Hugging Face's logo

colerobertson
/
distilhubert-finetuned-not-a-word2

Audio Classification
Transformers
TensorBoard
Safetensors
hubert
Inference Endpoints
Model card Files Metrics Community
colerobertson commited on
Commit
d89240d
•
1 Parent(s): 3369a48

Training in progress, epoch 1

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. model.safetensors +1 -1
  2. run-0/checkpoint-144/config.json +80 -0
  3. run-0/checkpoint-144/model.safetensors +3 -0
  4. run-0/checkpoint-144/optimizer.pt +3 -0
  5. run-0/checkpoint-144/preprocessor_config.json +9 -0
  6. run-0/checkpoint-144/rng_state.pth +3 -0
  7. run-0/checkpoint-144/scheduler.pt +3 -0
  8. run-0/checkpoint-144/trainer_state.json +247 -0
  9. run-0/checkpoint-144/training_args.bin +3 -0
  10. run-0/checkpoint-192/config.json +80 -0
  11. run-0/checkpoint-192/model.safetensors +3 -0
  12. run-0/checkpoint-192/optimizer.pt +3 -0
  13. run-0/checkpoint-192/preprocessor_config.json +9 -0
  14. run-0/checkpoint-192/rng_state.pth +3 -0
  15. run-0/checkpoint-192/scheduler.pt +3 -0
  16. run-0/checkpoint-192/trainer_state.json +326 -0
  17. run-0/checkpoint-192/training_args.bin +3 -0
  18. run-0/checkpoint-240/config.json +80 -0
  19. run-0/checkpoint-240/model.safetensors +3 -0
  20. run-0/checkpoint-240/optimizer.pt +3 -0
  21. run-0/checkpoint-240/preprocessor_config.json +9 -0
  22. run-0/checkpoint-240/rng_state.pth +3 -0
  23. run-0/checkpoint-240/scheduler.pt +3 -0
  24. run-0/checkpoint-240/trainer_state.json +405 -0
  25. run-0/checkpoint-240/training_args.bin +3 -0
  26. run-0/checkpoint-288/config.json +80 -0
  27. run-0/checkpoint-288/model.safetensors +3 -0
  28. run-0/checkpoint-288/optimizer.pt +3 -0
  29. run-0/checkpoint-288/preprocessor_config.json +9 -0
  30. run-0/checkpoint-288/rng_state.pth +3 -0
  31. run-0/checkpoint-288/scheduler.pt +3 -0
  32. run-0/checkpoint-288/trainer_state.json +477 -0
  33. run-0/checkpoint-288/training_args.bin +3 -0
  34. run-0/checkpoint-336/config.json +80 -0
  35. run-0/checkpoint-336/model.safetensors +3 -0
  36. run-0/checkpoint-336/optimizer.pt +3 -0
  37. run-0/checkpoint-336/preprocessor_config.json +9 -0
  38. run-0/checkpoint-336/rng_state.pth +3 -0
  39. run-0/checkpoint-336/scheduler.pt +3 -0
  40. run-0/checkpoint-336/trainer_state.json +556 -0
  41. run-0/checkpoint-336/training_args.bin +3 -0
  42. run-0/checkpoint-384/config.json +80 -0
  43. run-0/checkpoint-384/model.safetensors +3 -0
  44. run-0/checkpoint-384/optimizer.pt +3 -0
  45. run-0/checkpoint-384/preprocessor_config.json +9 -0
  46. run-0/checkpoint-384/rng_state.pth +3 -0
  47. run-0/checkpoint-384/scheduler.pt +3 -0
  48. run-0/checkpoint-384/trainer_state.json +628 -0
  49. run-0/checkpoint-384/training_args.bin +3 -0
  50. run-0/checkpoint-432/config.json +80 -0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:217f7edb6688a569bb52663d265ca3a8d89c0ec31c358861da469fe433c133a9
3
  size 94763496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42d09eafd2930cae4b789dc9b1dd21dab399f3369a7f7dab6297bd28e15031d7
3
  size 94763496
run-0/checkpoint-144/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-144/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbd249682f247d9175e94df5795ecb65a897def6f9bff4b424a8e531ea8cf115
3
+ size 94763496
run-0/checkpoint-144/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c6c6835b482b755ceac7aadfc71bf8e03ec03e36687c526b6537d3b7c5682d2
3
+ size 189552570
run-0/checkpoint-144/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-144/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87b8bdd7b4355fd23f0b8256efb0158e4240e11263e992a13d50944c37692500
3
+ size 14244
run-0/checkpoint-144/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f90516656a2816df31c7f558bfa8502c55f64cbf50846da7ef147552dd20ea1
3
+ size 1064
run-0/checkpoint-144/trainer_state.json ADDED
@@ -0,0 +1,247 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7326732673267327,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-48",
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 144,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ }
234
+ ],
235
+ "logging_steps": 5,
236
+ "max_steps": 480,
237
+ "num_input_tokens_seen": 0,
238
+ "num_train_epochs": 10,
239
+ "save_steps": 500,
240
+ "total_flos": 2121874430755872.0,
241
+ "train_batch_size": 4,
242
+ "trial_name": null,
243
+ "trial_params": {
244
+ "learning_rate": 1.6870979154575775e-05,
245
+ "per_device_train_batch_size": 4
246
+ }
247
+ }
run-0/checkpoint-144/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-192/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-192/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:494e7b76fd86f833ac7eb4b7796007e351e6e6040f334df059c0418338596b5a
3
+ size 94763496
run-0/checkpoint-192/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a5599fafb75801fd64d03f7ab6e4a17c3ab76bd34b2b6b0afbe62f56d37cdd1
3
+ size 189552570
run-0/checkpoint-192/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-192/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5aef020ca2df517540ac9ff4e195e1c41a7b85939e93195d118078f119bc949
3
+ size 14244
run-0/checkpoint-192/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3507798446b60ba34521f29ffbf6b83b03e0559fc04f5721956c76f3783718af
3
