File size: 25,332 Bytes
51eee24 |
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 |
2023-10-10 19:39:45,665 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,667 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 19:39:45,667 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,667 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
- NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
2023-10-10 19:39:45,667 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,667 Train: 7142 sentences
2023-10-10 19:39:45,667 (train_with_dev=False, train_with_test=False)
2023-10-10 19:39:45,668 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,668 Training Params:
2023-10-10 19:39:45,668 - learning_rate: "0.00015"
2023-10-10 19:39:45,668 - mini_batch_size: "8"
2023-10-10 19:39:45,668 - max_epochs: "10"
2023-10-10 19:39:45,668 - shuffle: "True"
2023-10-10 19:39:45,668 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,668 Plugins:
2023-10-10 19:39:45,668 - TensorboardLogger
2023-10-10 19:39:45,668 - LinearScheduler | warmup_fraction: '0.1'
2023-10-10 19:39:45,668 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,668 Final evaluation on model from best epoch (best-model.pt)
2023-10-10 19:39:45,668 - metric: "('micro avg', 'f1-score')"
2023-10-10 19:39:45,668 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,669 Computation:
2023-10-10 19:39:45,669 - compute on device: cuda:0
2023-10-10 19:39:45,669 - embedding storage: none
2023-10-10 19:39:45,669 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,669 Model training base path: "hmbench-newseye/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-1"
2023-10-10 19:39:45,669 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,669 ----------------------------------------------------------------------------------------------------
2023-10-10 19:39:45,669 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-10 19:40:37,221 epoch 1 - iter 89/893 - loss 2.82906495 - time (sec): 51.55 - samples/sec: 494.24 - lr: 0.000015 - momentum: 0.000000
2023-10-10 19:41:26,623 epoch 1 - iter 178/893 - loss 2.78110436 - time (sec): 100.95 - samples/sec: 499.36 - lr: 0.000030 - momentum: 0.000000
2023-10-10 19:42:16,721 epoch 1 - iter 267/893 - loss 2.60742789 - time (sec): 151.05 - samples/sec: 496.82 - lr: 0.000045 - momentum: 0.000000
2023-10-10 19:43:07,007 epoch 1 - iter 356/893 - loss 2.38685711 - time (sec): 201.34 - samples/sec: 494.15 - lr: 0.000060 - momentum: 0.000000
2023-10-10 19:44:00,063 epoch 1 - iter 445/893 - loss 2.13431932 - time (sec): 254.39 - samples/sec: 494.13 - lr: 0.000075 - momentum: 0.000000
2023-10-10 19:44:51,943 epoch 1 - iter 534/893 - loss 1.92436744 - time (sec): 306.27 - samples/sec: 485.92 - lr: 0.000090 - momentum: 0.000000
2023-10-10 19:45:41,674 epoch 1 - iter 623/893 - loss 1.74254208 - time (sec): 356.00 - samples/sec: 484.27 - lr: 0.000104 - momentum: 0.000000
2023-10-10 19:46:33,920 epoch 1 - iter 712/893 - loss 1.57825868 - time (sec): 408.25 - samples/sec: 484.14 - lr: 0.000119 - momentum: 0.000000
