Hugging Face's logo

marcel
/
wav2vec2-large-xlsr-german-demo

Automatic Speech Recognition
Transformers
PyTorch
JAX
German
wav2vec2
audio
speech
xlsr-fine-tuning-week
Eval Results
Inference Endpoints
Model card Files Community
marcel commited on
Commit
1b3dc8c
•
1 Parent(s): b82f29b

Add model files

Browse files
Files changed (18) hide show
  1. checkpoint-6800/config.json +76 -0
  2. checkpoint-6800/optimizer.pt +3 -0
  3. checkpoint-6800/preprocessor_config.json +8 -0
  4. checkpoint-6800/pytorch_model.bin +3 -0
  5. checkpoint-6800/scheduler.pt +3 -0
  6. checkpoint-6800/trainer_state.json +254 -0
  7. checkpoint-6800/training_args.bin +3 -0
  8. checkpoint-7200/config.json +76 -0
  9. checkpoint-7200/optimizer.pt +3 -0
  10. checkpoint-7200/preprocessor_config.json +8 -0
  11. checkpoint-7200/pytorch_model.bin +3 -0
  12. checkpoint-7200/scheduler.pt +3 -0
  13. checkpoint-7200/trainer_state.json +268 -0
  14. checkpoint-7200/training_args.bin +3 -0
  15. preprocessor_config.json +8 -0
  16. special_tokens_map.json +1 -0
  17. tokenizer_config.json +1 -0
  18. vocab.json +1 -0
checkpoint-6800/config.json ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
3
+ "activation_dropout": 0.0,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "Wav2Vec2ForCTC"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "conv_bias": true,
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
+ "ctc_loss_reduction": "mean",
39
+ "ctc_zero_infinity": false,
40
+ "do_stable_layer_norm": true,
41
+ "eos_token_id": 2,
42
+ "feat_extract_activation": "gelu",
43
+ "feat_extract_dropout": 0.0,
44
+ "feat_extract_norm": "layer",
45
+ "feat_proj_dropout": 0.0,
46
+ "final_dropout": 0.0,
47
+ "gradient_checkpointing": true,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 1024,
51
+ "initializer_range": 0.02,
52
+ "intermediate_size": 4096,
53
+ "layer_norm_eps": 1e-05,
54
+ "layerdrop": 0.1,
55
+ "mask_channel_length": 10,
56
+ "mask_channel_min_space": 1,
57
+ "mask_channel_other": 0.0,
58
+ "mask_channel_prob": 0.0,
59
+ "mask_channel_selection": "static",
60
+ "mask_feature_length": 10,
61
+ "mask_feature_prob": 0.0,
62
+ "mask_time_length": 10,
63
+ "mask_time_min_space": 1,
64
+ "mask_time_other": 0.0,
65
+ "mask_time_prob": 0.05,
66
+ "mask_time_selection": "static",
67
+ "model_type": "wav2vec2",
68
+ "num_attention_heads": 16,
69
+ "num_conv_pos_embedding_groups": 16,
70
+ "num_conv_pos_embeddings": 128,
71
+ "num_feat_extract_layers": 7,
72
+ "num_hidden_layers": 24,
73
+ "pad_token_id": 35,
74
+ "transformers_version": "4.5.0.dev0",
75
+ "vocab_size": 36
76
+ }
checkpoint-6800/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:546395b46d0d5c6b7fd31058167581989b47627fba654d7e57038345922ae1ee
3
+ size 2490372359
checkpoint-6800/preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": true,
7
+ "sampling_rate": 16000
8
+ }
checkpoint-6800/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e925627ecc7c3375a088da44d3b3ffab15fd4cc8481bc6adaec9c2b9ee59b4c4
3
+ size 1262081431