+ size 1064
run-0/checkpoint-192/trainer_state.json ADDED
@@ -0,0 +1,326 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.74,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-192",
4
+ "epoch": 4.0,
5
+ "eval_steps": 500,
6
+ "global_step": 192,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 2.6530747413635254,
237
+ "learning_rate": 1.3082819483293713e-05,
238
+ "loss": 0.6167,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 2.0653493404388428,
244
+ "learning_rate": 1.2887553520856493e-05,
245
+ "loss": 0.5266,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 1.5351653099060059,
251
+ "learning_rate": 1.2692287558419274e-05,
252
+ "loss": 0.5224,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 1.5351920127868652,
258
+ "learning_rate": 1.2497021595982055e-05,
259
+ "loss": 0.622,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.5447399616241455,
265
+ "learning_rate": 1.2301755633544835e-05,
266
+ "loss": 0.4725,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 4.063059329986572,
272
+ "learning_rate": 1.2106489671107616e-05,
273
+ "loss": 0.4089,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 2.517096519470215,
279
+ "learning_rate": 1.1911223708670397e-05,
280
+ "loss": 0.579,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 2.847086191177368,
286
+ "learning_rate": 1.1755010938720622e-05,
287
+ "loss": 0.4687,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 1.7841814756393433,
293
+ "learning_rate": 1.1559744976283403e-05,
294
+ "loss": 0.4846,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 3.5284550189971924,
300
+ "learning_rate": 1.1364479013846184e-05,
301
+ "loss": 0.5628,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.74,
307
+ "eval_loss": 0.6948537826538086,
308
+ "eval_runtime": 1.3673,
309
+ "eval_samples_per_second": 46.807,
310
+ "eval_steps_per_second": 5.851,
311
+ "step": 192
312
+ }
313
+ ],
314
+ "logging_steps": 5,
315
+ "max_steps": 480,
316
+ "num_input_tokens_seen": 0,
317
+ "num_train_epochs": 10,
318
+ "save_steps": 500,
319
+ "total_flos": 2891755054954176.0,
320
+ "train_batch_size": 4,
321
+ "trial_name": null,
322
+ "trial_params": {
323
+ "learning_rate": 1.6870979154575775e-05,
324
+ "per_device_train_batch_size": 4
325
+ }
326
+ }
run-0/checkpoint-192/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-240/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-240/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2f5a0d44ac15e10163d55c6da6d4c66a9d4d22cf684b03b50ef9957fc6bf783
3
+ size 94763496
run-0/checkpoint-240/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a31269492b9e77b7194c4eaa9c9069709ebad096d6bb21d7b64952397ffabbe
3
+ size 189552570
run-0/checkpoint-240/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-240/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8eb071c49709b6f4047e7f48105f0dd51daaf73e0a11fd742255aa4c3526f42
3
+ size 14244
run-0/checkpoint-240/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45c27d79369faa613fdab2b285fce73b5cc1b4f5e1086859a8482102877cc85d
3
+ size 1064
run-0/checkpoint-240/trainer_state.json ADDED
@@ -0,0 +1,405 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7551020408163265,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-240",
4
+ "epoch": 5.0,
5
+ "eval_steps": 500,
6
+ "global_step": 240,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 2.6530747413635254,
237
+ "learning_rate": 1.3082819483293713e-05,
238
+ "loss": 0.6167,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 2.0653493404388428,
244
+ "learning_rate": 1.2887553520856493e-05,
245
+ "loss": 0.5266,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 1.5351653099060059,
251
+ "learning_rate": 1.2692287558419274e-05,
252
+ "loss": 0.5224,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 1.5351920127868652,
258
+ "learning_rate": 1.2497021595982055e-05,
259
+ "loss": 0.622,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.5447399616241455,
265
+ "learning_rate": 1.2301755633544835e-05,
266
+ "loss": 0.4725,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 4.063059329986572,
272
+ "learning_rate": 1.2106489671107616e-05,
273
+ "loss": 0.4089,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 2.517096519470215,
279
+ "learning_rate": 1.1911223708670397e-05,
280
+ "loss": 0.579,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 2.847086191177368,
286
+ "learning_rate": 1.1755010938720622e-05,
287
+ "loss": 0.4687,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 1.7841814756393433,
293
+ "learning_rate": 1.1559744976283403e-05,
294
+ "loss": 0.4846,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 3.5284550189971924,
300
+ "learning_rate": 1.1364479013846184e-05,
301
+ "loss": 0.5628,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.74,
307
+ "eval_loss": 0.6948537826538086,
308
+ "eval_runtime": 1.3673,
309
+ "eval_samples_per_second": 46.807,
310
+ "eval_steps_per_second": 5.851,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 1.6642357110977173,
316
+ "learning_rate": 1.1169213051408962e-05,
317
+ "loss": 0.5529,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 3.0695078372955322,
323
+ "learning_rate": 1.0973947088971741e-05,
324
+ "loss": 0.3534,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 2.5939130783081055,