2023-10-10 19:47:28,413 epoch 1 - iter 801/893 - loss 1.44698773 - time (sec): 462.74 - samples/sec: 482.79 - lr: 0.000134 - momentum: 0.000000
2023-10-10 19:48:22,731 epoch 1 - iter 890/893 - loss 1.34261657 - time (sec): 517.06 - samples/sec: 479.59 - lr: 0.000149 - momentum: 0.000000
2023-10-10 19:48:24,352 ----------------------------------------------------------------------------------------------------
2023-10-10 19:48:24,352 EPOCH 1 done: loss 1.3396 - lr: 0.000149
2023-10-10 19:48:45,236 DEV : loss 0.2843281924724579 - f1-score (micro avg) 0.2763
2023-10-10 19:48:45,266 saving best model
2023-10-10 19:48:46,188 ----------------------------------------------------------------------------------------------------
2023-10-10 19:49:39,164 epoch 2 - iter 89/893 - loss 0.31930581 - time (sec): 52.97 - samples/sec: 499.28 - lr: 0.000148 - momentum: 0.000000
2023-10-10 19:50:29,513 epoch 2 - iter 178/893 - loss 0.31347332 - time (sec): 103.32 - samples/sec: 488.21 - lr: 0.000147 - momentum: 0.000000
2023-10-10 19:51:19,488 epoch 2 - iter 267/893 - loss 0.29696194 - time (sec): 153.30 - samples/sec: 484.60 - lr: 0.000145 - momentum: 0.000000
2023-10-10 19:52:10,169 epoch 2 - iter 356/893 - loss 0.27774706 - time (sec): 203.98 - samples/sec: 487.01 - lr: 0.000143 - momentum: 0.000000
2023-10-10 19:53:00,310 epoch 2 - iter 445/893 - loss 0.26146640 - time (sec): 254.12 - samples/sec: 488.59 - lr: 0.000142 - momentum: 0.000000
2023-10-10 19:53:52,223 epoch 2 - iter 534/893 - loss 0.25100648 - time (sec): 306.03 - samples/sec: 484.05 - lr: 0.000140 - momentum: 0.000000
2023-10-10 19:54:44,777 epoch 2 - iter 623/893 - loss 0.23745422 - time (sec): 358.59 - samples/sec: 483.90 - lr: 0.000138 - momentum: 0.000000
2023-10-10 19:55:37,395 epoch 2 - iter 712/893 - loss 0.22599111 - time (sec): 411.20 - samples/sec: 485.53 - lr: 0.000137 - momentum: 0.000000
2023-10-10 19:56:28,840 epoch 2 - iter 801/893 - loss 0.21632815 - time (sec): 462.65 - samples/sec: 483.50 - lr: 0.000135 - momentum: 0.000000
2023-10-10 19:57:21,815 epoch 2 - iter 890/893 - loss 0.20811631 - time (sec): 515.62 - samples/sec: 481.20 - lr: 0.000133 - momentum: 0.000000
2023-10-10 19:57:23,363 ----------------------------------------------------------------------------------------------------
2023-10-10 19:57:23,364 EPOCH 2 done: loss 0.2078 - lr: 0.000133
2023-10-10 19:57:46,164 DEV : loss 0.11342751234769821 - f1-score (micro avg) 0.737
2023-10-10 19:57:46,198 saving best model
2023-10-10 19:57:56,671 ----------------------------------------------------------------------------------------------------
2023-10-10 19:58:46,865 epoch 3 - iter 89/893 - loss 0.09254059 - time (sec): 50.19 - samples/sec: 475.99 - lr: 0.000132 - momentum: 0.000000
2023-10-10 19:59:37,979 epoch 3 - iter 178/893 - loss 0.08867836 - time (sec): 101.30 - samples/sec: 487.47 - lr: 0.000130 - momentum: 0.000000
2023-10-10 20:00:28,421 epoch 3 - iter 267/893 - loss 0.08987985 - time (sec): 151.75 - samples/sec: 487.94 - lr: 0.000128 - momentum: 0.000000
