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
fef3774
•
1 Parent(s): 4dc7c40

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-12/checkpoint-144/config.json +80 -0
  3. run-12/checkpoint-144/model.safetensors +3 -0
  4. run-12/checkpoint-144/optimizer.pt +3 -0
  5. run-12/checkpoint-144/preprocessor_config.json +9 -0
  6. run-12/checkpoint-144/rng_state.pth +3 -0
  7. run-12/checkpoint-144/scheduler.pt +3 -0
  8. run-12/checkpoint-144/trainer_state.json +247 -0
  9. run-12/checkpoint-144/training_args.bin +3 -0
  10. run-12/checkpoint-192/config.json +80 -0
  11. run-12/checkpoint-192/model.safetensors +3 -0
  12. run-12/checkpoint-192/optimizer.pt +3 -0
  13. run-12/checkpoint-192/preprocessor_config.json +9 -0
  14. run-12/checkpoint-192/rng_state.pth +3 -0
  15. run-12/checkpoint-192/scheduler.pt +3 -0
  16. run-12/checkpoint-192/trainer_state.json +326 -0
  17. run-12/checkpoint-192/training_args.bin +3 -0
  18. run-12/checkpoint-240/config.json +80 -0
  19. run-12/checkpoint-240/model.safetensors +3 -0
  20. run-12/checkpoint-240/optimizer.pt +3 -0
  21. run-12/checkpoint-240/preprocessor_config.json +9 -0
  22. run-12/checkpoint-240/rng_state.pth +3 -0
  23. run-12/checkpoint-240/scheduler.pt +3 -0
  24. run-12/checkpoint-240/trainer_state.json +405 -0
  25. run-12/checkpoint-240/training_args.bin +3 -0
  26. run-12/checkpoint-288/config.json +80 -0
  27. run-12/checkpoint-288/model.safetensors +3 -0
  28. run-12/checkpoint-288/optimizer.pt +3 -0
  29. run-12/checkpoint-288/preprocessor_config.json +9 -0
  30. run-12/checkpoint-288/rng_state.pth +3 -0
  31. run-12/checkpoint-288/scheduler.pt +3 -0
  32. run-12/checkpoint-288/trainer_state.json +477 -0
  33. run-12/checkpoint-288/training_args.bin +3 -0
  34. run-12/checkpoint-336/config.json +80 -0
  35. run-12/checkpoint-336/model.safetensors +3 -0
  36. run-12/checkpoint-336/optimizer.pt +3 -0
  37. run-12/checkpoint-336/preprocessor_config.json +9 -0
  38. run-12/checkpoint-336/rng_state.pth +3 -0
  39. run-12/checkpoint-336/scheduler.pt +3 -0
  40. run-12/checkpoint-336/trainer_state.json +556 -0
  41. run-12/checkpoint-336/training_args.bin +3 -0
  42. run-12/checkpoint-384/config.json +80 -0
  43. run-12/checkpoint-384/model.safetensors +3 -0
  44. run-12/checkpoint-384/optimizer.pt +3 -0
  45. run-12/checkpoint-384/preprocessor_config.json +9 -0
  46. run-12/checkpoint-384/rng_state.pth +3 -0
  47. run-12/checkpoint-384/scheduler.pt +3 -0
  48. run-12/checkpoint-384/trainer_state.json +628 -0
  49. run-12/checkpoint-384/training_args.bin +3 -0
  50. run-12/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:e0ba6fbcbb9cb83150dc1524c9934db5438ee17d96fb72054875ac1e12dab680
3
  size 94763496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4e98bc19214e703abd204c0ffebd30d38df81e5c2e8f315a5706737f9081bfb
3
  size 94763496
run-12/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-12/checkpoint-144/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eaafb011b379cf62cd419e44806cffe5c40f0ebd1494eb84cc11412907c1ce86
3
+ size 94763496
run-12/checkpoint-144/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a42d64a2ac9a1d13bb5c26fffe6ea5381a46d607590c0f1ecab42ec2e44a17c4
3
+ size 189552570
run-12/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-12/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-12/checkpoint-144/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ff75340c86479687882ce65f14a1d7a456a7caa002168531c1e959e3e0c6f81
3
+ size 1064
run-12/checkpoint-144/trainer_state.json ADDED
@@ -0,0 +1,247 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.74,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/checkpoint-96",
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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
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": 5.6339237645170805e-05,
245
+ "per_device_train_batch_size": 4
246
+ }
247
+ }
run-12/checkpoint-144/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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-12/checkpoint-192/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d31cb05aeba6e2292924bbc04460e937269bf312e23f53065bce60cfa3ef407
3
+ size 94763496
run-12/checkpoint-192/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b4e79cde86596b89d27a8c6e6cbd607aad30e711651a975a8bedeb06a6d7491
3
+ size 189552570
run-12/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-12/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-12/checkpoint-192/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a28c17e73d522b84903a6b6c8b074965e775699a23d01d2ac1ac8476b3765fc
3
+ size 1064
run-12/checkpoint-192/trainer_state.json ADDED
@@ -0,0 +1,326 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7500000000000001,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 3.1919455528259277,
237
+ "learning_rate": 4.421065176322431e-05,
238
+ "loss": 0.485,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 56.47049331665039,
