File size: 25,623 Bytes
67f6fea
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
2023-10-10 18:16:34,319 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,322 Model: "SequenceTagger(
  (embeddings): ByT5Embeddings(
    (model): T5EncoderModel(
      (shared): Embedding(384, 1472)
      (encoder): T5Stack(
        (embed_tokens): Embedding(384, 1472)
        (block): ModuleList(
          (0): T5Block(
            (layer): ModuleList(
              (0): T5LayerSelfAttention(
                (SelfAttention): T5Attention(
                  (q): Linear(in_features=1472, out_features=384, bias=False)
                  (k): Linear(in_features=1472, out_features=384, bias=False)
                  (v): Linear(in_features=1472, out_features=384, bias=False)
                  (o): Linear(in_features=384, out_features=1472, bias=False)
                  (relative_attention_bias): Embedding(32, 6)
                )
                (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
                (dropout): Dropout(p=0.1, inplace=False)
              )
              (1): T5LayerFF(
                (DenseReluDense): T5DenseGatedActDense(
                  (wi_0): Linear(in_features=1472, out_features=3584, bias=False)
                  (wi_1): Linear(in_features=1472, out_features=3584, bias=False)
                  (wo): Linear(in_features=3584, out_features=1472, bias=False)
                  (dropout): Dropout(p=0.1, inplace=False)
                  (act): NewGELUActivation()
                )
                (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
                (dropout): Dropout(p=0.1, inplace=False)
              )
            )
          )
          (1-11): 11 x T5Block(
            (layer): ModuleList(
              (0): T5LayerSelfAttention(
                (SelfAttention): T5Attention(
                  (q): Linear(in_features=1472, out_features=384, bias=False)
                  (k): Linear(in_features=1472, out_features=384, bias=False)
                  (v): Linear(in_features=1472, out_features=384, bias=False)
                  (o): Linear(in_features=384, out_features=1472, bias=False)
                )
                (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
                (dropout): Dropout(p=0.1, inplace=False)
              )
              (1): T5LayerFF(
                (DenseReluDense): T5DenseGatedActDense(
                  (wi_0): Linear(in_features=1472, out_features=3584, bias=False)
                  (wi_1): Linear(in_features=1472, out_features=3584, bias=False)
                  (wo): Linear(in_features=3584, out_features=1472, bias=False)
                  (dropout): Dropout(p=0.1, inplace=False)
                  (act): NewGELUActivation()
                )
                (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
                (dropout): Dropout(p=0.1, inplace=False)
              )
            )
          )
        )
        (final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
        (dropout): Dropout(p=0.1, inplace=False)
      )
    )
  )
  (locked_dropout): LockedDropout(p=0.5)
  (linear): Linear(in_features=1472, out_features=17, bias=True)
  (loss_function): CrossEntropyLoss()
)"
2023-10-10 18:16:34,322 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,322 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
 - NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
2023-10-10 18:16:34,322 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,322 Train:  20847 sentences