2023-10-10 20:01:17,835 epoch 3 - iter 356/893 - loss 0.09300563 - time (sec): 201.16 - samples/sec: 484.15 - lr: 0.000127 - momentum: 0.000000
2023-10-10 20:02:09,174 epoch 3 - iter 445/893 - loss 0.09146828 - time (sec): 252.50 - samples/sec: 488.22 - lr: 0.000125 - momentum: 0.000000
2023-10-10 20:02:59,499 epoch 3 - iter 534/893 - loss 0.08889789 - time (sec): 302.82 - samples/sec: 489.12 - lr: 0.000123 - momentum: 0.000000
2023-10-10 20:03:50,308 epoch 3 - iter 623/893 - loss 0.08580958 - time (sec): 353.63 - samples/sec: 488.96 - lr: 0.000122 - momentum: 0.000000
2023-10-10 20:04:42,049 epoch 3 - iter 712/893 - loss 0.08460481 - time (sec): 405.37 - samples/sec: 489.39 - lr: 0.000120 - momentum: 0.000000
2023-10-10 20:05:35,281 epoch 3 - iter 801/893 - loss 0.08297470 - time (sec): 458.61 - samples/sec: 490.86 - lr: 0.000118 - momentum: 0.000000
2023-10-10 20:06:25,590 epoch 3 - iter 890/893 - loss 0.08298388 - time (sec): 508.91 - samples/sec: 487.34 - lr: 0.000117 - momentum: 0.000000
2023-10-10 20:06:27,249 ----------------------------------------------------------------------------------------------------
2023-10-10 20:06:27,249 EPOCH 3 done: loss 0.0829 - lr: 0.000117
2023-10-10 20:06:49,372 DEV : loss 0.1113942340016365 - f1-score (micro avg) 0.7559
2023-10-10 20:06:49,404 saving best model
2023-10-10 20:06:56,053 ----------------------------------------------------------------------------------------------------
2023-10-10 20:07:48,821 epoch 4 - iter 89/893 - loss 0.05345803 - time (sec): 52.76 - samples/sec: 472.25 - lr: 0.000115 - momentum: 0.000000
2023-10-10 20:08:39,233 epoch 4 - iter 178/893 - loss 0.05431726 - time (sec): 103.17 - samples/sec: 476.99 - lr: 0.000113 - momentum: 0.000000
2023-10-10 20:09:31,627 epoch 4 - iter 267/893 - loss 0.05425313 - time (sec): 155.57 - samples/sec: 474.35 - lr: 0.000112 - momentum: 0.000000
2023-10-10 20:10:23,324 epoch 4 - iter 356/893 - loss 0.05786397 - time (sec): 207.26 - samples/sec: 478.70 - lr: 0.000110 - momentum: 0.000000
2023-10-10 20:11:15,726 epoch 4 - iter 445/893 - loss 0.05597032 - time (sec): 259.67 - samples/sec: 483.48 - lr: 0.000108 - momentum: 0.000000
2023-10-10 20:12:07,752 epoch 4 - iter 534/893 - loss 0.05501856 - time (sec): 311.69 - samples/sec: 483.60 - lr: 0.000107 - momentum: 0.000000
2023-10-10 20:13:01,380 epoch 4 - iter 623/893 - loss 0.05376043 - time (sec): 365.32 - samples/sec: 484.68 - lr: 0.000105 - momentum: 0.000000
2023-10-10 20:13:51,750 epoch 4 - iter 712/893 - loss 0.05374869 - time (sec): 415.69 - samples/sec: 484.81 - lr: 0.000103 - momentum: 0.000000
2023-10-10 20:14:42,974 epoch 4 - iter 801/893 - loss 0.05390271 - time (sec): 466.91 - samples/sec: 481.79 - lr: 0.000102 - momentum: 0.000000
2023-10-10 20:15:33,480 epoch 4 - iter 890/893 - loss 0.05392475 - time (sec): 517.42 - samples/sec: 479.54 - lr: 0.000100 - momentum: 0.000000
2023-10-10 20:15:34,961 ----------------------------------------------------------------------------------------------------
2023-10-10 20:15:34,962 EPOCH 4 done: loss 0.0540 - lr: 0.000100
2023-10-10 20:15:56,301 DEV : loss 0.11903274804353714 - f1-score (micro avg) 0.7738