244
+ "learning_rate": 4.355857725344224e-05,
245
+ "loss": 0.3607,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 4.917665481567383,
251
+ "learning_rate": 4.290650274366017e-05,
252
+ "loss": 0.2077,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 8.88060188293457,
258
+ "learning_rate": 4.238484313583451e-05,
259
+ "loss": 0.6496,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.9647566080093384,
265
+ "learning_rate": 4.1732768626052447e-05,
266
+ "loss": 0.1394,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 50.96482849121094,
272
+ "learning_rate": 4.1080694116270374e-05,
273
+ "loss": 0.2834,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 132.5663299560547,
279
+ "learning_rate": 4.042861960648831e-05,
280
+ "loss": 0.485,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 3.1624388694763184,
286
+ "learning_rate": 3.977654509670624e-05,
287
+ "loss": 0.2353,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 0.5671400427818298,
293
+ "learning_rate": 3.912447058692417e-05,
294
+ "loss": 0.3058,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 23.86330223083496,
300
+ "learning_rate": 3.84723960771421e-05,
301
+ "loss": 0.3764,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.7500000000000001,
307
+ "eval_loss": 1.1437296867370605,
308
+ "eval_runtime": 1.4019,
309
+ "eval_samples_per_second": 45.651,
310
+ "eval_steps_per_second": 5.706,
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": 5.6339237645170805e-05,
324
+ "per_device_train_batch_size": 4
325
+ }
326
+ }
run-12/checkpoint-192/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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-12/checkpoint-240/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe13919b8a2e8d78cedcf94e20fd8cc69cbfb8388d40b30dd66e644b8f4244be
3
+ size 94763496
run-12/checkpoint-240/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e96c3e7d3c1371959d9508cb869c652cac871af40a27e34b7b41aa900fc88fd7
3
+ size 189552570
run-12/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-12/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-12/checkpoint-240/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd508f8a2232ad18a737df054d955b6333c9edb38074719f2e3c8bb421c75a1e
3
+ size 1064
run-12/checkpoint-240/trainer_state.json ADDED
@@ -0,0 +1,405 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7865168539325843,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 3.1919455528259277,
237
+ "learning_rate": 4.421065176322431e-05,
238
+ "loss": 0.485,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 56.47049331665039,
244
+ "learning_rate": 4.355857725344224e-05,
245
+ "loss": 0.3607,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 4.917665481567383,
251
+ "learning_rate": 4.290650274366017e-05,
252
+ "loss": 0.2077,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 8.88060188293457,
258
+ "learning_rate": 4.238484313583451e-05,
259
+ "loss": 0.6496,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.9647566080093384,
265
+ "learning_rate": 4.1732768626052447e-05,
266
+ "loss": 0.1394,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 50.96482849121094,
272
+ "learning_rate": 4.1080694116270374e-05,
273
+ "loss": 0.2834,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 132.5663299560547,
279
+ "learning_rate": 4.042861960648831e-05,
280
+ "loss": 0.485,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 3.1624388694763184,
286
+ "learning_rate": 3.977654509670624e-05,
287
+ "loss": 0.2353,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 0.5671400427818298,
293
+ "learning_rate": 3.912447058692417e-05,
294
+ "loss": 0.3058,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 23.86330223083496,
300
+ "learning_rate": 3.84723960771421e-05,
301
+ "loss": 0.3764,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.7500000000000001,
307
+ "eval_loss": 1.1437296867370605,
308
+ "eval_runtime": 1.4019,
309
+ "eval_samples_per_second": 45.651,
310
+ "eval_steps_per_second": 5.706,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 0.443531334400177,
316
+ "learning_rate": 3.782032156736003e-05,
317
+ "loss": 0.2652,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 5.118373394012451,
323
+ "learning_rate": 3.716824705757796e-05,
324
+ "loss": 0.3585,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 0.7104140520095825,
330
+ "learning_rate": 3.651617254779589e-05,
331
+ "loss": 0.2533,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": Infinity,
337
+ "learning_rate": 3.599451293997023e-05,
338
+ "loss": 0.1963,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 1.0030689239501953,
344
+ "learning_rate": 3.5342438430188165e-05,
345
+ "loss": 0.0602,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 193.33819580078125,
351
+ "learning_rate": 3.469036392040609e-05,
352
+ "loss": 0.2754,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.3048127889633179,