2023-10-10 18:16:34,322         (train_with_dev=False, train_with_test=False)
2023-10-10 18:16:34,322 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,322 Training Params:
2023-10-10 18:16:34,322  - learning_rate: "0.00015" 
2023-10-10 18:16:34,323  - mini_batch_size: "4"
2023-10-10 18:16:34,323  - max_epochs: "10"
2023-10-10 18:16:34,323  - shuffle: "True"
2023-10-10 18:16:34,323 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,323 Plugins:
2023-10-10 18:16:34,323  - TensorboardLogger
2023-10-10 18:16:34,323  - LinearScheduler | warmup_fraction: '0.1'
2023-10-10 18:16:34,323 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,323 Final evaluation on model from best epoch (best-model.pt)
2023-10-10 18:16:34,323  - metric: "('micro avg', 'f1-score')"
2023-10-10 18:16:34,323 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,323 Computation:
2023-10-10 18:16:34,323  - compute on device: cuda:0
2023-10-10 18:16:34,323  - embedding storage: none
2023-10-10 18:16:34,323 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,324 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-2"
2023-10-10 18:16:34,324 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,324 ----------------------------------------------------------------------------------------------------
2023-10-10 18:16:34,324 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-10 18:19:07,343 epoch 1 - iter 521/5212 - loss 2.81016252 - time (sec): 153.02 - samples/sec: 234.74 - lr: 0.000015 - momentum: 0.000000
2023-10-10 18:21:43,109 epoch 1 - iter 1042/5212 - loss 2.37850019 - time (sec): 308.78 - samples/sec: 233.59 - lr: 0.000030 - momentum: 0.000000
2023-10-10 18:24:21,241 epoch 1 - iter 1563/5212 - loss 1.84782296 - time (sec): 466.92 - samples/sec: 234.30 - lr: 0.000045 - momentum: 0.000000
2023-10-10 18:27:01,083 epoch 1 - iter 2084/5212 - loss 1.48063193 - time (sec): 626.76 - samples/sec: 237.27 - lr: 0.000060 - momentum: 0.000000
2023-10-10 18:29:38,764 epoch 1 - iter 2605/5212 - loss 1.27253415 - time (sec): 784.44 - samples/sec: 236.54 - lr: 0.000075 - momentum: 0.000000
2023-10-10 18:32:17,153 epoch 1 - iter 3126/5212 - loss 1.11681269 - time (sec): 942.83 - samples/sec: 237.00 - lr: 0.000090 - momentum: 0.000000
2023-10-10 18:34:54,795 epoch 1 - iter 3647/5212 - loss 1.00835432 - time (sec): 1100.47 - samples/sec: 236.27 - lr: 0.000105 - momentum: 0.000000
2023-10-10 18:37:32,394 epoch 1 - iter 4168/5212 - loss 0.91865937 - time (sec): 1258.07 - samples/sec: 234.97 - lr: 0.000120 - momentum: 0.000000
2023-10-10 18:40:00,428 epoch 1 - iter 4689/5212 - loss 0.84845181 - time (sec): 1406.10 - samples/sec: 234.64 - lr: 0.000135 - momentum: 0.000000
2023-10-10 18:42:26,316 epoch 1 - iter 5210/5212 - loss 0.78330933 - time (sec): 1551.99 - samples/sec: 236.66 - lr: 0.000150 - momentum: 0.000000
2023-10-10 18:42:26,799 ----------------------------------------------------------------------------------------------------
2023-10-10 18:42:26,800 EPOCH 1 done: loss 0.7830 - lr: 0.000150
2023-10-10 18:43:02,191 DEV : loss 0.14865536987781525 - f1-score (micro avg)  0.2591
2023-10-10 18:43:02,244 saving best model
2023-10-10 18:43:03,204 ----------------------------------------------------------------------------------------------------
2023-10-10 18:45:23,109 epoch 2 - iter 521/5212 - loss 0.18267050 - time (sec): 139.90 - samples/sec: 250.67 - lr: 0.000148 - momentum: 0.000000