checkpoint-6800/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4516191e0bac48e8193846dbc65d5025657680877d832316f6480ebd3c516831
3
+ size 623
checkpoint-6800/trainer_state.json ADDED
@@ -0,0 +1,254 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 27.642276422764226,
5
+ "global_step": 6800,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.63,
12
+ "learning_rate": 0.00023999999999999998,
13
+ "loss": 5.7155,
14
+ "step": 400
15
+ },
16
+ {
17
+ "epoch": 1.63,
18
+ "eval_loss": 3.039255142211914,
19
+ "eval_runtime": 56.5805,
20
+ "eval_samples_per_second": 8.271,
21
+ "eval_wer": 1.0,
22
+ "step": 400
23
+ },
24
+ {
25
+ "epoch": 3.25,
26
+ "learning_rate": 0.0002869186046511628,
27
+ "loss": 1.6883,
28
+ "step": 800
29
+ },
30
+ {
31
+ "epoch": 3.25,
32
+ "eval_loss": 0.7244638800621033,
33
+ "eval_runtime": 58.2783,
34
+ "eval_samples_per_second": 8.03,
35
+ "eval_wer": 0.5629911280101394,
36
+ "step": 800
37
+ },
38
+ {
39
+ "epoch": 4.88,
40
+ "learning_rate": 0.0002694767441860465,
41
+ "loss": 0.3776,
42
+ "step": 1200
43
+ },
44
+ {
45
+ "epoch": 4.88,
46
+ "eval_loss": 0.6811308264732361,
47
+ "eval_runtime": 57.5618,
48
+ "eval_samples_per_second": 8.13,
49
+ "eval_wer": 0.49150823827629914,
50
+ "step": 1200
51
+ },
52
+ {
53
+ "epoch": 6.5,
54
+ "learning_rate": 0.00025203488372093025,
55
+ "loss": 0.2425,
56
+ "step": 1600
57
+ },
58
+ {
59
+ "epoch": 6.5,
60
+ "eval_loss": 0.629974365234375,
61
+ "eval_runtime": 57.7361,
62
+ "eval_samples_per_second": 8.106,
63
+ "eval_wer": 0.45449936628643856,
64
+ "step": 1600
65
+ },
66
+ {
67
+ "epoch": 8.13,
68
+ "learning_rate": 0.00023459302325581392,
69
+ "loss": 0.1872,
70
+ "step": 2000
71
+ },
72
+ {
73
+ "epoch": 8.13,
74
+ "eval_loss": 0.7140640616416931,
75
+ "eval_runtime": 58.4584,
76
+ "eval_samples_per_second": 8.006,
77
+ "eval_wer": 0.44283903675538655,
78
+ "step": 2000
79
+ },
80
+ {
81
+ "epoch": 9.76,
82
+ "learning_rate": 0.00021715116279069765,
83
+ "loss": 0.1565,
84
+ "step": 2400
85
+ },
86
+ {
87
+ "epoch": 9.76,
88
+ "eval_loss": 0.7015160322189331,
89
+ "eval_runtime": 57.3174,
90
+ "eval_samples_per_second": 8.165,
91
+ "eval_wer": 0.4385297845373891,
92
+ "step": 2400
93
+ },
94
+ {
95
+ "epoch": 11.38,
96
+ "learning_rate": 0.00019970930232558135,
97
+ "loss": 0.1232,
98
+ "step": 2800
99
+ },
100
+ {
101
+ "epoch": 11.38,
102
+ "eval_loss": 0.7360302805900574,
103
+ "eval_runtime": 57.5205,
104
+ "eval_samples_per_second": 8.136,
105
+ "eval_wer": 0.4326996197718631,
106
+ "step": 2800
107
+ },
108
+ {
109
+ "epoch": 13.01,
110
+ "learning_rate": 0.00018226744186046508,
111
+ "loss": 0.1131,
112
+ "step": 3200
113
+ },
114
+ {
115
+ "epoch": 13.01,
116
+ "eval_loss": 0.921540379524231,
117
+ "eval_runtime": 57.6005,
118
+ "eval_samples_per_second": 8.125,
119
+ "eval_wer": 0.44714828897338404,
120
+ "step": 3200
121
+ },
122
+ {
123
+ "epoch": 14.63,
124
+ "learning_rate": 0.0001648255813953488,
125
+ "loss": 0.1004,
126
+ "step": 3600
127
+ },
128
+ {
129
+ "epoch": 14.63,