358
+ "learning_rate": 3.403828941062403e-05,
359
+ "loss": 0.1878,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 0.18795377016067505,
365
+ "learning_rate": 3.3386214900841956e-05,
366
+ "loss": 0.0245,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 0.7579631805419922,
372
+ "learning_rate": 3.2864555293016304e-05,
373
+ "loss": 0.3502,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 0.39707404375076294,
379
+ "learning_rate": 3.221248078323424e-05,
380
+ "loss": 0.2378,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7865168539325843,
386
+ "eval_loss": 1.2380468845367432,
387
+ "eval_runtime": 1.3627,
388
+ "eval_samples_per_second": 46.967,
389
+ "eval_steps_per_second": 5.871,
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": 5.6339237645170805e-05,
403
+ "per_device_train_batch_size": 4
404
+ }
405
+ }
run-12/checkpoint-240/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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-12/checkpoint-288/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c093bd7bab44904ca418c380647c2afe644592f1142ec0fb6b8b814b518ddfaf
3
+ size 94763496
run-12/checkpoint-288/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca9b5da025052b4a803c74df25704dc62cdaef93efa14bfad3a43bc7e5cb566a
3
+ size 189552570
run-12/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-12/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-12/checkpoint-288/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1a8b51dc202eacec90154ad2e823e318f7d3712d39d66c177f64f58d1dd484f
3
+ size 1064
run-12/checkpoint-288/trainer_state.json ADDED
@@ -0,0 +1,477 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7865168539325843,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 3.1919455528259277,
237
+ "learning_rate": 4.421065176322431e-05,
238
+ "loss": 0.485,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 56.47049331665039,
244
+ "learning_rate": 4.355857725344224e-05,
245
+ "loss": 0.3607,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 4.917665481567383,
251
+ "learning_rate": 4.290650274366017e-05,
252
+ "loss": 0.2077,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 8.88060188293457,
258
+ "learning_rate": 4.238484313583451e-05,
259
+ "loss": 0.6496,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.9647566080093384,
265
+ "learning_rate": 4.1732768626052447e-05,
266
+ "loss": 0.1394,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 50.96482849121094,
272
+ "learning_rate": 4.1080694116270374e-05,
273
+ "loss": 0.2834,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 132.5663299560547,
279
+ "learning_rate": 4.042861960648831e-05,
280
+ "loss": 0.485,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 3.1624388694763184,
286
+ "learning_rate": 3.977654509670624e-05,
287
+ "loss": 0.2353,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 0.5671400427818298,
293
+ "learning_rate": 3.912447058692417e-05,
294
+ "loss": 0.3058,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 23.86330223083496,
300
+ "learning_rate": 3.84723960771421e-05,
301
+ "loss": 0.3764,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.7500000000000001,
307
+ "eval_loss": 1.1437296867370605,
308
+ "eval_runtime": 1.4019,
309
+ "eval_samples_per_second": 45.651,
310
+ "eval_steps_per_second": 5.706,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 0.443531334400177,
316
+ "learning_rate": 3.782032156736003e-05,
317
+ "loss": 0.2652,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 5.118373394012451,
323
+ "learning_rate": 3.716824705757796e-05,
324
+ "loss": 0.3585,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 0.7104140520095825,
330
+ "learning_rate": 3.651617254779589e-05,
331
+ "loss": 0.2533,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": Infinity,
337
+ "learning_rate": 3.599451293997023e-05,
338
+ "loss": 0.1963,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 1.0030689239501953,
344
+ "learning_rate": 3.5342438430188165e-05,
345
+ "loss": 0.0602,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 193.33819580078125,
351
+ "learning_rate": 3.469036392040609e-05,
352
+ "loss": 0.2754,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.3048127889633179,
358
+ "learning_rate": 3.403828941062403e-05,
359
+ "loss": 0.1878,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 0.18795377016067505,
365
+ "learning_rate": 3.3386214900841956e-05,
366
+ "loss": 0.0245,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 0.7579631805419922,
372
+ "learning_rate": 3.2864555293016304e-05,
373
+ "loss": 0.3502,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 0.39707404375076294,
379
+ "learning_rate": 3.221248078323424e-05,
380
+ "loss": 0.2378,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7865168539325843,
386
+ "eval_loss": 1.2380468845367432,
387
+ "eval_runtime": 1.3627,
388
+ "eval_samples_per_second": 46.967,
389