2023-10-10 18:47:50,051 epoch 2 - iter 1042/5212 - loss 0.18153459 - time (sec): 286.84 - samples/sec: 247.64 - lr: 0.000147 - momentum: 0.000000
2023-10-10 18:50:09,700 epoch 2 - iter 1563/5212 - loss 0.17717219 - time (sec): 426.49 - samples/sec: 252.57 - lr: 0.000145 - momentum: 0.000000
2023-10-10 18:52:31,360 epoch 2 - iter 2084/5212 - loss 0.17309531 - time (sec): 568.15 - samples/sec: 251.18 - lr: 0.000143 - momentum: 0.000000
2023-10-10 18:54:52,430 epoch 2 - iter 2605/5212 - loss 0.16981704 - time (sec): 709.22 - samples/sec: 253.68 - lr: 0.000142 - momentum: 0.000000
2023-10-10 18:57:12,245 epoch 2 - iter 3126/5212 - loss 0.16298321 - time (sec): 849.04 - samples/sec: 256.48 - lr: 0.000140 - momentum: 0.000000
2023-10-10 18:59:31,681 epoch 2 - iter 3647/5212 - loss 0.16081033 - time (sec): 988.47 - samples/sec: 258.82 - lr: 0.000138 - momentum: 0.000000
2023-10-10 19:01:48,864 epoch 2 - iter 4168/5212 - loss 0.15845681 - time (sec): 1125.66 - samples/sec: 258.75 - lr: 0.000137 - momentum: 0.000000
2023-10-10 19:04:06,270 epoch 2 - iter 4689/5212 - loss 0.15711708 - time (sec): 1263.06 - samples/sec: 258.81 - lr: 0.000135 - momentum: 0.000000
2023-10-10 19:06:27,101 epoch 2 - iter 5210/5212 - loss 0.15390781 - time (sec): 1403.89 - samples/sec: 261.55 - lr: 0.000133 - momentum: 0.000000
2023-10-10 19:06:27,658 ----------------------------------------------------------------------------------------------------
2023-10-10 19:06:27,658 EPOCH 2 done: loss 0.1539 - lr: 0.000133
2023-10-10 19:07:07,419 DEV : loss 0.12877270579338074 - f1-score (micro avg)  0.3352
2023-10-10 19:07:07,472 saving best model
2023-10-10 19:07:10,082 ----------------------------------------------------------------------------------------------------
2023-10-10 19:09:26,104 epoch 3 - iter 521/5212 - loss 0.09359735 - time (sec): 136.02 - samples/sec: 262.05 - lr: 0.000132 - momentum: 0.000000
2023-10-10 19:11:43,039 epoch 3 - iter 1042/5212 - loss 0.09901175 - time (sec): 272.95 - samples/sec: 266.45 - lr: 0.000130 - momentum: 0.000000
2023-10-10 19:14:01,365 epoch 3 - iter 1563/5212 - loss 0.10198032 - time (sec): 411.28 - samples/sec: 267.17 - lr: 0.000128 - momentum: 0.000000
2023-10-10 19:16:17,180 epoch 3 - iter 2084/5212 - loss 0.10644730 - time (sec): 547.09 - samples/sec: 266.24 - lr: 0.000127 - momentum: 0.000000
2023-10-10 19:18:36,955 epoch 3 - iter 2605/5212 - loss 0.10979467 - time (sec): 686.87 - samples/sec: 270.23 - lr: 0.000125 - momentum: 0.000000
2023-10-10 19:20:57,796 epoch 3 - iter 3126/5212 - loss 0.10703931 - time (sec): 827.71 - samples/sec: 269.96 - lr: 0.000123 - momentum: 0.000000
2023-10-10 19:23:16,304 epoch 3 - iter 3647/5212 - loss 0.10648081 - time (sec): 966.22 - samples/sec: 269.63 - lr: 0.000122 - momentum: 0.000000
2023-10-10 19:25:35,494 epoch 3 - iter 4168/5212 - loss 0.10571107 - time (sec): 1105.41 - samples/sec: 268.40 - lr: 0.000120 - momentum: 0.000000
2023-10-10 19:27:52,717 epoch 3 - iter 4689/5212 - loss 0.10470171 - time (sec): 1242.63 - samples/sec: 265.47 - lr: 0.000118 - momentum: 0.000000
2023-10-10 19:30:15,588 epoch 3 - iter 5210/5212 - loss 0.10374270 - time (sec): 1385.50 - samples/sec: 265.05 - lr: 0.000117 - momentum: 0.000000
2023-10-10 19:30:16,133 ----------------------------------------------------------------------------------------------------
2023-10-10 19:30:16,133 EPOCH 3 done: loss 0.1037 - lr: 0.000117
2023-10-10 19:30:54,988 DEV : loss 0.17644649744033813 - f1-score (micro avg)  0.3625
2023-10-10 19:30:55,039 saving best model
