple_iterator: false
83
+ chunk_length: 500
84
+ chunk_shift_ratio: 0.5
85
+ num_cache_chunks: 1024
86
+ train_data_path_and_name_and_type:
87
+ - - dump/raw/cheat_sp/wav.scp
88
+ - speech
89
+ - sound
90
+ - - dump/raw/cheat_sp/text
91
+ - text
92
+ - text
93
+ valid_data_path_and_name_and_type:
94
+ - - dump/raw/test/wav.scp
95
+ - speech
96
+ - sound
97
+ - - dump/raw/test/text
98
+ - text
99
+ - text
100
+ allow_variable_data_keys: false
101
+ max_cache_size: 0.0
102
+ max_cache_fd: 32
103
+ valid_max_cache_size: null
104
+ optim: adam
105
+ optim_conf:
106
+ lr: 0.0005
107
+ weight_decay: 1.0e-06
108
+ scheduler: warmuplr
109
+ scheduler_conf:
110
+ warmup_steps: 5000
111
+ token_list:
112
+ - <blank>
113
+ - <unk>
114
+ - <space>
115
+ - า
116
+ - น
117
+ - ่
118
+ - อ
119
+ - ร
120
+ - ้
121
+ - ง
122
+ - ก
123
+ - เ
124
+ - ม
125
+ - ย
126
+ - ั
127
+ - ล
128
+ - ว
129
+ - ี
130
+ - ห
131
+ - ด
132
+ - บ
133
+ - ส
134
+ - ะ
135
+ - ท
136
+ - ิ
137
+ - ค
138
+ - ต
139
+ - ป
140
+ - ข
141
+ - แ
142
+ - จ
143
+ - ไ
144
+ - พ
145
+ - ื
146
+ - ุ
147
+ - ู
148
+ - ็
149
+ - ใ
150
+ - ช
151
+ - ์
152
+ - ำ
153
+ - ึ
154
+ - โ
155
+ - ผ
156
+ - ถ
157
+ - ซ
158
+ - ณ
159
+ - ศ
160
+ - ฟ
161
+ - ญ
162
+ - ธ
163
+ - ภ
164
+ - ษ
165
+ - ฉ
166
+ - ๊
167
+ - ฝ
168
+ - ฮ
169
+ - ๋
170
+ - ฐ
171
+ - ฤ
172
+ - ฏ
173
+ - ฎ
174
+ - ฆ
175
+ - ฒ
176
+ - ฑ
177
+ - ฬ
178
+ - ฌ
179
+ - ฯ
180
+ - <sos/eos>
181
+ init: null
182
+ input_size: 1
183
+ ctc_conf:
184
+ dropout_rate: 0.0
185
+ ctc_type: builtin
186
+ reduce: true
187
+ ignore_nan_grad: true
188
+ joint_net_conf: null
189
+ use_preprocessor: true
190
+ token_type: char
191
+ bpemodel: null
192
+ non_linguistic_symbols: null
193
+ cleaner: null
194
+ g2p: null
195
+ speech_volume_normalize: null
196
+ rir_scp: null
197
+ rir_apply_prob: 1.0
198
+ noise_scp: null
199
+ noise_apply_prob: 1.0
200
+ noise_db_range: '13_15'
201
+ frontend: null
202
+ frontend_conf: {}
203
+ specaug: null
204
+ specaug_conf: {}
205
+ normalize: utterance_mvn
206
+ normalize_conf: {}
207
+ model: espnet
208
+ model_conf:
209
+ ctc_weight: 0.3
210
+ lsm_weight: 0.1
211
+ length_normalized_loss: false
212
+ preencoder: null
213
+ preencoder_conf: {}
214
+ encoder: hubert
215
+ encoder_conf:
216
+ output_size: 256
217
+ normalize_before: false
218
+ hubert_url: https://dl.fbaipublicfiles.com/hubert/hubert_base_ls960.pt
219
+ hubert_dir_path: ./downloads/hubert_pretrained_models/hubert_base_ls960.pt
220
+ postencoder: null
221
+ postencoder_conf: {}
222
+ decoder: transformer
223
+ decoder_conf:
224
+ attention_heads: 4
225
+ linear_units: 2048
226
+ num_blocks: 6
227
+ dropout_rate: 0.1
228
+ positional_dropout_rate: 0.1
229
+ self_attention_dropout_rate: 0.1
230
+ src_attention_dropout_rate: 0.1
231
+ required:
232
+ - output_dir
233
+ - token_list
234
+ version: 0.10.7a1
235
+ distributed: true
thai-central/valid.acc.ave_10best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:821b1a8dc73ceb4f6dc004d0217cc2f6dd19c6705f206f80798de12a3eb421e3
3
+ size 417827465