+ "eval_steps_per_second": 5.871,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 0.18084272742271423,
395
+ "learning_rate": 3.1560406273452166e-05,
396
+ "loss": 0.0143,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 0.6576858758926392,
402
+ "learning_rate": 3.0908331763670094e-05,
403
+ "loss": 0.0151,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 1.1509562730789185,
409
+ "learning_rate": 3.025625725388803e-05,
410
+ "loss": 0.3467,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": 29.21381187438965,
416
+ "learning_rate": 2.9604182744105953e-05,
417
+ "loss": 0.0524,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 0.1710425168275833,
423
+ "learning_rate": 2.8952108234323884e-05,
424
+ "loss": 0.0106,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 0.14679908752441406,
430
+ "learning_rate": 2.8300033724541812e-05,
431
+ "loss": 0.0411,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 85.6775894165039,
437
+ "learning_rate": 2.7647959214759747e-05,
438
+ "loss": 0.4682,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 0.12999875843524933,
444
+ "learning_rate": 2.6995884704977678e-05,
445
+ "loss": 0.2671,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 23.26946258544922,
451
+ "learning_rate": 2.634381019519561e-05,
452
+ "loss": 0.2224,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7741935483870968,
458
+ "eval_loss": 1.5522700548171997,
459
+ "eval_runtime": 1.3653,
460
+ "eval_samples_per_second": 46.878,
461
+ "eval_steps_per_second": 5.86,
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": 5.6339237645170805e-05,
475
+ "per_device_train_batch_size": 4
476
+ }
477
+ }
run-12/checkpoint-288/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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-12/checkpoint-336/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6fbfe341e802e8c1c9141fd1674a0212dbf9350613781f59e36db618af34432d
3
+ size 94763496
run-12/checkpoint-336/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f6024e606695ce49d9cbde3810cc0a02be602cc3527e7c8f397a7dd0a75f218
3
+ size 189552570
run-12/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-12/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-12/checkpoint-336/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb236290338c4bdc836db4b1cd729b56ec7945118ed02875a5414df94a328339
3
+ size 1064
run-12/checkpoint-336/trainer_state.json ADDED
@@ -0,0 +1,556 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7865168539325843,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 3.1919455528259277,
237
+ "learning_rate": 4.421065176322431e-05,
238
+ "loss": 0.485,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 56.47049331665039,
244
+ "learning_rate": 4.355857725344224e-05,
245
+ "loss": 0.3607,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 4.917665481567383,
251
+ "learning_rate": 4.290650274366017e-05,
252
+ "loss": 0.2077,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 8.88060188293457,
258
+ "learning_rate": 4.238484313583451e-05,
259
+ "loss": 0.6496,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.9647566080093384,
265
+ "learning_rate": 4.1732768626052447e-05,
266
+ "loss": 0.1394,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 50.96482849121094,
272
+ "learning_rate": 4.1080694116270374e-05,
273
+ "loss": 0.2834,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 132.5663299560547,
279
+ "learning_rate": 4.042861960648831e-05,
280
+ "loss": 0.485,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 3.1624388694763184,
286
+ "learning_rate": 3.977654509670624e-05,
287
+ "loss": 0.2353,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 0.5671400427818298,
293
+ "learning_rate": 3.912447058692417e-05,
294
+ "loss": 0.3058,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 23.86330223083496,
300
+ "learning_rate": 3.84723960771421e-05,
301
+ "loss": 0.3764,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.7500000000000001,
307
+ "eval_loss": 1.1437296867370605,
308
+ "eval_runtime": 1.4019,
309
+ "eval_samples_per_second": 45.651,
310
+ "eval_steps_per_second": 5.706,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 0.443531334400177,
316
+ "learning_rate": 3.782032156736003e-05,
317
+ "loss": 0.2652,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 5.118373394012451,
323
+ "learning_rate": 3.716824705757796e-05,
324
+ "loss": 0.3585,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 0.7104140520095825,
330
+ "learning_rate": 3.651617254779589e-05,
331
+ "loss": 0.2533,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": Infinity,
337
+ "learning_rate": 3.599451293997023e-05,
338
+ "loss": 0.1963,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 1.0030689239501953,
344
+ "learning_rate": 3.5342438430188165e-05,
345
+ "loss": 0.0602,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 193.33819580078125,
351
+ "learning_rate": 3.469036392040609e-05,
352