2023-10-10 19:30:57,643 ----------------------------------------------------------------------------------------------------
2023-10-10 19:33:24,242 epoch 4 - iter 521/5212 - loss 0.06203084 - time (sec): 146.59 - samples/sec: 251.99 - lr: 0.000115 - momentum: 0.000000
2023-10-10 19:35:52,103 epoch 4 - iter 1042/5212 - loss 0.07415039 - time (sec): 294.45 - samples/sec: 260.15 - lr: 0.000113 - momentum: 0.000000
2023-10-10 19:38:15,428 epoch 4 - iter 1563/5212 - loss 0.07137797 - time (sec): 437.78 - samples/sec: 257.60 - lr: 0.000112 - momentum: 0.000000
2023-10-10 19:40:37,400 epoch 4 - iter 2084/5212 - loss 0.07121646 - time (sec): 579.75 - samples/sec: 256.34 - lr: 0.000110 - momentum: 0.000000
2023-10-10 19:43:03,829 epoch 4 - iter 2605/5212 - loss 0.06855636 - time (sec): 726.18 - samples/sec: 255.84 - lr: 0.000108 - momentum: 0.000000
2023-10-10 19:45:28,588 epoch 4 - iter 3126/5212 - loss 0.07215692 - time (sec): 870.94 - samples/sec: 255.84 - lr: 0.000107 - momentum: 0.000000
2023-10-10 19:47:53,752 epoch 4 - iter 3647/5212 - loss 0.07245324 - time (sec): 1016.10 - samples/sec: 256.59 - lr: 0.000105 - momentum: 0.000000
2023-10-10 19:50:17,550 epoch 4 - iter 4168/5212 - loss 0.07195821 - time (sec): 1159.90 - samples/sec: 254.60 - lr: 0.000103 - momentum: 0.000000
2023-10-10 19:52:43,827 epoch 4 - iter 4689/5212 - loss 0.07354350 - time (sec): 1306.18 - samples/sec: 254.26 - lr: 0.000102 - momentum: 0.000000
2023-10-10 19:55:06,401 epoch 4 - iter 5210/5212 - loss 0.07389688 - time (sec): 1448.75 - samples/sec: 253.55 - lr: 0.000100 - momentum: 0.000000
2023-10-10 19:55:06,857 ----------------------------------------------------------------------------------------------------
2023-10-10 19:55:06,857 EPOCH 4 done: loss 0.0739 - lr: 0.000100
2023-10-10 19:55:48,160 DEV : loss 0.3035148084163666 - f1-score (micro avg)  0.3425
2023-10-10 19:55:48,214 ----------------------------------------------------------------------------------------------------
2023-10-10 19:58:10,368 epoch 5 - iter 521/5212 - loss 0.04380137 - time (sec): 142.15 - samples/sec: 241.95 - lr: 0.000098 - momentum: 0.000000
2023-10-10 20:00:32,233 epoch 5 - iter 1042/5212 - loss 0.04814301 - time (sec): 284.02 - samples/sec: 247.92 - lr: 0.000097 - momentum: 0.000000
2023-10-10 20:02:56,247 epoch 5 - iter 1563/5212 - loss 0.04738308 - time (sec): 428.03 - samples/sec: 254.16 - lr: 0.000095 - momentum: 0.000000
2023-10-10 20:05:18,346 epoch 5 - iter 2084/5212 - loss 0.04616747 - time (sec): 570.13 - samples/sec: 254.93 - lr: 0.000093 - momentum: 0.000000
2023-10-10 20:07:43,212 epoch 5 - iter 2605/5212 - loss 0.04857278 - time (sec): 715.00 - samples/sec: 255.31 - lr: 0.000092 - momentum: 0.000000
2023-10-10 20:10:05,909 epoch 5 - iter 3126/5212 - loss 0.04886862 - time (sec): 857.69 - samples/sec: 255.88 - lr: 0.000090 - momentum: 0.000000
2023-10-10 20:12:25,629 epoch 5 - iter 3647/5212 - loss 0.04941233 - time (sec): 997.41 - samples/sec: 255.51 - lr: 0.000088 - momentum: 0.000000
2023-10-10 20:15:05,698 epoch 5 - iter 4168/5212 - loss 0.04979465 - time (sec): 1157.48 - samples/sec: 253.62 - lr: 0.000087 - momentum: 0.000000
2023-10-10 20:17:41,182 epoch 5 - iter 4689/5212 - loss 0.04912451 - time (sec): 1312.97 - samples/sec: 252.86 - lr: 0.000085 - momentum: 0.000000
2023-10-10 20:20:07,092 epoch 5 - iter 5210/5212 - loss 0.04860995 - time (sec): 1458.88 - samples/sec: 251.66 - lr: 0.000083 - momentum: 0.000000
2023-10-10 20:20:07,750 ----------------------------------------------------------------------------------------------------