2023-10-10 20:15:56,332 saving best model
2023-10-10 20:16:02,540 ----------------------------------------------------------------------------------------------------
2023-10-10 20:16:54,836 epoch 5 - iter 89/893 - loss 0.03937382 - time (sec): 52.29 - samples/sec: 486.27 - lr: 0.000098 - momentum: 0.000000
2023-10-10 20:17:46,858 epoch 5 - iter 178/893 - loss 0.03701172 - time (sec): 104.31 - samples/sec: 466.44 - lr: 0.000097 - momentum: 0.000000
2023-10-10 20:18:41,817 epoch 5 - iter 267/893 - loss 0.03814808 - time (sec): 159.27 - samples/sec: 467.57 - lr: 0.000095 - momentum: 0.000000
2023-10-10 20:19:35,645 epoch 5 - iter 356/893 - loss 0.04034811 - time (sec): 213.10 - samples/sec: 472.42 - lr: 0.000093 - momentum: 0.000000
2023-10-10 20:20:27,648 epoch 5 - iter 445/893 - loss 0.04030794 - time (sec): 265.10 - samples/sec: 465.76 - lr: 0.000092 - momentum: 0.000000
2023-10-10 20:21:17,438 epoch 5 - iter 534/893 - loss 0.03985960 - time (sec): 314.89 - samples/sec: 468.02 - lr: 0.000090 - momentum: 0.000000
2023-10-10 20:22:09,623 epoch 5 - iter 623/893 - loss 0.03993508 - time (sec): 367.08 - samples/sec: 469.85 - lr: 0.000088 - momentum: 0.000000
2023-10-10 20:23:02,735 epoch 5 - iter 712/893 - loss 0.03990589 - time (sec): 420.19 - samples/sec: 471.66 - lr: 0.000087 - momentum: 0.000000
2023-10-10 20:23:56,026 epoch 5 - iter 801/893 - loss 0.03955761 - time (sec): 473.48 - samples/sec: 471.16 - lr: 0.000085 - momentum: 0.000000
2023-10-10 20:24:48,739 epoch 5 - iter 890/893 - loss 0.03910254 - time (sec): 526.20 - samples/sec: 471.38 - lr: 0.000083 - momentum: 0.000000
2023-10-10 20:24:50,326 ----------------------------------------------------------------------------------------------------
2023-10-10 20:24:50,326 EPOCH 5 done: loss 0.0391 - lr: 0.000083
2023-10-10 20:25:13,897 DEV : loss 0.12937377393245697 - f1-score (micro avg) 0.788
2023-10-10 20:25:13,940 saving best model
2023-10-10 20:25:16,976 ----------------------------------------------------------------------------------------------------
2023-10-10 20:26:09,547 epoch 6 - iter 89/893 - loss 0.02822273 - time (sec): 52.57 - samples/sec: 474.24 - lr: 0.000082 - momentum: 0.000000
2023-10-10 20:27:00,916 epoch 6 - iter 178/893 - loss 0.02803573 - time (sec): 103.93 - samples/sec: 476.62 - lr: 0.000080 - momentum: 0.000000
2023-10-10 20:27:52,841 epoch 6 - iter 267/893 - loss 0.02761811 - time (sec): 155.86 - samples/sec: 482.22 - lr: 0.000078 - momentum: 0.000000
2023-10-10 20:28:41,858 epoch 6 - iter 356/893 - loss 0.02913699 - time (sec): 204.88 - samples/sec: 484.20 - lr: 0.000077 - momentum: 0.000000
2023-10-10 20:29:32,333 epoch 6 - iter 445/893 - loss 0.02861398 - time (sec): 255.35 - samples/sec: 481.80 - lr: 0.000075 - momentum: 0.000000
2023-10-10 20:30:22,684 epoch 6 - iter 534/893 - loss 0.02936834 - time (sec): 305.70 - samples/sec: 483.71 - lr: 0.000073 - momentum: 0.000000
2023-10-10 20:31:16,351 epoch 6 - iter 623/893 - loss 0.02911242 - time (sec): 359.37 - samples/sec: 484.80 - lr: 0.000072 - momentum: 0.000000