+ "loss": 0.2754,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.3048127889633179,
358
+ "learning_rate": 3.403828941062403e-05,
359
+ "loss": 0.1878,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 0.18795377016067505,
365
+ "learning_rate": 3.3386214900841956e-05,
366
+ "loss": 0.0245,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 0.7579631805419922,
372
+ "learning_rate": 3.2864555293016304e-05,
373
+ "loss": 0.3502,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 0.39707404375076294,
379
+ "learning_rate": 3.221248078323424e-05,
380
+ "loss": 0.2378,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7865168539325843,
386
+ "eval_loss": 1.2380468845367432,
387
+ "eval_runtime": 1.3627,
388
+ "eval_samples_per_second": 46.967,
389
+ "eval_steps_per_second": 5.871,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 0.18084272742271423,
395
+ "learning_rate": 3.1560406273452166e-05,
396
+ "loss": 0.0143,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 0.6576858758926392,
402
+ "learning_rate": 3.0908331763670094e-05,
403
+ "loss": 0.0151,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 1.1509562730789185,
409
+ "learning_rate": 3.025625725388803e-05,
410
+ "loss": 0.3467,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": 29.21381187438965,
416
+ "learning_rate": 2.9604182744105953e-05,
417
+ "loss": 0.0524,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 0.1710425168275833,
423
+ "learning_rate": 2.8952108234323884e-05,
424
+ "loss": 0.0106,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 0.14679908752441406,
430
+ "learning_rate": 2.8300033724541812e-05,
431
+ "loss": 0.0411,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 85.6775894165039,
437
+ "learning_rate": 2.7647959214759747e-05,
438
+ "loss": 0.4682,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 0.12999875843524933,
444
+ "learning_rate": 2.6995884704977678e-05,
445
+ "loss": 0.2671,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 23.26946258544922,
451
+ "learning_rate": 2.634381019519561e-05,
452
+ "loss": 0.2224,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7741935483870968,
458
+ "eval_loss": 1.5522700548171997,
459
+ "eval_runtime": 1.3653,
460
+ "eval_samples_per_second": 46.878,
461
+ "eval_steps_per_second": 5.86,
462
+ "step": 288
463
+ },
464
+ {
465
+ "epoch": 6.04,
466
+ "grad_norm": 0.12312161922454834,
467
+ "learning_rate": 2.569173568541354e-05,
468
+ "loss": 0.2473,
469
+ "step": 290
470
+ },
471
+ {
472
+ "epoch": 6.15,
473
+ "grad_norm": 0.21050110459327698,
474
+ "learning_rate": 2.503966117563147e-05,
475
+ "loss": 0.0108,
476
+ "step": 295
477
+ },
478
+ {
479
+ "epoch": 6.25,
480
+ "grad_norm": 0.1071925163269043,
481
+ "learning_rate": 2.43875866658494e-05,
482
+ "loss": 0.0104,
483
+ "step": 300
484
+ },
485
+ {
486
+ "epoch": 6.35,
487
+ "grad_norm": 0.5196150541305542,
488
+ "learning_rate": 2.3735512156067328e-05,
489
+ "loss": 0.0197,
490
+ "step": 305
491
+ },
492
+ {
493
+ "epoch": 6.46,
494
+ "grad_norm": 0.13510240614414215,
495
+ "learning_rate": 2.308343764628526e-05,
496
+ "loss": 0.023,
497
+ "step": 310
498
+ },
499
+ {
500
+ "epoch": 6.56,
501
+ "grad_norm": 0.20159611105918884,
502
+ "learning_rate": 2.243136313650319e-05,
503
+ "loss": 0.2273,
504
+ "step": 315
505
+ },
506
+ {
507
+ "epoch": 6.67,
508
+ "grad_norm": 106.47545623779297,
509
+ "learning_rate": 2.177928862672112e-05,
510
+ "loss": 0.3449,
511
+ "step": 320
512
+ },
513
+ {
514
+ "epoch": 6.77,
515
+ "grad_norm": 0.2443103939294815,
516
+ "learning_rate": 2.1127214116939053e-05,
517
+ "loss": 0.3164,
518
+ "step": 325
519
+ },
520
+ {
521
+ "epoch": 6.88,
522
+ "grad_norm": 19.54550552368164,
523
+ "learning_rate": 2.0475139607156984e-05,
524
+ "loss": 0.0099,
525
+ "step": 330
526
+ },
527
+ {
528
+ "epoch": 6.98,
529
+ "grad_norm": 0.08734071254730225,
530
+ "learning_rate": 1.9823065097374912e-05,
531
+ "loss": 0.2767,
532
+ "step": 335
533
+ },
534
+ {
535
+ "epoch": 7.0,
536
+ "eval_f1": 0.7578947368421053,
537
+ "eval_loss": 1.8833692073822021,
538
+ "eval_runtime": 1.3822,
539
+ "eval_samples_per_second": 46.303,
540
+ "eval_steps_per_second": 5.788,
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": 5.6339237645170805e-05,
554
+ "per_device_train_batch_size": 4
555
+ }
556
+ }
run-12/checkpoint-336/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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-12/checkpoint-384/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c52ac05124e3456aeeed025037b8df2daf555ceaa8751d40c7ac779c709cd1e
3
+ size 94763496
run-12/checkpoint-384/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce2fd05edbaf36ed4b62c69e5797f13aa3dae16f759a02cacd0c3ff703dd1a21
3
+ size 189552570
run-12/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-12/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-12/checkpoint-384/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee4d64b72a2ca5dc39026024619a44349bc69713d7f3ca04f4ffcb12da95046c