2023-10-10 20:20:07,750 EPOCH 5 done: loss 0.0486 - lr: 0.000083
2023-10-10 20:20:50,848 DEV : loss 0.3323441743850708 - f1-score (micro avg)  0.3643
2023-10-10 20:20:50,907 saving best model
2023-10-10 20:20:53,520 ----------------------------------------------------------------------------------------------------
2023-10-10 20:23:09,350 epoch 6 - iter 521/5212 - loss 0.03801063 - time (sec): 135.83 - samples/sec: 258.05 - lr: 0.000082 - momentum: 0.000000
2023-10-10 20:25:25,462 epoch 6 - iter 1042/5212 - loss 0.03880207 - time (sec): 271.94 - samples/sec: 257.64 - lr: 0.000080 - momentum: 0.000000
2023-10-10 20:27:44,831 epoch 6 - iter 1563/5212 - loss 0.03989806 - time (sec): 411.31 - samples/sec: 263.80 - lr: 0.000078 - momentum: 0.000000
2023-10-10 20:30:04,726 epoch 6 - iter 2084/5212 - loss 0.04174066 - time (sec): 551.20 - samples/sec: 266.80 - lr: 0.000077 - momentum: 0.000000
2023-10-10 20:32:24,885 epoch 6 - iter 2605/5212 - loss 0.03948956 - time (sec): 691.36 - samples/sec: 269.01 - lr: 0.000075 - momentum: 0.000000
2023-10-10 20:34:43,523 epoch 6 - iter 3126/5212 - loss 0.03830296 - time (sec): 830.00 - samples/sec: 267.05 - lr: 0.000073 - momentum: 0.000000
2023-10-10 20:37:01,312 epoch 6 - iter 3647/5212 - loss 0.03849237 - time (sec): 967.79 - samples/sec: 266.77 - lr: 0.000072 - momentum: 0.000000
2023-10-10 20:39:20,384 epoch 6 - iter 4168/5212 - loss 0.03899918 - time (sec): 1106.86 - samples/sec: 266.65 - lr: 0.000070 - momentum: 0.000000
2023-10-10 20:41:38,051 epoch 6 - iter 4689/5212 - loss 0.03784107 - time (sec): 1244.53 - samples/sec: 264.93 - lr: 0.000068 - momentum: 0.000000
2023-10-10 20:43:57,783 epoch 6 - iter 5210/5212 - loss 0.03843205 - time (sec): 1384.26 - samples/sec: 265.36 - lr: 0.000067 - momentum: 0.000000
2023-10-10 20:43:58,230 ----------------------------------------------------------------------------------------------------
2023-10-10 20:43:58,230 EPOCH 6 done: loss 0.0384 - lr: 0.000067
2023-10-10 20:44:38,095 DEV : loss 0.3565690815448761 - f1-score (micro avg)  0.3714
2023-10-10 20:44:38,147 saving best model
2023-10-10 20:44:40,782 ----------------------------------------------------------------------------------------------------
2023-10-10 20:47:08,584 epoch 7 - iter 521/5212 - loss 0.02908204 - time (sec): 147.80 - samples/sec: 240.48 - lr: 0.000065 - momentum: 0.000000
2023-10-10 20:49:40,233 epoch 7 - iter 1042/5212 - loss 0.02884625 - time (sec): 299.45 - samples/sec: 238.89 - lr: 0.000063 - momentum: 0.000000
2023-10-10 20:52:06,698 epoch 7 - iter 1563/5212 - loss 0.02640536 - time (sec): 445.91 - samples/sec: 242.78 - lr: 0.000062 - momentum: 0.000000
2023-10-10 20:54:30,081 epoch 7 - iter 2084/5212 - loss 0.02749239 - time (sec): 589.29 - samples/sec: 245.46 - lr: 0.000060 - momentum: 0.000000
2023-10-10 20:56:53,383 epoch 7 - iter 2605/5212 - loss 0.02821275 - time (sec): 732.59 - samples/sec: 249.90 - lr: 0.000058 - momentum: 0.000000
2023-10-10 20:59:13,941 epoch 7 - iter 3126/5212 - loss 0.02790788 - time (sec): 873.15 - samples/sec: 251.10 - lr: 0.000057 - momentum: 0.000000
2023-10-10 21:01:41,509 epoch 7 - iter 3647/5212 - loss 0.02797348 - time (sec): 1020.72 - samples/sec: 250.96 - lr: 0.000055 - momentum: 0.000000
2023-10-10 21:04:04,892 epoch 7 - iter 4168/5212 - loss 0.02744857 - time (sec): 1164.10 - samples/sec: 249.39 - lr: 0.000053 - momentum: 0.000000