330
+ "learning_rate": 1.0778681126534522e-05,
331
+ "loss": 0.4534,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": 9.777613639831543,
337
+ "learning_rate": 1.0583415164097303e-05,
338
+ "loss": 0.5345,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 4.490750312805176,
344
+ "learning_rate": 1.0388149201660083e-05,
345
+ "loss": 0.4854,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 3.7475597858428955,
351
+ "learning_rate": 1.0192883239222864e-05,
352
+ "loss": 0.4174,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.7532075643539429,
358
+ "learning_rate": 9.997617276785644e-06,
359
+ "loss": 0.4379,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 1.4538604021072388,
365
+ "learning_rate": 9.802351314348425e-06,
366
+ "loss": 0.3269,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 3.058314085006714,
372
+ "learning_rate": 9.607085351911206e-06,
373
+ "loss": 0.3553,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 5.367821216583252,
379
+ "learning_rate": 9.45087258196143e-06,
380
+ "loss": 0.4753,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7551020408163265,
386
+ "eval_loss": 0.7308206558227539,
387
+ "eval_runtime": 1.3546,
388
+ "eval_samples_per_second": 47.246,
389
+ "eval_steps_per_second": 5.906,
390
+ "step": 240
391
+ }
392
+ ],
393
+ "logging_steps": 5,
394
+ "max_steps": 480,
395
+ "num_input_tokens_seen": 0,
396
+ "num_train_epochs": 10,
397
+ "save_steps": 500,
398
+ "total_flos": 3654362860415712.0,
399
+ "train_batch_size": 4,
400
+ "trial_name": null,
401
+ "trial_params": {
402
+ "learning_rate": 1.6870979154575775e-05,
403
+ "per_device_train_batch_size": 4
404
+ }
405
+ }
run-0/checkpoint-240/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-288/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-288/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd92ffe27fe5c27bb997309c72b306c63a4a60fce68074ddc6c24650c0091ebc
3
+ size 94763496
run-0/checkpoint-288/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:adc9d8d06d58ee532c0027ebafbd4db55eceece415615ab965aad2b7bfd72538
3
+ size 189552570
run-0/checkpoint-288/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-288/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9903236b654011babeaee26ea70e1c6278fa670549b900c6df1d64732428a642
3
+ size 14244
run-0/checkpoint-288/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32c65caf188260d9f32d1c3709b7f2af991b3129c506772b942a9fa90c297ddb
3
+ size 1064
run-0/checkpoint-288/trainer_state.json ADDED
@@ -0,0 +1,477 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7551020408163265,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-240",
4
+ "epoch": 6.0,
5
+ "eval_steps": 500,
6
+ "global_step": 288,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 2.6530747413635254,
237
+ "learning_rate": 1.3082819483293713e-05,
238
+ "loss": 0.6167,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 2.0653493404388428,
244
+ "learning_rate": 1.2887553520856493e-05,
245
+ "loss": 0.5266,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 1.5351653099060059,
251
+ "learning_rate": 1.2692287558419274e-05,
252
+ "loss": 0.5224,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 1.5351920127868652,
258
+ "learning_rate": 1.2497021595982055e-05,
259
+ "loss": 0.622,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.5447399616241455,
265
+ "learning_rate": 1.2301755633544835e-05,
266
+ "loss": 0.4725,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 4.063059329986572,
272
+ "learning_rate": 1.2106489671107616e-05,
273
+ "loss": 0.4089,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 2.517096519470215,
279
+ "learning_rate": 1.1911223708670397e-05,
280
+ "loss": 0.579,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 2.847086191177368,
286
+ "learning_rate": 1.1755010938720622e-05,
287
+ "loss": 0.4687,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 1.7841814756393433,
293
+ "learning_rate": 1.1559744976283403e-05,
294
+ "loss": 0.4846,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 3.5284550189971924,
300
+ "learning_rate": 1.1364479013846184e-05,
301
+ "loss": 0.5628,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.74,
307
+ "eval_loss": 0.6948537826538086,
308
+ "eval_runtime": 1.3673,
309
+ "eval_samples_per_second": 46.807,
310
+ "eval_steps_per_second": 5.851,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 1.6642357110977173,
316
+ "learning_rate": 1.1169213051408962e-05,
317
+ "loss": 0.5529,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 3.0695078372955322,
323
+ "learning_rate": 1.0973947088971741e-05,
324
+ "loss": 0.3534,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 2.5939130783081055,
330
+ "learning_rate": 1.0778681126534522e-05,
331
+ "loss": 0.4534,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": 9.777613639831543,
337
+ "learning_rate": 1.0583415164097303e-05,
338
+ "loss": 0.5345,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 4.490750312805176,
344
+ "learning_rate": 1.0388149201660083e-05,
345
+ "loss": 0.4854,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 3.7475597858428955,
351
+ "learning_rate": 1.0192883239222864e-05,
352
+ "loss": 0.4174,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.7532075643539429,