130
+ "eval_loss": 0.8807307481765747,
131
+ "eval_runtime": 57.8571,
132
+ "eval_samples_per_second": 8.089,
133
+ "eval_wer": 0.4187579214195184,
134
+ "step": 3600
135
+ },
136
+ {
137
+ "epoch": 16.26,
138
+ "learning_rate": 0.00014738372093023254,
139
+ "loss": 0.0851,
140
+ "step": 4000
141
+ },
142
+ {
143
+ "epoch": 16.26,
144
+ "eval_loss": 0.7986534237861633,
145
+ "eval_runtime": 57.6046,
146
+ "eval_samples_per_second": 8.124,
147
+ "eval_wer": 0.43067173637515843,
148
+ "step": 4000
149
+ },
150
+ {
151
+ "epoch": 17.89,
152
+ "learning_rate": 0.00012994186046511627,
153
+ "loss": 0.0763,
154
+ "step": 4400
155
+ },
156
+ {
157
+ "epoch": 17.89,
158
+ "eval_loss": 0.8154810667037964,
159
+ "eval_runtime": 57.7278,
160
+ "eval_samples_per_second": 8.107,
161
+ "eval_wer": 0.41698352344740175,
162
+ "step": 4400
163
+ },
164
+ {
165
+ "epoch": 19.51,
166
+ "learning_rate": 0.0001125,
167
+ "loss": 0.0664,
168
+ "step": 4800
169
+ },
170
+ {
171
+ "epoch": 19.51,
172
+ "eval_loss": 0.8099461197853088,
173
+ "eval_runtime": 57.6128,
174
+ "eval_samples_per_second": 8.123,
175
+ "eval_wer": 0.4124207858048162,
176
+ "step": 4800
177
+ },
178
+ {
179
+ "epoch": 21.14,
180
+ "learning_rate": 9.505813953488371e-05,
181
+ "loss": 0.0611,
182
+ "step": 5200
183
+ },
184
+ {
185
+ "epoch": 21.14,
186
+ "eval_loss": 0.8145996928215027,
187
+ "eval_runtime": 57.1919,
188
+ "eval_samples_per_second": 8.183,
189
+ "eval_wer": 0.4114068441064639,
190
+ "step": 5200
191
+ },
192
+ {
193
+ "epoch": 22.76,
194
+ "learning_rate": 7.761627906976744e-05,
195
+ "loss": 0.0573,
196
+ "step": 5600
197
+ },
198
+ {
199
+ "epoch": 22.76,
200
+ "eval_loss": 0.8513393402099609,
201
+ "eval_runtime": 58.2107,
202
+ "eval_samples_per_second": 8.04,
203
+ "eval_wer": 0.4126742712294043,
204
+ "step": 5600
205
+ },
206
+ {
207
+ "epoch": 24.39,
208
+ "learning_rate": 6.0174418604651156e-05,
209
+ "loss": 0.0531,
210
+ "step": 6000
211
+ },
212
+ {
213
+ "epoch": 24.39,
214
+ "eval_loss": 0.8830446600914001,
215
+ "eval_runtime": 57.6531,
216
+ "eval_samples_per_second": 8.118,
217
+ "eval_wer": 0.4050697084917617,
218
+ "step": 6000
219
+ },
220
+ {
221
+ "epoch": 26.02,
222
+ "learning_rate": 4.273255813953488e-05,
223
+ "loss": 0.0498,
224
+ "step": 6400
225
+ },
226
+ {
227
+ "epoch": 26.02,
228
+ "eval_loss": 0.8851639628410339,
229
+ "eval_runtime": 57.3676,
230
+ "eval_samples_per_second": 8.158,
231
+ "eval_wer": 0.40481622306717363,
232
+ "step": 6400
233
+ },
234
+ {
235
+ "epoch": 27.64,
236
+ "learning_rate": 2.52906976744186e-05,
237
+ "loss": 0.0444,
238
+ "step": 6800
239
+ },
240
+ {
241
+ "epoch": 27.64,
242
+ "eval_loss": 0.936349093914032,
243
+ "eval_runtime": 57.896,
244
+ "eval_samples_per_second": 8.083,
245
+ "eval_wer": 0.40887198986058304,
246
+ "step": 6800
247
+ }
248
+ ],
249
+ "max_steps": 7380,
250
+ "num_train_epochs": 30,
251
+ "total_flos": 3.795659162347106e+19,
252
+ "trial_name": null,
253
+ "trial_params": null
254
+ }