2023-10-10 21:06:29,203 epoch 7 - iter 4689/5212 - loss 0.02707634 - time (sec): 1308.42 - samples/sec: 250.75 - lr: 0.000052 - momentum: 0.000000
2023-10-10 21:08:56,201 epoch 7 - iter 5210/5212 - loss 0.02676611 - time (sec): 1455.41 - samples/sec: 252.36 - lr: 0.000050 - momentum: 0.000000
2023-10-10 21:08:56,684 ----------------------------------------------------------------------------------------------------
2023-10-10 21:08:56,684 EPOCH 7 done: loss 0.0268 - lr: 0.000050
2023-10-10 21:09:37,357 DEV : loss 0.3953941762447357 - f1-score (micro avg)  0.3828
2023-10-10 21:09:37,410 saving best model
2023-10-10 21:09:40,213 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:07,582 epoch 8 - iter 521/5212 - loss 0.01910833 - time (sec): 147.37 - samples/sec: 271.50 - lr: 0.000048 - momentum: 0.000000
2023-10-10 21:14:32,536 epoch 8 - iter 1042/5212 - loss 0.01909368 - time (sec): 292.32 - samples/sec: 264.69 - lr: 0.000047 - momentum: 0.000000
2023-10-10 21:16:51,098 epoch 8 - iter 1563/5212 - loss 0.01709210 - time (sec): 430.88 - samples/sec: 260.58 - lr: 0.000045 - momentum: 0.000000
2023-10-10 21:19:14,800 epoch 8 - iter 2084/5212 - loss 0.01718362 - time (sec): 574.58 - samples/sec: 261.23 - lr: 0.000043 - momentum: 0.000000
2023-10-10 21:21:38,710 epoch 8 - iter 2605/5212 - loss 0.01688276 - time (sec): 718.49 - samples/sec: 258.02 - lr: 0.000042 - momentum: 0.000000
2023-10-10 21:24:04,659 epoch 8 - iter 3126/5212 - loss 0.01713091 - time (sec): 864.44 - samples/sec: 258.08 - lr: 0.000040 - momentum: 0.000000
2023-10-10 21:26:29,475 epoch 8 - iter 3647/5212 - loss 0.01686813 - time (sec): 1009.26 - samples/sec: 256.90 - lr: 0.000038 - momentum: 0.000000
2023-10-10 21:28:50,236 epoch 8 - iter 4168/5212 - loss 0.01687351 - time (sec): 1150.02 - samples/sec: 255.80 - lr: 0.000037 - momentum: 0.000000
2023-10-10 21:31:09,057 epoch 8 - iter 4689/5212 - loss 0.01695639 - time (sec): 1288.84 - samples/sec: 254.16 - lr: 0.000035 - momentum: 0.000000
2023-10-10 21:33:31,826 epoch 8 - iter 5210/5212 - loss 0.01809776 - time (sec): 1431.61 - samples/sec: 256.63 - lr: 0.000033 - momentum: 0.000000
2023-10-10 21:33:32,225 ----------------------------------------------------------------------------------------------------
2023-10-10 21:33:32,225 EPOCH 8 done: loss 0.0181 - lr: 0.000033
2023-10-10 21:34:11,286 DEV : loss 0.43120914697647095 - f1-score (micro avg)  0.3696
2023-10-10 21:34:11,342 ----------------------------------------------------------------------------------------------------
2023-10-10 21:36:31,101 epoch 9 - iter 521/5212 - loss 0.01561923 - time (sec): 139.76 - samples/sec: 266.39 - lr: 0.000032 - momentum: 0.000000
2023-10-10 21:38:54,563 epoch 9 - iter 1042/5212 - loss 0.01332527 - time (sec): 283.22 - samples/sec: 273.46 - lr: 0.000030 - momentum: 0.000000
2023-10-10 21:41:13,645 epoch 9 - iter 1563/5212 - loss 0.01209082 - time (sec): 422.30 - samples/sec: 268.95 - lr: 0.000028 - momentum: 0.000000
2023-10-10 21:43:28,845 epoch 9 - iter 2084/5212 - loss 0.01241173 - time (sec): 557.50 - samples/sec: 263.32 - lr: 0.000027 - momentum: 0.000000
2023-10-10 21:45:49,437 epoch 9 - iter 2605/5212 - loss 0.01364082 - time (sec): 698.09 - samples/sec: 265.94 - lr: 0.000025 - momentum: 0.000000
2023-10-10 21:48:05,337 epoch 9 - iter 3126/5212 - loss 0.01364299 - time (sec): 833.99 - samples/sec: 264.53 - lr: 0.000023 - momentum: 0.000000
2023-10-10 21:50:21,661 epoch 9 - iter 3647/5212 - loss 0.01332216 - time (sec): 970.32 - samples/sec: 263.44 - lr: 0.000022 - momentum: 0.000000