358
+ "learning_rate": 9.997617276785644e-06,
359
+ "loss": 0.4379,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 1.4538604021072388,
365
+ "learning_rate": 9.802351314348425e-06,
366
+ "loss": 0.3269,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 3.058314085006714,
372
+ "learning_rate": 9.607085351911206e-06,
373
+ "loss": 0.3553,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 5.367821216583252,
379
+ "learning_rate": 9.45087258196143e-06,
380
+ "loss": 0.4753,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7551020408163265,
386
+ "eval_loss": 0.7308206558227539,
387
+ "eval_runtime": 1.3546,
388
+ "eval_samples_per_second": 47.246,
389
+ "eval_steps_per_second": 5.906,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 1.079416275024414,
395
+ "learning_rate": 9.25560661952421e-06,
396
+ "loss": 0.2547,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 6.890663146972656,
402
+ "learning_rate": 9.060340657086991e-06,
403
+ "loss": 0.3043,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 5.391942977905273,
409
+ "learning_rate": 8.86507469464977e-06,
410
+ "loss": 0.4958,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": Infinity,
416
+ "learning_rate": 8.708861924699996e-06,
417
+ "loss": 0.4157,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 2.2877392768859863,
423
+ "learning_rate": 8.513595962262776e-06,
424
+ "loss": 0.289,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 6.283646106719971,
430
+ "learning_rate": 8.318329999825557e-06,
431
+ "loss": 0.4315,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 8.02907657623291,
437
+ "learning_rate": 8.123064037388336e-06,
438
+ "loss": 0.3901,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 5.0022664070129395,
444
+ "learning_rate": 7.927798074951117e-06,
445
+ "loss": 0.3458,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 9.924915313720703,
451
+ "learning_rate": 7.732532112513897e-06,
452
+ "loss": 0.3239,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7526881720430108,
458
+ "eval_loss": 0.734130859375,
459
+ "eval_runtime": 1.3629,
460
+ "eval_samples_per_second": 46.958,
461
+ "eval_steps_per_second": 5.87,
462
+ "step": 288
463
+ }
464
+ ],
465
+ "logging_steps": 5,
466
+ "max_steps": 480,
467
+ "num_input_tokens_seen": 0,
468
+ "num_train_epochs": 10,
469
+ "save_steps": 500,
470
+ "total_flos": 4356539523260784.0,
471
+ "train_batch_size": 4,
472
+ "trial_name": null,
473
+ "trial_params": {
474
+ "learning_rate": 1.6870979154575775e-05,
475
+ "per_device_train_batch_size": 4
476
+ }
477
+ }
run-0/checkpoint-288/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-336/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-336/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95f98888d1c53d8cb63ff0487309ac7c8ace158d50b36a052facc806c08fc38e
3
+ size 94763496
run-0/checkpoint-336/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de78168bf4c4d94b7f32b614e818d24e08905a32a56475b3ae5f951fd6d5c4e2
3
+ size 189552570
run-0/checkpoint-336/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-336/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d12884ae20f0c926a355fda8650edc055a398d4c7c42545ccdb7d60bd202452
3
+ size 14244
run-0/checkpoint-336/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a93e746e734dd25bfab62ebb5e7354881f43d8c5f1c834242798dfd2074bc2ce
3
+ size 1064
run-0/checkpoint-336/trainer_state.json ADDED
@@ -0,0 +1,556 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7551020408163265,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-240",
4
+ "epoch": 7.0,
5
+ "eval_steps": 500,
6
+ "global_step": 336,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 2.6530747413635254,
237
+ "learning_rate": 1.3082819483293713e-05,
238
+ "loss": 0.6167,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 2.0653493404388428,
244
+ "learning_rate": 1.2887553520856493e-05,
245
+ "loss": 0.5266,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 1.5351653099060059,
251
+ "learning_rate": 1.2692287558419274e-05,
252
+ "loss": 0.5224,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 1.5351920127868652,
258
+ "learning_rate": 1.2497021595982055e-05,
259
+ "loss": 0.622,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.5447399616241455,
265
+ "learning_rate": 1.2301755633544835e-05,
266
+ "loss": 0.4725,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 4.063059329986572,
272
+ "learning_rate": 1.2106489671107616e-05,
273
+ "loss": 0.4089,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 2.517096519470215,
279
+ "learning_rate": 1.1911223708670397e-05,
280
+ "loss": 0.579,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 2.847086191177368,
286
+ "learning_rate": 1.1755010938720622e-05,
287
+ "loss": 0.4687,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 1.7841814756393433,
293
+ "learning_rate": 1.1559744976283403e-05,
294
+ "loss": 0.4846,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 3.5284550189971924,
300
+ "learning_rate": 1.1364479013846184e-05,
301
+ "loss": 0.5628,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.74,
307
+ "eval_loss": 0.6948537826538086,
308
+ "eval_runtime": 1.3673,
309
+ "eval_samples_per_second": 46.807,
310