3
+ size 1064
run-12/checkpoint-384/trainer_state.json ADDED
@@ -0,0 +1,628 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7865168539325843,
3
+ "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/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": 1.2605946063995361,
14
+ "learning_rate": 5.868670588038625e-06,
15
+ "loss": 0.7019,
16
+ "step": 5
17
+ },
18
+ {
19
+ "epoch": 0.21,
20
+ "grad_norm": 1.1554573774337769,
21
+ "learning_rate": 1.173734117607725e-05,
22
+ "loss": 0.6822,
23
+ "step": 10
24
+ },
25
+ {
26
+ "epoch": 0.31,
27
+ "grad_norm": 1.2631292343139648,
28
+ "learning_rate": 1.7606011764115876e-05,
29
+ "loss": 0.6823,
30
+ "step": 15
31
+ },
32
+ {
33
+ "epoch": 0.42,
34
+ "grad_norm": 0.8986643552780151,
35
+ "learning_rate": 2.34746823521545e-05,
36
+ "loss": 0.6573,
37
+ "step": 20
38
+ },
39
+ {
40
+ "epoch": 0.52,
41
+ "grad_norm": 0.8833038210868835,
42
+ "learning_rate": 2.934335294019313e-05,
43
+ "loss": 0.6302,
44
+ "step": 25
45
+ },
46
+ {
47
+ "epoch": 0.62,
48
+ "grad_norm": 0.6974115967750549,
49
+ "learning_rate": 3.521202352823175e-05,
50
+ "loss": 0.6181,
51
+ "step": 30
52
+ },
53
+ {
54
+ "epoch": 0.73,
55
+ "grad_norm": 2.8611502647399902,
56
+ "learning_rate": 4.1080694116270374e-05,
57
+ "loss": 0.5431,
58
+ "step": 35
59
+ },
60
+ {
61
+ "epoch": 0.83,
62
+ "grad_norm": 1.6229726076126099,
63
+ "learning_rate": 4.6949364704309e-05,
64
+ "loss": 0.6722,
65
+ "step": 40
66
+ },
67
+ {
68
+ "epoch": 0.94,
69
+ "grad_norm": 3.6481218338012695,
70
+ "learning_rate": 5.281803529234763e-05,
71
+ "loss": 0.6342,
72
+ "step": 45
73
+ },
74
+ {
75
+ "epoch": 1.0,
76
+ "eval_f1": 0.7326732673267327,
77
+ "eval_loss": 0.7596015930175781,
78
+ "eval_runtime": 1.3819,
79
+ "eval_samples_per_second": 46.313,
80
+ "eval_steps_per_second": 5.789,
81
+ "step": 48
82
+ },
83
+ {
84
+ "epoch": 1.04,
85
+ "grad_norm": 4.841191291809082,
86
+ "learning_rate": 5.607840784125798e-05,
87
+ "loss": 0.7106,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 1.15,
92
+ "grad_norm": 0.9242434501647949,
93
+ "learning_rate": 5.542633333147591e-05,
94
+ "loss": 0.703,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 1.25,
99
+ "grad_norm": 1.9776262044906616,
100
+ "learning_rate": 5.4774258821693835e-05,
101
+ "loss": 0.4098,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 1.35,
106
+ "grad_norm": 0.7563148140907288,
107
+ "learning_rate": 5.412218431191177e-05,
108
+ "loss": 0.5599,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 1.46,
113
+ "grad_norm": 1.44520103931427,
114
+ "learning_rate": 5.34701098021297e-05,
115
+ "loss": 0.4823,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 1.56,
120
+ "grad_norm": Infinity,
121
+ "learning_rate": 5.294845019430404e-05,
122
+ "loss": 0.7782,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 1.67,
127
+ "grad_norm": 2.0363686084747314,
128
+ "learning_rate": 5.229637568452197e-05,
129
+ "loss": 0.6186,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 1.77,
134
+ "grad_norm": 1.6177257299423218,
135
+ "learning_rate": 5.16443011747399e-05,
136
+ "loss": 0.5907,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 1.88,
141
+ "grad_norm": 1.8490890264511108,
142
+ "learning_rate": 5.0992226664957835e-05,
143
+ "loss": 0.5909,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 1.98,
148
+ "grad_norm": 6.961443901062012,
149
+ "learning_rate": 5.034015215517576e-05,
150
+ "loss": 0.603,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 2.0,
155
+ "eval_f1": 0.74,
156
+ "eval_loss": 0.6603889465332031,
157
+ "eval_runtime": 1.4385,
158
+ "eval_samples_per_second": 44.492,
159
+ "eval_steps_per_second": 5.561,
160
+ "step": 96
161
+ },
162
+ {
163
+ "epoch": 2.08,
164
+ "grad_norm": 1.232475757598877,
165
+ "learning_rate": 4.981849254735011e-05,
166
+ "loss": 0.572,
167
+ "step": 100
168
+ },
169
+ {
170
+ "epoch": 2.19,
171
+ "grad_norm": 1.0792688131332397,
172
+ "learning_rate": 4.9166418037568046e-05,
173
+ "loss": 0.4763,
174
+ "step": 105
175
+ },
176
+ {
177
+ "epoch": 2.29,
178
+ "grad_norm": 1.4945162534713745,
179
+ "learning_rate": 4.8514343527785974e-05,
180
+ "loss": 0.5616,
181
+ "step": 110
182
+ },
183
+ {
184
+ "epoch": 2.4,
185
+ "grad_norm": 1.323120355606079,
186
+ "learning_rate": 4.78622690180039e-05,
187
+ "loss": 0.5489,
188
+ "step": 115
189
+ },
190
+ {
191
+ "epoch": 2.5,
192
+ "grad_norm": 5.674253463745117,
193
+ "learning_rate": 4.721019450822183e-05,
194
+ "loss": 0.5401,
195
+ "step": 120
196
+ },
197
+ {
198
+ "epoch": 2.6,
199
+ "grad_norm": 4.727626323699951,
200
+ "learning_rate": 4.668853490039618e-05,
201
+ "loss": 0.4278,
202
+ "step": 125
203
+ },
204
+ {
205
+ "epoch": 2.71,
206
+ "grad_norm": 8.496376037597656,
207
+ "learning_rate": 4.6036460390614105e-05,
208
+ "loss": 0.6216,
209
+ "step": 130
210
+ },
211
+ {
212
+ "epoch": 2.81,
213