2023-10-10 20:32:07,955 epoch 6 - iter 712/893 - loss 0.02921508 - time (sec): 410.97 - samples/sec: 484.48 - lr: 0.000070 - momentum: 0.000000
2023-10-10 20:32:59,979 epoch 6 - iter 801/893 - loss 0.02955267 - time (sec): 463.00 - samples/sec: 485.26 - lr: 0.000068 - momentum: 0.000000
2023-10-10 20:33:49,274 epoch 6 - iter 890/893 - loss 0.03022345 - time (sec): 512.29 - samples/sec: 484.15 - lr: 0.000067 - momentum: 0.000000
2023-10-10 20:33:50,856 ----------------------------------------------------------------------------------------------------
2023-10-10 20:33:50,857 EPOCH 6 done: loss 0.0301 - lr: 0.000067
2023-10-10 20:34:13,518 DEV : loss 0.16373801231384277 - f1-score (micro avg) 0.7847
2023-10-10 20:34:13,556 ----------------------------------------------------------------------------------------------------
2023-10-10 20:35:05,833 epoch 7 - iter 89/893 - loss 0.01814226 - time (sec): 52.28 - samples/sec: 485.39 - lr: 0.000065 - momentum: 0.000000
2023-10-10 20:35:55,252 epoch 7 - iter 178/893 - loss 0.02138159 - time (sec): 101.69 - samples/sec: 479.04 - lr: 0.000063 - momentum: 0.000000
2023-10-10 20:36:47,244 epoch 7 - iter 267/893 - loss 0.02102136 - time (sec): 153.69 - samples/sec: 482.73 - lr: 0.000062 - momentum: 0.000000
2023-10-10 20:37:37,631 epoch 7 - iter 356/893 - loss 0.02233733 - time (sec): 204.07 - samples/sec: 484.89 - lr: 0.000060 - momentum: 0.000000
2023-10-10 20:38:28,716 epoch 7 - iter 445/893 - loss 0.02235734 - time (sec): 255.16 - samples/sec: 481.90 - lr: 0.000058 - momentum: 0.000000
2023-10-10 20:39:19,282 epoch 7 - iter 534/893 - loss 0.02197453 - time (sec): 305.72 - samples/sec: 485.02 - lr: 0.000057 - momentum: 0.000000
2023-10-10 20:40:10,988 epoch 7 - iter 623/893 - loss 0.02262488 - time (sec): 357.43 - samples/sec: 484.42 - lr: 0.000055 - momentum: 0.000000
2023-10-10 20:41:00,028 epoch 7 - iter 712/893 - loss 0.02251825 - time (sec): 406.47 - samples/sec: 483.34 - lr: 0.000053 - momentum: 0.000000
2023-10-10 20:41:52,314 epoch 7 - iter 801/893 - loss 0.02299809 - time (sec): 458.76 - samples/sec: 485.89 - lr: 0.000052 - momentum: 0.000000
2023-10-10 20:42:43,299 epoch 7 - iter 890/893 - loss 0.02327864 - time (sec): 509.74 - samples/sec: 486.71 - lr: 0.000050 - momentum: 0.000000
2023-10-10 20:42:44,897 ----------------------------------------------------------------------------------------------------
2023-10-10 20:42:44,897 EPOCH 7 done: loss 0.0233 - lr: 0.000050
2023-10-10 20:43:07,633 DEV : loss 0.17558923363685608 - f1-score (micro avg) 0.7827
2023-10-10 20:43:07,671 ----------------------------------------------------------------------------------------------------
2023-10-10 20:44:00,852 epoch 8 - iter 89/893 - loss 0.01531205 - time (sec): 53.18 - samples/sec: 461.75 - lr: 0.000048 - momentum: 0.000000
2023-10-10 20:44:51,517 epoch 8 - iter 178/893 - loss 0.01532229 - time (sec): 103.84 - samples/sec: 467.12 - lr: 0.000047 - momentum: 0.000000
2023-10-10 20:45:42,040 epoch 8 - iter 267/893 - loss 0.01676001 - time (sec): 154.37 - samples/sec: 466.78 - lr: 0.000045 - momentum: 0.000000