+ "eval_steps_per_second": 5.851,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 1.6642357110977173,
316
+ "learning_rate": 1.1169213051408962e-05,
317
+ "loss": 0.5529,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 3.0695078372955322,
323
+ "learning_rate": 1.0973947088971741e-05,
324
+ "loss": 0.3534,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 2.5939130783081055,
330
+ "learning_rate": 1.0778681126534522e-05,
331
+ "loss": 0.4534,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": 9.777613639831543,
337
+ "learning_rate": 1.0583415164097303e-05,
338
+ "loss": 0.5345,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 4.490750312805176,
344
+ "learning_rate": 1.0388149201660083e-05,
345
+ "loss": 0.4854,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 3.7475597858428955,
351
+ "learning_rate": 1.0192883239222864e-05,
352
+ "loss": 0.4174,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.7532075643539429,
358
+ "learning_rate": 9.997617276785644e-06,
359
+ "loss": 0.4379,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 1.4538604021072388,
365
+ "learning_rate": 9.802351314348425e-06,
366
+ "loss": 0.3269,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 3.058314085006714,
372
+ "learning_rate": 9.607085351911206e-06,
373
+ "loss": 0.3553,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 5.367821216583252,
379
+ "learning_rate": 9.45087258196143e-06,
380
+ "loss": 0.4753,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7551020408163265,
386
+ "eval_loss": 0.7308206558227539,
387
+ "eval_runtime": 1.3546,
388
+ "eval_samples_per_second": 47.246,
389
+ "eval_steps_per_second": 5.906,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 1.079416275024414,
395
+ "learning_rate": 9.25560661952421e-06,
396
+ "loss": 0.2547,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 6.890663146972656,
402
+ "learning_rate": 9.060340657086991e-06,
403
+ "loss": 0.3043,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 5.391942977905273,
409
+ "learning_rate": 8.86507469464977e-06,
410
+ "loss": 0.4958,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": Infinity,
416
+ "learning_rate": 8.708861924699996e-06,
417
+ "loss": 0.4157,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 2.2877392768859863,
423
+ "learning_rate": 8.513595962262776e-06,
424
+ "loss": 0.289,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 6.283646106719971,
430
+ "learning_rate": 8.318329999825557e-06,
431
+ "loss": 0.4315,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 8.02907657623291,
437
+ "learning_rate": 8.123064037388336e-06,
438
+ "loss": 0.3901,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 5.0022664070129395,
444
+ "learning_rate": 7.927798074951117e-06,
445
+ "loss": 0.3458,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 9.924915313720703,
451
+ "learning_rate": 7.732532112513897e-06,
452
+ "loss": 0.3239,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7526881720430108,
458
+ "eval_loss": 0.734130859375,
459
+ "eval_runtime": 1.3629,
460
+ "eval_samples_per_second": 46.958,
461
+ "eval_steps_per_second": 5.87,
462
+ "step": 288
463
+ },
464
+ {
465
+ "epoch": 6.04,
466
+ "grad_norm": 6.063592910766602,
467
+ "learning_rate": 7.537266150076677e-06,
468
+ "loss": 0.4556,
469
+ "step": 290
470
+ },
471
+ {
472
+ "epoch": 6.15,
473
+ "grad_norm": 5.061245441436768,
474
+ "learning_rate": 7.3420001876394575e-06,
475
+ "loss": 0.2344,
476
+ "step": 295
477
+ },
478
+ {
479
+ "epoch": 6.25,
480
+ "grad_norm": 1.229814052581787,
481
+ "learning_rate": 7.146734225202238e-06,
482
+ "loss": 0.2622,
483
+ "step": 300
484
+ },
485
+ {
486
+ "epoch": 6.35,
487
+ "grad_norm": 10.423860549926758,
488
+ "learning_rate": 6.951468262765019e-06,
489
+ "loss": 0.3694,
490
+ "step": 305
491
+ },
492
+ {
493
+ "epoch": 6.46,
494
+ "grad_norm": 9.655594825744629,
495
+ "learning_rate": 6.756202300327799e-06,
496
+ "loss": 0.3442,
497
+ "step": 310
498
+ },
499
+ {
500
+ "epoch": 6.56,
501
+ "grad_norm": 3.6673190593719482,
502
+ "learning_rate": 6.560936337890579e-06,
503
+ "loss": 0.3164,
504
+ "step": 315
505
+ },
506
+ {
507
+ "epoch": 6.67,
508
+ "grad_norm": 11.887277603149414,
509
+ "learning_rate": 6.36567037545336e-06,
510
+ "loss": 0.2361,
511
+ "step": 320
512
+ },
513
+ {
514
+ "epoch": 6.77,
515
+ "grad_norm": 12.80982780456543,
516
+ "learning_rate": 6.17040441301614e-06,
517
+ "loss": 0.3502,
518
+ "step": 325
519
+ },
520
+ {
521
+ "epoch": 6.88,
522
+ "grad_norm": 14.678775787353516,
523
+ "learning_rate": 5.975138450578921e-06,
524
+ "loss": 0.2829,
525
+ "step": 330
526
+ },
527
+ {
528
+ "epoch": 6.98,
529
+ "grad_norm": 2.002340078353882,
530
+ "learning_rate": 5.7798724881417015e-06,
531
+ "loss": 0.3153,
532
+ "step": 335
533
+ },
534
+ {
535
+ "epoch": 7.0,
536
+ "eval_f1": 0.7368421052631577,
537
+ "eval_loss": 0.8391213417053223,
538
+ "eval_runtime": 1.3632,
539
+ "eval_samples_per_second": 46.947,
540
+ "eval_steps_per_second": 5.868,
541
+ "step": 336
542
+ }
543
+ ],
544
+ "logging_steps": 5,
545
+ "max_steps": 480,
546
+ "num_input_tokens_seen": 0,
547
+ "num_train_epochs": 10,
548
+ "save_steps": 500,
549
+ "total_flos": 5116387614670704.0,
550
+ "train_batch_size": 4,
551
+ "trial_name": null,
552
+ "trial_params": {
553