+ "grad_norm": 2.796064615249634,
214
+ "learning_rate": 4.538438588083204e-05,
215
+ "loss": 0.3347,
216
+ "step": 135
217
+ },
218
+ {
219
+ "epoch": 2.92,
220
+ "grad_norm": 7.401052951812744,
221
+ "learning_rate": 4.486272627300638e-05,
222
+ "loss": 0.3109,
223
+ "step": 140
224
+ },
225
+ {
226
+ "epoch": 3.0,
227
+ "eval_f1": 0.72,
228
+ "eval_loss": 0.8806066513061523,
229
+ "eval_runtime": 1.3788,
230
+ "eval_samples_per_second": 46.418,
231
+ "eval_steps_per_second": 5.802,
232
+ "step": 144
233
+ },
234
+ {
235
+ "epoch": 3.02,
236
+ "grad_norm": 3.1919455528259277,
237
+ "learning_rate": 4.421065176322431e-05,
238
+ "loss": 0.485,
239
+ "step": 145
240
+ },
241
+ {
242
+ "epoch": 3.12,
243
+ "grad_norm": 56.47049331665039,
244
+ "learning_rate": 4.355857725344224e-05,
245
+ "loss": 0.3607,
246
+ "step": 150
247
+ },
248
+ {
249
+ "epoch": 3.23,
250
+ "grad_norm": 4.917665481567383,
251
+ "learning_rate": 4.290650274366017e-05,
252
+ "loss": 0.2077,
253
+ "step": 155
254
+ },
255
+ {
256
+ "epoch": 3.33,
257
+ "grad_norm": 8.88060188293457,
258
+ "learning_rate": 4.238484313583451e-05,
259
+ "loss": 0.6496,
260
+ "step": 160
261
+ },
262
+ {
263
+ "epoch": 3.44,
264
+ "grad_norm": 1.9647566080093384,
265
+ "learning_rate": 4.1732768626052447e-05,
266
+ "loss": 0.1394,
267
+ "step": 165
268
+ },
269
+ {
270
+ "epoch": 3.54,
271
+ "grad_norm": 50.96482849121094,
272
+ "learning_rate": 4.1080694116270374e-05,
273
+ "loss": 0.2834,
274
+ "step": 170
275
+ },
276
+ {
277
+ "epoch": 3.65,
278
+ "grad_norm": 132.5663299560547,
279
+ "learning_rate": 4.042861960648831e-05,
280
+ "loss": 0.485,
281
+ "step": 175
282
+ },
283
+ {
284
+ "epoch": 3.75,
285
+ "grad_norm": 3.1624388694763184,
286
+ "learning_rate": 3.977654509670624e-05,
287
+ "loss": 0.2353,
288
+ "step": 180
289
+ },
290
+ {
291
+ "epoch": 3.85,
292
+ "grad_norm": 0.5671400427818298,
293
+ "learning_rate": 3.912447058692417e-05,
294
+ "loss": 0.3058,
295
+ "step": 185
296
+ },
297
+ {
298
+ "epoch": 3.96,
299
+ "grad_norm": 23.86330223083496,
300
+ "learning_rate": 3.84723960771421e-05,
301
+ "loss": 0.3764,
302
+ "step": 190
303
+ },
304
+ {
305
+ "epoch": 4.0,
306
+ "eval_f1": 0.7500000000000001,
307
+ "eval_loss": 1.1437296867370605,
308
+ "eval_runtime": 1.4019,
309
+ "eval_samples_per_second": 45.651,
310
+ "eval_steps_per_second": 5.706,
311
+ "step": 192
312
+ },
313
+ {
314
+ "epoch": 4.06,
315
+ "grad_norm": 0.443531334400177,
316
+ "learning_rate": 3.782032156736003e-05,
317
+ "loss": 0.2652,
318
+ "step": 195
319
+ },
320
+ {
321
+ "epoch": 4.17,
322
+ "grad_norm": 5.118373394012451,
323
+ "learning_rate": 3.716824705757796e-05,
324
+ "loss": 0.3585,
325
+ "step": 200
326
+ },
327
+ {
328
+ "epoch": 4.27,
329
+ "grad_norm": 0.7104140520095825,
330
+ "learning_rate": 3.651617254779589e-05,
331
+ "loss": 0.2533,
332
+ "step": 205
333
+ },
334
+ {
335
+ "epoch": 4.38,
336
+ "grad_norm": Infinity,
337
+ "learning_rate": 3.599451293997023e-05,
338
+ "loss": 0.1963,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 4.48,
343
+ "grad_norm": 1.0030689239501953,
344
+ "learning_rate": 3.5342438430188165e-05,
345
+ "loss": 0.0602,
346
+ "step": 215
347
+ },
348
+ {
349
+ "epoch": 4.58,
350
+ "grad_norm": 193.33819580078125,
351
+ "learning_rate": 3.469036392040609e-05,
352
+ "loss": 0.2754,
353
+ "step": 220
354
+ },
355
+ {
356
+ "epoch": 4.69,
357
+ "grad_norm": 1.3048127889633179,
358
+ "learning_rate": 3.403828941062403e-05,
359
+ "loss": 0.1878,
360
+ "step": 225
361
+ },
362
+ {
363
+ "epoch": 4.79,
364
+ "grad_norm": 0.18795377016067505,
365
+ "learning_rate": 3.3386214900841956e-05,
366
+ "loss": 0.0245,
367
+ "step": 230
368
+ },
369
+ {
370
+ "epoch": 4.9,
371
+ "grad_norm": 0.7579631805419922,
372
+ "learning_rate": 3.2864555293016304e-05,
373
+ "loss": 0.3502,
374
+ "step": 235
375
+ },
376
+ {
377
+ "epoch": 5.0,
378
+ "grad_norm": 0.39707404375076294,
379
+ "learning_rate": 3.221248078323424e-05,
380
+ "loss": 0.2378,
381
+ "step": 240
382
+ },
383
+ {
384
+ "epoch": 5.0,
385
+ "eval_f1": 0.7865168539325843,
386
+ "eval_loss": 1.2380468845367432,
387
+ "eval_runtime": 1.3627,
388
+ "eval_samples_per_second": 46.967,
389
+ "eval_steps_per_second": 5.871,
390
+ "step": 240
391
+ },
392
+ {
393
+ "epoch": 5.1,
394
+ "grad_norm": 0.18084272742271423,
395
+ "learning_rate": 3.1560406273452166e-05,
396
+ "loss": 0.0143,
397
+ "step": 245
398
+ },
399
+ {
400
+ "epoch": 5.21,
401
+ "grad_norm": 0.6576858758926392,
402
+ "learning_rate": 3.0908331763670094e-05,
403
+ "loss": 0.0151,
404
+ "step": 250
405
+ },
406
+ {
407
+ "epoch": 5.31,
408
+ "grad_norm": 1.1509562730789185,
409
+ "learning_rate": 3.025625725388803e-05,
410
+ "loss": 0.3467,
411
+ "step": 255
412
+ },
413
+ {
414
+ "epoch": 5.42,
415
+ "grad_norm": 29.21381187438965,
416
+ "learning_rate": 2.9604182744105953e-05,
417
+ "loss": 0.0524,
418
+ "step": 260
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "grad_norm": 0.1710425168275833,