checkpoint-6800/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28a85071638bb9d7d3f4b39303e2c56cdd6ea61baa2df1558d1e42576140f653
3
+ size 2351
checkpoint-7200/config.json ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
3
+ "activation_dropout": 0.0,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "Wav2Vec2ForCTC"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "conv_bias": true,
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
+ "ctc_loss_reduction": "mean",
39
+ "ctc_zero_infinity": false,
40
+ "do_stable_layer_norm": true,
41
+ "eos_token_id": 2,
42
+ "feat_extract_activation": "gelu",
43
+ "feat_extract_dropout": 0.0,
44
+ "feat_extract_norm": "layer",
45
+ "feat_proj_dropout": 0.0,
46
+ "final_dropout": 0.0,
47
+ "gradient_checkpointing": true,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 1024,
51
+ "initializer_range": 0.02,
52
+ "intermediate_size": 4096,
53
+ "layer_norm_eps": 1e-05,
54
+ "layerdrop": 0.1,
55
+ "mask_channel_length": 10,
56
+ "mask_channel_min_space": 1,
57
+ "mask_channel_other": 0.0,
58
+ "mask_channel_prob": 0.0,
59
+ "mask_channel_selection": "static",
60
+ "mask_feature_length": 10,
61
+ "mask_feature_prob": 0.0,
62
+ "mask_time_length": 10,
63
+ "mask_time_min_space": 1,
64
+ "mask_time_other": 0.0,
65
+ "mask_time_prob": 0.05,
66
+ "mask_time_selection": "static",
67
+ "model_type": "wav2vec2",
68
+ "num_attention_heads": 16,
69
+ "num_conv_pos_embedding_groups": 16,
70
+ "num_conv_pos_embeddings": 128,
71
+ "num_feat_extract_layers": 7,
72
+ "num_hidden_layers": 24,
73
+ "pad_token_id": 35,
74
+ "transformers_version": "4.5.0.dev0",
75
+ "vocab_size": 36
76
+ }
checkpoint-7200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5611e1ddb82cae93ff1c507d79a084ff9bdce32ded98cfba1c18075766b1a0a8
3
+ size 2490372359
checkpoint-7200/preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": true,
7
+ "sampling_rate": 16000
8
+ }
checkpoint-7200/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e94b7c454e44c0c3a87e18e88184ed7244669ee53c751523dbf5379ed6a9093c
3
+ size 1262081431
checkpoint-7200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2836df60d721674b336f35a8ff7cd9b9b8820a755354213b45e45fecb05bf80
3
+ size 623
checkpoint-7200/trainer_state.json ADDED
@@ -0,0 +1,268 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 29.26829268292683,
5
+ "global_step": 7200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.63,
12
+ "learning_rate": 0.00023999999999999998,
13
+ "loss": 5.7155,
14
+ "step": 400
15
+ },
16
+ {
17
+ "epoch": 1.63,
18
+ "eval_loss": 3.039255142211914,
19
+ "eval_runtime": 56.5805,
20
+ "eval_samples_per_second": 8.271,
21
+ "eval_wer": 1.0,
22
+ "step": 400
23
+ },
24
+ {
25
+ "epoch": 3.25,
26
+ "learning_rate": 0.0002869186046511628,
27
+ "loss": 1.6883,
28
+ "step": 800
29
+ },
30
+ {
31
+ "epoch": 3.25,
32
+ "eval_loss": 0.7244638800621033,
33
+ "eval_runtime": 58.2783,
34
+ "eval_samples_per_second": 8.03,
35
+ "eval_wer": 0.5629911280101394,
36
+ "step": 800
37
+ },
38
+ {
39
+ "epoch": 4.88,
40