+ "learning_rate": 1.6870979154575775e-05,
554
+ "per_device_train_batch_size": 4
555
+ }
556
+ }
run-0/checkpoint-336/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-384/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }
run-0/checkpoint-384/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2916687feb7cc17d32b2a31e17ad5512424a6fa2fa8da91f15e461c8e89af87
3
+ size 94763496
run-0/checkpoint-384/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:755f0f8d2a05c523eb52bc9e0ef5b6eaa7b698f71fb4db6e2e264bc9014c1b6f
3
+ size 189552570
run-0/checkpoint-384/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
run-0/checkpoint-384/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5326b9611b4fb9dc5dc0b29580e7e48abf50913e44071592799c052bebfbacd7
3
+ size 14244
run-0/checkpoint-384/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c4157f3302c981a6f9ee8b08a83b0b6bf226da9e9c00e1267ffa9b3b1d1e447
3
+ size 1064
run-0/checkpoint-384/trainer_state.json ADDED
@@ -0,0 +1,628 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7551020408163265,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-240",
4
+ "epoch": 8.0,
5
+ "eval_steps": 500,
6
+ "global_step": 384,
7
+ "is_hyper_param_search": true,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1,
13
+ "grad_norm": 1.2701194286346436,
14
+ "learning_rate": 1.7573936619349767e-06,
15
+ "loss": 0.7024,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.218135952949524,
21
+ "learning_rate": 3.5147873238699533e-06,
22
+ "loss": 0.6932,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.5910921096801758,
28
+ "learning_rate": 5.2721809858049295e-06,
29
+ "loss": 0.6916,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 1.0592715740203857,
35
+ "learning_rate": 7.029574647739907e-06,
36
+ "loss": 0.6854,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 1.132177472114563,
42
+ "learning_rate": 8.786968309674883e-06,
43
+ "loss": 0.6719,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.9851714372634888,
49
+ "learning_rate": 1.0544361971609859e-05,
50
+ "loss": 0.6633,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 1.5592440366744995,
56
+ "learning_rate": 1.2301755633544835e-05,
57
+ "loss": 0.6288,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.8793619871139526,
63
+ "learning_rate": 1.4059149295479813e-05,
64
+ "loss": 0.6561,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 1.987308382987976,
70
+ "learning_rate": 1.581654295741479e-05,
71
+ "loss": 0.6343,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.6868629455566406,
78
+ "eval_runtime": 1.3443,
79
+ "eval_samples_per_second": 47.609,
80
+ "eval_steps_per_second": 5.951,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 3.1652989387512207,
86
+ "learning_rate": 1.6792872769600888e-05,
87
+ "loss": 0.6712,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9990371465682983,
93
+ "learning_rate": 1.6597606807163667e-05,
94
+ "loss": 0.6657,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.8210965394973755,
100
+ "learning_rate": 1.640234084472645e-05,
101
+ "loss": 0.476,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.5364782810211182,
107
+ "learning_rate": 1.6207074882289228e-05,
108
+ "loss": 0.5768,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.6772537231445312,
114
+ "learning_rate": 1.6011808919852007e-05,
115
+ "loss": 0.5039,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": 1.0453933477401733,
121
+ "learning_rate": 1.581654295741479e-05,
122
+ "loss": 0.7182,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 1.621067762374878,
128
+ "learning_rate": 1.562127699497757e-05,
129
+ "loss": 0.6178,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.7878597974777222,
135
+ "learning_rate": 1.542601103254035e-05,
136
+ "loss": 0.6192,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.6824157238006592,
142
+ "learning_rate": 1.5230745070103131e-05,
143
+ "loss": 0.6109,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 0.7989784479141235,
149
+ "learning_rate": 1.5035479107665912e-05,
150
+ "loss": 0.6367,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.7326732673267327,
156
+ "eval_loss": 0.7258987426757812,
157
+ "eval_runtime": 1.3947,
158
+ "eval_samples_per_second": 45.888,
159
+ "eval_steps_per_second": 5.736,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 0.6805263757705688,
165
+ "learning_rate": 1.484021314522869e-05,
166
+ "loss": 0.6271,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 0.6469630002975464,
172
+ "learning_rate": 1.4644947182791471e-05,
173
+ "loss": 0.5085,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 0.9272030591964722,
179
+ "learning_rate": 1.4449681220354252e-05,
180
+ "loss": 0.6377,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 0.7515610456466675,
186
+ "learning_rate": 1.4254415257917033e-05,
187
+ "loss": 0.655,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 2.1927835941314697,
193
+ "learning_rate": 1.4059149295479813e-05,
194
+ "loss": 0.5797,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 1.0289098024368286,
200
+ "learning_rate": 1.3863883333042594e-05,
201
+ "loss": 0.5291,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 1.7173566818237305,
207
+ "learning_rate": 1.3668617370605374e-05,
208