423
+ "learning_rate": 2.8952108234323884e-05,
424
+ "loss": 0.0106,
425
+ "step": 265
426
+ },
427
+ {
428
+ "epoch": 5.62,
429
+ "grad_norm": 0.14679908752441406,
430
+ "learning_rate": 2.8300033724541812e-05,
431
+ "loss": 0.0411,
432
+ "step": 270
433
+ },
434
+ {
435
+ "epoch": 5.73,
436
+ "grad_norm": 85.6775894165039,
437
+ "learning_rate": 2.7647959214759747e-05,
438
+ "loss": 0.4682,
439
+ "step": 275
440
+ },
441
+ {
442
+ "epoch": 5.83,
443
+ "grad_norm": 0.12999875843524933,
444
+ "learning_rate": 2.6995884704977678e-05,
445
+ "loss": 0.2671,
446
+ "step": 280
447
+ },
448
+ {
449
+ "epoch": 5.94,
450
+ "grad_norm": 23.26946258544922,
451
+ "learning_rate": 2.634381019519561e-05,
452
+ "loss": 0.2224,
453
+ "step": 285
454
+ },
455
+ {
456
+ "epoch": 6.0,
457
+ "eval_f1": 0.7741935483870968,
458
+ "eval_loss": 1.5522700548171997,
459
+ "eval_runtime": 1.3653,
460
+ "eval_samples_per_second": 46.878,
461
+ "eval_steps_per_second": 5.86,
462
+ "step": 288
463
+ },
464
+ {
465
+ "epoch": 6.04,
466
+ "grad_norm": 0.12312161922454834,
467
+ "learning_rate": 2.569173568541354e-05,
468
+ "loss": 0.2473,
469
+ "step": 290
470
+ },
471
+ {
472
+ "epoch": 6.15,
473
+ "grad_norm": 0.21050110459327698,
474
+ "learning_rate": 2.503966117563147e-05,
475
+ "loss": 0.0108,
476
+ "step": 295
477
+ },
478
+ {
479
+ "epoch": 6.25,
480
+ "grad_norm": 0.1071925163269043,
481
+ "learning_rate": 2.43875866658494e-05,
482
+ "loss": 0.0104,
483
+ "step": 300
484
+ },
485
+ {
486
+ "epoch": 6.35,
487
+ "grad_norm": 0.5196150541305542,
488
+ "learning_rate": 2.3735512156067328e-05,
489
+ "loss": 0.0197,
490
+ "step": 305
491
+ },
492
+ {
493
+ "epoch": 6.46,
494
+ "grad_norm": 0.13510240614414215,
495
+ "learning_rate": 2.308343764628526e-05,
496
+ "loss": 0.023,
497
+ "step": 310
498
+ },
499
+ {
500
+ "epoch": 6.56,
501
+ "grad_norm": 0.20159611105918884,
502
+ "learning_rate": 2.243136313650319e-05,
503
+ "loss": 0.2273,
504
+ "step": 315
505
+ },
506
+ {
507
+ "epoch": 6.67,
508
+ "grad_norm": 106.47545623779297,
509
+ "learning_rate": 2.177928862672112e-05,
510
+ "loss": 0.3449,
511
+ "step": 320
512
+ },
513
+ {
514
+ "epoch": 6.77,
515
+ "grad_norm": 0.2443103939294815,
516
+ "learning_rate": 2.1127214116939053e-05,
517
+ "loss": 0.3164,
518
+ "step": 325
519
+ },
520
+ {
521
+ "epoch": 6.88,
522
+ "grad_norm": 19.54550552368164,
523
+ "learning_rate": 2.0475139607156984e-05,
524
+ "loss": 0.0099,
525
+ "step": 330
526
+ },
527
+ {
528
+ "epoch": 6.98,
529
+ "grad_norm": 0.08734071254730225,
530
+ "learning_rate": 1.9823065097374912e-05,
531
+ "loss": 0.2767,
532
+ "step": 335
533
+ },
534
+ {
535
+ "epoch": 7.0,
536
+ "eval_f1": 0.7578947368421053,
537
+ "eval_loss": 1.8833692073822021,
538
+ "eval_runtime": 1.3822,
539
+ "eval_samples_per_second": 46.303,
540
+ "eval_steps_per_second": 5.788,
541
+ "step": 336
542
+ },
543
+ {
544
+ "epoch": 7.08,
545
+ "grad_norm": 19.562559127807617,
546
+ "learning_rate": 1.9170990587592843e-05,
547
+ "loss": 0.2792,
548
+ "step": 340
549
+ },
550
+ {
551
+ "epoch": 7.19,
552
+ "grad_norm": 0.08157272636890411,
553
+ "learning_rate": 1.8518916077810774e-05,
554
+ "loss": 0.031,
555
+ "step": 345
556
+ },
557
+ {
558
+ "epoch": 7.29,
559
+ "grad_norm": 0.14632558822631836,
560
+ "learning_rate": 1.7866841568028706e-05,
561
+ "loss": 0.0108,
562
+ "step": 350
563
+ },
564
+ {
565
+ "epoch": 7.4,
566
+ "grad_norm": 0.40814200043678284,
567
+ "learning_rate": 1.7214767058246637e-05,
568
+ "loss": 0.0057,
569
+ "step": 355
570
+ },
571
+ {
572
+ "epoch": 7.5,
573
+ "grad_norm": 0.07355981320142746,
574
+ "learning_rate": 1.6562692548464565e-05,
575
+ "loss": 0.0056,
576
+ "step": 360
577
+ },
578
+ {
579
+ "epoch": 7.6,
580
+ "grad_norm": 0.10633735358715057,
581
+ "learning_rate": 1.5910618038682496e-05,
582
+ "loss": 0.005,
583
+ "step": 365
584
+ },
585
+ {
586
+ "epoch": 7.71,
587
+ "grad_norm": 0.4227176308631897,
588
+ "learning_rate": 1.5258543528900426e-05,
589
+ "loss": 0.0337,
590
+ "step": 370
591
+ },
592
+ {
593
+ "epoch": 7.81,
594
+ "grad_norm": 0.06743155419826508,
595
+ "learning_rate": 1.4606469019118355e-05,
596
+ "loss": 0.2789,
597
+ "step": 375
598
+ },
599
+ {
600
+ "epoch": 7.92,
601
+ "grad_norm": 0.07822620868682861,
602
+ "learning_rate": 1.3954394509336287e-05,
603
+ "loss": 0.0048,
604
+ "step": 380
605
+ },
606
+ {
607
+ "epoch": 8.0,
608
+ "eval_f1": 0.782608695652174,
609
+ "eval_loss": 1.4861700534820557,
610
+ "eval_runtime": 1.3792,
611
+ "eval_samples_per_second": 46.405,
612
+ "eval_steps_per_second": 5.801,
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": 5.6339237645170805e-05,
626
+ "per_device_train_batch_size": 4
627
+ }
628
+ }
run-12/checkpoint-384/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a13ddf2852b749b2ed775695ad5ddfecc242a8e8d8b1546d5e7357434a3ed37
3
+ size 4920
run-12/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
+ }