2023-10-10 21:52:40,256 epoch 9 - iter 4168/5212 - loss 0.01347305 - time (sec): 1108.91 - samples/sec: 263.38 - lr: 0.000020 - momentum: 0.000000
2023-10-10 21:54:57,896 epoch 9 - iter 4689/5212 - loss 0.01371078 - time (sec): 1246.55 - samples/sec: 263.47 - lr: 0.000018 - momentum: 0.000000
2023-10-10 21:57:17,103 epoch 9 - iter 5210/5212 - loss 0.01386029 - time (sec): 1385.76 - samples/sec: 265.11 - lr: 0.000017 - momentum: 0.000000
2023-10-10 21:57:17,512 ----------------------------------------------------------------------------------------------------
2023-10-10 21:57:17,513 EPOCH 9 done: loss 0.0139 - lr: 0.000017
2023-10-10 21:57:57,009 DEV : loss 0.4479060769081116 - f1-score (micro avg)  0.3711
2023-10-10 21:57:57,061 ----------------------------------------------------------------------------------------------------
2023-10-10 22:00:15,416 epoch 10 - iter 521/5212 - loss 0.00629895 - time (sec): 138.35 - samples/sec: 276.50 - lr: 0.000015 - momentum: 0.000000
2023-10-10 22:02:36,634 epoch 10 - iter 1042/5212 - loss 0.00672524 - time (sec): 279.57 - samples/sec: 275.18 - lr: 0.000013 - momentum: 0.000000
2023-10-10 22:04:56,189 epoch 10 - iter 1563/5212 - loss 0.00663359 - time (sec): 419.13 - samples/sec: 269.04 - lr: 0.000012 - momentum: 0.000000
2023-10-10 22:07:12,353 epoch 10 - iter 2084/5212 - loss 0.00656745 - time (sec): 555.29 - samples/sec: 261.90 - lr: 0.000010 - momentum: 0.000000
2023-10-10 22:09:33,505 epoch 10 - iter 2605/5212 - loss 0.00650200 - time (sec): 696.44 - samples/sec: 262.57 - lr: 0.000008 - momentum: 0.000000
2023-10-10 22:11:53,364 epoch 10 - iter 3126/5212 - loss 0.00681067 - time (sec): 836.30 - samples/sec: 261.75 - lr: 0.000007 - momentum: 0.000000
2023-10-10 22:14:13,626 epoch 10 - iter 3647/5212 - loss 0.00705344 - time (sec): 976.56 - samples/sec: 262.86 - lr: 0.000005 - momentum: 0.000000
2023-10-10 22:16:35,112 epoch 10 - iter 4168/5212 - loss 0.00721008 - time (sec): 1118.05 - samples/sec: 264.27 - lr: 0.000003 - momentum: 0.000000
2023-10-10 22:18:53,877 epoch 10 - iter 4689/5212 - loss 0.00709267 - time (sec): 1256.81 - samples/sec: 263.70 - lr: 0.000002 - momentum: 0.000000
2023-10-10 22:21:11,081 epoch 10 - iter 5210/5212 - loss 0.00738418 - time (sec): 1394.02 - samples/sec: 263.35 - lr: 0.000000 - momentum: 0.000000
2023-10-10 22:21:11,710 ----------------------------------------------------------------------------------------------------
2023-10-10 22:21:11,710 EPOCH 10 done: loss 0.0074 - lr: 0.000000
2023-10-10 22:21:50,043 DEV : loss 0.47310149669647217 - f1-score (micro avg)  0.3745
2023-10-10 22:21:51,032 ----------------------------------------------------------------------------------------------------
2023-10-10 22:21:51,034 Loading model from best epoch ...
2023-10-10 22:21:56,334 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-10 22:23:37,448 
Results:
- F-score (micro) 0.4359
- F-score (macro) 0.2967
- Accuracy 0.2824

By class:
              precision    recall  f1-score   support

         LOC     0.4896    0.5066    0.4980      1214
         PER     0.4020    0.4567    0.4276       808
         ORG     0.2487    0.2748    0.2611       353
   HumanProd     0.0000    0.0000    0.0000        15

   micro avg     0.4206    0.4523    0.4359      2390
   macro avg     0.2851    0.3095    0.2967      2390
weighted avg     0.4213    0.4523    0.4361      2390

2023-10-10 22:23:37,448 ----------------------------------------------------------------------------------------------------