+ "loss": 0.6829,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 0.8574482202529907,
214
+ "learning_rate": 1.3473351408168155e-05,
215
+ "loss": 0.4509,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 1.1971725225448608,
221
+ "learning_rate": 1.3278085445730936e-05,
222
+ "loss": 0.4369,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.7326732673267327,
228
+ "eval_loss": 0.7368831634521484,
229
+ "eval_runtime": 1.4074,
230
+ "eval_samples_per_second": 45.475,
231
+ "eval_steps_per_second": 5.684,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 2.6530747413635254,
237
+ "learning_rate": 1.3082819483293713e-05,
238
+ "loss": 0.6167,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 2.0653493404388428,
244
+ "learning_rate": 1.2887553520856493e-05,
245
+ "loss": 0.5266,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 1.5351653099060059,
251
+ "learning_rate": 1.2692287558419274e-05,
252
+ "loss": 0.5224,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 1.5351920127868652,
258
+ "learning_rate": 1.2497021595982055e-05,
259
+ "loss": 0.622,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.5447399616241455,
265
+ "learning_rate": 1.2301755633544835e-05,
266
+ "loss": 0.4725,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 4.063059329986572,
272
+ "learning_rate": 1.2106489671107616e-05,
273
+ "loss": 0.4089,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 2.517096519470215,
279
+ "learning_rate": 1.1911223708670397e-05,
280
+ "loss": 0.579,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 2.847086191177368,
286
+ "learning_rate": 1.1755010938720622e-05,
287
+ "loss": 0.4687,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 1.7841814756393433,
293
+ "learning_rate": 1.1559744976283403e-05,
294
+ "loss": 0.4846,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 3.5284550189971924,
300
+ "learning_rate": 1.1364479013846184e-05,
301
+ "loss": 0.5628,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.74,
307
+ "eval_loss": 0.6948537826538086,
308
+ "eval_runtime": 1.3673,
309
+ "eval_samples_per_second": 46.807,
310
+ "eval_steps_per_second": 5.851,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 1.6642357110977173,
316
+ "learning_rate": 1.1169213051408962e-05,
317
+ "loss": 0.5529,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 3.0695078372955322,
323
+ "learning_rate": 1.0973947088971741e-05,
324
+ "loss": 0.3534,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 2.5939130783081055,
330
+ "learning_rate": 1.0778681126534522e-05,
331
+ "loss": 0.4534,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": 9.777613639831543,
337
+ "learning_rate": 1.0583415164097303e-05,
338
+ "loss": 0.5345,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 4.490750312805176,
344
+ "learning_rate": 1.0388149201660083e-05,
345
+ "loss": 0.4854,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 3.7475597858428955,
351
+ "learning_rate": 1.0192883239222864e-05,
352
+ "loss": 0.4174,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.7532075643539429,
358
+ "learning_rate": 9.997617276785644e-06,
359
+ "loss": 0.4379,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 1.4538604021072388,
365
+ "learning_rate": 9.802351314348425e-06,
366
+ "loss": 0.3269,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 3.058314085006714,
372
+ "learning_rate": 9.607085351911206e-06,
373
+ "loss": 0.3553,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 5.367821216583252,
379
+ "learning_rate": 9.45087258196143e-06,
380
+ "loss": 0.4753,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7551020408163265,
386
+ "eval_loss": 0.7308206558227539,
387
+ "eval_runtime": 1.3546,
388
+ "eval_samples_per_second": 47.246,
389
+ "eval_steps_per_second": 5.906,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 1.079416275024414,
395
+ "learning_rate": 9.25560661952421e-06,
396
+ "loss": 0.2547,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 6.890663146972656,
402
+ "learning_rate": 9.060340657086991e-06,
403
+ "loss": 0.3043,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 5.391942977905273,
409
+ "learning_rate": 8.86507469464977e-06,
410
+ "loss": 0.4958,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": Infinity,
416
+ "learning_rate": 8.708861924699996e-06,
417
+ "loss": 0.4157,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 2.2877392768859863,
423
+ "learning_rate": 8.513595962262776e-06,
424
+ "loss": 0.289,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 6.283646106719971,
430
+ "learning_rate": 8.318329999825557e-06,
431
+ "loss": 0.4315,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 8.02907657623291,
437
+ "learning_rate": 8.123064037388336e-06,
438
+ "loss": 0.3901,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 5.0022664070129395,
444
+ "learning_rate": 7.927798074951117e-06,
445
+ "loss": 0.3458,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 9.924915313720703,
451
+ "learning_rate": 7.732532112513897e-06,
452
+ "loss": 0.3239,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7526881720430108,
458
+ "eval_loss": 0.734130859375,
459
+ "eval_runtime": 1.3629,
460
+ "eval_samples_per_second": 46.958,
461
+ "eval_steps_per_second": 5.87,
462
+ "step": 288
463
+ },
464
+ {
465
+ "epoch": 6.04,