+ "learning_rate": 0.0002694767441860465,
41
+ "loss": 0.3776,
42
+ "step": 1200
43
+ },
44
+ {
45
+ "epoch": 4.88,
46
+ "eval_loss": 0.6811308264732361,
47
+ "eval_runtime": 57.5618,
48
+ "eval_samples_per_second": 8.13,
49
+ "eval_wer": 0.49150823827629914,
50
+ "step": 1200
51
+ },
52
+ {
53
+ "epoch": 6.5,
54
+ "learning_rate": 0.00025203488372093025,
55
+ "loss": 0.2425,
56
+ "step": 1600
57
+ },
58
+ {
59
+ "epoch": 6.5,
60
+ "eval_loss": 0.629974365234375,
61
+ "eval_runtime": 57.7361,
62
+ "eval_samples_per_second": 8.106,
63
+ "eval_wer": 0.45449936628643856,
64
+ "step": 1600
65
+ },
66
+ {
67
+ "epoch": 8.13,
68
+ "learning_rate": 0.00023459302325581392,
69
+ "loss": 0.1872,
70
+ "step": 2000
71
+ },
72
+ {
73
+ "epoch": 8.13,
74
+ "eval_loss": 0.7140640616416931,
75
+ "eval_runtime": 58.4584,
76
+ "eval_samples_per_second": 8.006,
77
+ "eval_wer": 0.44283903675538655,
78
+ "step": 2000
79
+ },
80
+ {
81
+ "epoch": 9.76,
82
+ "learning_rate": 0.00021715116279069765,
83
+ "loss": 0.1565,
84
+ "step": 2400
85
+ },
86
+ {
87
+ "epoch": 9.76,
88
+ "eval_loss": 0.7015160322189331,
89
+ "eval_runtime": 57.3174,
90
+ "eval_samples_per_second": 8.165,
91
+ "eval_wer": 0.4385297845373891,
92
+ "step": 2400
93
+ },
94
+ {
95
+ "epoch": 11.38,
96
+ "learning_rate": 0.00019970930232558135,
97
+ "loss": 0.1232,
98
+ "step": 2800
99
+ },
100
+ {
101
+ "epoch": 11.38,
102
+ "eval_loss": 0.7360302805900574,
103
+ "eval_runtime": 57.5205,
104
+ "eval_samples_per_second": 8.136,
105
+ "eval_wer": 0.4326996197718631,
106
+ "step": 2800
107
+ },
108
+ {
109
+ "epoch": 13.01,
110
+ "learning_rate": 0.00018226744186046508,
111
+ "loss": 0.1131,
112
+ "step": 3200
113
+ },
114
+ {
115
+ "epoch": 13.01,
116
+ "eval_loss": 0.921540379524231,
117
+ "eval_runtime": 57.6005,
118
+ "eval_samples_per_second": 8.125,
119
+ "eval_wer": 0.44714828897338404,
120
+ "step": 3200
121
+ },
122
+ {
123
+ "epoch": 14.63,
124
+ "learning_rate": 0.0001648255813953488,
125
+ "loss": 0.1004,
126
+ "step": 3600
127
+ },
128
+ {
129
+ "epoch": 14.63,
130
+ "eval_loss": 0.8807307481765747,
131
+ "eval_runtime": 57.8571,
132
+ "eval_samples_per_second": 8.089,
133
+ "eval_wer": 0.4187579214195184,
134
+ "step": 3600
135
+ },
136
+ {
137
+ "epoch": 16.26,
138
+ "learning_rate": 0.00014738372093023254,
139
+ "loss": 0.0851,
140
+ "step": 4000
141
+ },
142
+ {
143
+ "epoch": 16.26,
144
+ "eval_loss": 0.7986534237861633,
145
+ "eval_runtime": 57.6046,
146
+ "eval_samples_per_second": 8.124,
147
+ "eval_wer": 0.43067173637515843,
148
+ "step": 4000
149
+ },
150
+ {
151
+ "epoch": 17.89,
152
+ "learning_rate": 0.00012994186046511627,
153
+ "loss": 0.0763,
154
+ "step": 4400
155
+ },
156
+ {
157
+ "epoch": 17.89,
158
+ "eval_loss": 0.8154810667037964,
159
+ "eval_runtime": 57.7278,
160
+ "eval_samples_per_second": 8.107,
161
+ "eval_wer": 0.41698352344740175,
162
+ "step": 4400
163
+ },
164
+ {
165
+ "epoch": 19.51,
166
+ "learning_rate": 0.0001125,
167
+ "loss": 0.0664,
168
+ "step": 4800
169
+ },
170
+ {
171
+ "epoch": 19.51,
172
+ "eval_loss": 0.8099461197853088,
173
+ "eval_runtime": 57.6128,
174
+ "eval_samples_per_second": 8.123,
175
+ "eval_wer": 0.4124207858048162,
176
+ "step": 4800
177
+ },
178
+ {
179
+ "epoch": 21.14,
180
+ "learning_rate": 9.505813953488371e-05,
181
+ "loss": 0.0611,
182
+ "step": 5200
183
+ },
184
+ {
185
+ "epoch": 21.14,
186
+ "eval_loss": 0.8145996928215027,
187
+ "eval_runtime": 57.1919,
188
+ "eval_samples_per_second": 8.183,
189
+ "eval_wer": 0.4114068441064639,
190
+ "step": 5200
191
+ },
192
+ {
193
+ "epoch": 22.76,
194
+ "learning_rate": 7.761627906976744e-05,
195
+ "loss": 0.0573,
196
+ "step": 5600
197
+ },
198
+ {
199
+ "epoch": 22.76,
200
+ "eval_loss": 0.8513393402099609,
201
+ "eval_runtime": 58.2107,
202
+ "eval_samples_per_second": 8.04,
203
+ "eval_wer": 0.4126742712294043,
204
+ "step": 5600
205
+ },
206
+ {
207
+ "epoch": 24.39,
208
+ "learning_rate": 6.0174418604651156e-05,
209
+ "loss": 0.0531,
210
+ "step": 6000
211
+ },
212
+ {
213
+ "epoch": 24.39,
214
+ "eval_loss": 0.8830446600914001,
215
+ "eval_runtime": 57.6531,
216
+ "eval_samples_per_second": 8.118,
217
+ "eval_wer": 0.4050697084917617,
218
+ "step": 6000
219
+ },
220
+ {
221
+ "epoch": 26.02,
222
+ "learning_rate": 4.273255813953488e-05,
223
+ "loss": 0.0498,
224
+ "step": 6400
225
+ },
226
+ {
227
+ "epoch": 26.02,
228
+ "eval_loss": 0.8851639628410339,
229
+ "eval_runtime": 57.3676,
230
+ "eval_samples_per_second": 8.158,
231
+ "eval_wer": 0.40481622306717363,
232
+ "step": 6400
233
+ },
234
+ {
235
+ "epoch": 27.64,
236
+ "learning_rate": 2.52906976744186e-05,
237
+ "loss": 0.0444,
238
+ "step": 6800
239
+ },
240
+ {
241
+ "epoch": 27.64,
242
+ "eval_loss": 0.936349093914032,
243
+ "eval_runtime": 57.896,
244
+ "eval_samples_per_second": 8.083,
245
+ "eval_wer": 0.40887198986058304,
246
+ "step": 6800
247
+ },
248
+ {
249
+ "epoch": 29.27,
250
+ "learning_rate": 7.848837209302325e-06,
251
+ "loss": 0.0392,
252
+ "step": 7200
253
+ },
254
+ {
255
+ "epoch": 29.27,
256
+ "eval_loss": 0.9341886043548584,
257
+ "eval_runtime": 57.5969,
258
+ "eval_samples_per_second": 8.125,
259
+ "eval_wer": 0.40126742712294045,
260
+ "step": 7200
261
+ }
262
+ ],
263
+ "max_steps": 7380,
264
+ "num_train_epochs": 30,
265
+ "total_flos": 4.018481177702621e+19,
266
+ "trial_name": null,
267
+ "trial_params": null
268
+ }
checkpoint-7200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28a85071638bb9d7d3f4b39303e2c56cdd6ea61baa2df1558d1e42576140f653
3
+ size 2351
preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": true,
7
+ "sampling_rate": 16000
8
+ }
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": "|"}
vocab.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 0, "t": 1, "y": 2, "ö": 3, "v": 4, "ʻ": 5, "r": 6, "m": 7, "x": 8, "w": 9, "'": 10, "c": 11, "g": 12, "h": 13, "f": 14, "l": 15, "s": 16, "o": 17, "u": 18, "d": 19, "a": 20, "ä": 21, "ѹ": 22, "j": 23, "i": 24, "о": 25, "e": 26, "n": 28, "ü": 29, "q": 30, "k": 31, "z": 32, "b": 33, "|": 27, "[UNK]": 34, "[PAD]": 35}