466
+ "grad_norm": 6.063592910766602,
467
+ "learning_rate": 7.537266150076677e-06,
468
+ "loss": 0.4556,
469
+ "step": 290
470
+ },
471
+ {
472
+ "epoch": 6.15,
473
+ "grad_norm": 5.061245441436768,
474
+ "learning_rate": 7.3420001876394575e-06,
475
+ "loss": 0.2344,
476
+ "step": 295
477
+ },
478
+ {
479
+ "epoch": 6.25,
480
+ "grad_norm": 1.229814052581787,
481
+ "learning_rate": 7.146734225202238e-06,
482
+ "loss": 0.2622,
483
+ "step": 300
484
+ },
485
+ {
486
+ "epoch": 6.35,
487
+ "grad_norm": 10.423860549926758,
488
+ "learning_rate": 6.951468262765019e-06,
489
+ "loss": 0.3694,
490
+ "step": 305
491
+ },
492
+ {
493
+ "epoch": 6.46,
494
+ "grad_norm": 9.655594825744629,
495
+ "learning_rate": 6.756202300327799e-06,
496
+ "loss": 0.3442,
497
+ "step": 310
498
+ },
499
+ {
500
+ "epoch": 6.56,
501
+ "grad_norm": 3.6673190593719482,
502
+ "learning_rate": 6.560936337890579e-06,
503
+ "loss": 0.3164,
504
+ "step": 315
505
+ },
506
+ {
507
+ "epoch": 6.67,
508
+ "grad_norm": 11.887277603149414,
509
+ "learning_rate": 6.36567037545336e-06,
510
+ "loss": 0.2361,
511
+ "step": 320
512
+ },
513
+ {
514
+ "epoch": 6.77,
515
+ "grad_norm": 12.80982780456543,
516
+ "learning_rate": 6.17040441301614e-06,
517
+ "loss": 0.3502,
518
+ "step": 325
519
+ },
520
+ {
521
+ "epoch": 6.88,
522
+ "grad_norm": 14.678775787353516,
523
+ "learning_rate": 5.975138450578921e-06,
524
+ "loss": 0.2829,
525
+ "step": 330
526
+ },
527
+ {
528
+ "epoch": 6.98,
529
+ "grad_norm": 2.002340078353882,
530
+ "learning_rate": 5.7798724881417015e-06,
531
+ "loss": 0.3153,
532
+ "step": 335
533
+ },
534
+ {
535
+ "epoch": 7.0,
536
+ "eval_f1": 0.7368421052631577,
537
+ "eval_loss": 0.8391213417053223,
538
+ "eval_runtime": 1.3632,
539
+ "eval_samples_per_second": 46.947,
540
+ "eval_steps_per_second": 5.868,
541
+ "step": 336
542
+ },
543
+ {
544
+ "epoch": 7.08,
545
+ "grad_norm": 14.519176483154297,
546
+ "learning_rate": 5.584606525704481e-06,
547
+ "loss": 0.1675,
548
+ "step": 340
549
+ },
550
+ {
551
+ "epoch": 7.19,
552
+ "grad_norm": 1.7075341939926147,
553
+ "learning_rate": 5.389340563267261e-06,
554
+ "loss": 0.3502,
555
+ "step": 345
556
+ },
557
+ {
558
+ "epoch": 7.29,
559
+ "grad_norm": 1.1975409984588623,
560
+ "learning_rate": 5.194074600830042e-06,
561
+ "loss": 0.1761,
562
+ "step": 350
563
+ },
564
+ {
565
+ "epoch": 7.4,
566
+ "grad_norm": 24.773109436035156,
567
+ "learning_rate": 4.998808638392822e-06,
568
+ "loss": 0.2723,
569
+ "step": 355
570
+ },
571
+ {
572
+ "epoch": 7.5,
573
+ "grad_norm": 1.3228851556777954,
574
+ "learning_rate": 4.803542675955603e-06,
575
+ "loss": 0.2997,
576
+ "step": 360
577
+ },
578
+ {
579
+ "epoch": 7.6,
580
+ "grad_norm": 11.504044532775879,
581
+ "learning_rate": 4.608276713518383e-06,
582
+ "loss": 0.167,
583
+ "step": 365
584
+ },
585
+ {
586
+ "epoch": 7.71,
587
+ "grad_norm": 4.855327129364014,
588
+ "learning_rate": 4.413010751081163e-06,
589
+ "loss": 0.1576,
590
+ "step": 370
591
+ },
592
+ {
593
+ "epoch": 7.81,
594
+ "grad_norm": 3.231921434402466,
595
+ "learning_rate": 4.217744788643944e-06,
596
+ "loss": 0.3102,
597
+ "step": 375
598
+ },
599
+ {
600
+ "epoch": 7.92,
601
+ "grad_norm": 11.368690490722656,
602
+ "learning_rate": 4.022478826206724e-06,
603
+ "loss": 0.3451,
604
+ "step": 380
605
+ },
606
+ {
607
+ "epoch": 8.0,
608
+ "eval_f1": 0.7333333333333333,
609
+ "eval_loss": 0.8304653167724609,
610
+ "eval_runtime": 1.3846,
611
+ "eval_samples_per_second": 46.222,
612
+ "eval_steps_per_second": 5.778,
613
+ "step": 384
614
+ }
615
+ ],
616
+ "logging_steps": 5,
617
+ "max_steps": 480,
618
+ "num_input_tokens_seen": 0,
619
+ "num_train_epochs": 10,
620
+ "save_steps": 500,
621
+ "total_flos": 5808783041309760.0,
622
+ "train_batch_size": 4,
623
+ "trial_name": null,
624
+ "trial_params": {
625
+ "learning_rate": 1.6870979154575775e-05,
626
+ "per_device_train_batch_size": 4
627
+ }
628
+ }
run-0/checkpoint-384/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3adcce4c12e301b4690ffc59bfab1ebdfc4aeec76c20952369aa2f6c1a91a7
3
+ size 4920
run-0/checkpoint-432/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "ntu-spml/distilhubert",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": false,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": false,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_norm": "group",
45
+ "feat_proj_dropout": 0.0,
46
+ "feat_proj_layer_norm": false,
47
+ "final_dropout": 0.0,
48
+ "hidden_act": "gelu",
49
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "NOT_WORD",
53
+ "1": "WORD"
54
+ },
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "label2id": {
58
+ "NOT_WORD": "0",
59
+ "WORD": "1"
60
+ },
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_feature_length": 10,
64
+ "mask_feature_min_masks": 0,
65
+ "mask_feature_prob": 0.0,
66
+ "mask_time_length": 10,
67
+ "mask_time_min_masks": 2,
68
+ "mask_time_prob": 0.05,
69
+ "model_type": "hubert",
70
+ "num_attention_heads": 12,
71
+ "num_conv_pos_embedding_groups": 16,
72
+ "num_conv_pos_embeddings": 128,
73
+ "num_feat_extract_layers": 7,
74
+ "num_hidden_layers": 2,
75
+ "pad_token_id": 0,
76
+ "torch_dtype": "float32",
77
+ "transformers_version": "4.38.1",
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 32
80
+ }