2023-10-10 20:46:33,877 epoch 8 - iter 356/893 - loss 0.01635882 - time (sec): 206.20 - samples/sec: 476.10 - lr: 0.000043 - momentum: 0.000000
2023-10-10 20:47:24,646 epoch 8 - iter 445/893 - loss 0.01650485 - time (sec): 256.97 - samples/sec: 477.47 - lr: 0.000042 - momentum: 0.000000
2023-10-10 20:48:14,620 epoch 8 - iter 534/893 - loss 0.01709783 - time (sec): 306.95 - samples/sec: 475.39 - lr: 0.000040 - momentum: 0.000000
2023-10-10 20:49:06,724 epoch 8 - iter 623/893 - loss 0.01756030 - time (sec): 359.05 - samples/sec: 475.74 - lr: 0.000038 - momentum: 0.000000
2023-10-10 20:49:58,454 epoch 8 - iter 712/893 - loss 0.01732975 - time (sec): 410.78 - samples/sec: 475.08 - lr: 0.000037 - momentum: 0.000000
2023-10-10 20:50:49,934 epoch 8 - iter 801/893 - loss 0.01736550 - time (sec): 462.26 - samples/sec: 478.47 - lr: 0.000035 - momentum: 0.000000
2023-10-10 20:51:42,783 epoch 8 - iter 890/893 - loss 0.01724690 - time (sec): 515.11 - samples/sec: 480.96 - lr: 0.000033 - momentum: 0.000000
2023-10-10 20:51:44,567 ----------------------------------------------------------------------------------------------------
2023-10-10 20:51:44,568 EPOCH 8 done: loss 0.0174 - lr: 0.000033
2023-10-10 20:52:07,454 DEV : loss 0.17957079410552979 - f1-score (micro avg) 0.7849
2023-10-10 20:52:07,486 ----------------------------------------------------------------------------------------------------
2023-10-10 20:52:59,178 epoch 9 - iter 89/893 - loss 0.01800137 - time (sec): 51.69 - samples/sec: 481.68 - lr: 0.000032 - momentum: 0.000000
2023-10-10 20:53:49,523 epoch 9 - iter 178/893 - loss 0.01641771 - time (sec): 102.03 - samples/sec: 478.59 - lr: 0.000030 - momentum: 0.000000
2023-10-10 20:54:39,930 epoch 9 - iter 267/893 - loss 0.01721855 - time (sec): 152.44 - samples/sec: 490.93 - lr: 0.000028 - momentum: 0.000000
2023-10-10 20:55:28,936 epoch 9 - iter 356/893 - loss 0.01603931 - time (sec): 201.45 - samples/sec: 485.42 - lr: 0.000027 - momentum: 0.000000
2023-10-10 20:56:17,981 epoch 9 - iter 445/893 - loss 0.01512545 - time (sec): 250.49 - samples/sec: 486.34 - lr: 0.000025 - momentum: 0.000000
2023-10-10 20:57:09,343 epoch 9 - iter 534/893 - loss 0.01508396 - time (sec): 301.86 - samples/sec: 484.37 - lr: 0.000023 - momentum: 0.000000
2023-10-10 20:57:58,870 epoch 9 - iter 623/893 - loss 0.01469697 - time (sec): 351.38 - samples/sec: 484.78 - lr: 0.000022 - momentum: 0.000000
2023-10-10 20:58:50,106 epoch 9 - iter 712/893 - loss 0.01406217 - time (sec): 402.62 - samples/sec: 486.23 - lr: 0.000020 - momentum: 0.000000
2023-10-10 20:59:40,498 epoch 9 - iter 801/893 - loss 0.01413326 - time (sec): 453.01 - samples/sec: 488.22 - lr: 0.000019 - momentum: 0.000000
2023-10-10 21:00:33,325 epoch 9 - iter 890/893 - loss 0.01414608 - time (sec): 505.84 - samples/sec: 490.12 - lr: 0.000017 - momentum: 0.000000
2023-10-10 21:00:35,031 ----------------------------------------------------------------------------------------------------
2023-10-10 21:00:35,031 EPOCH 9 done: loss 0.0142 - lr: 0.000017
2023-10-10 21:00:56,874 DEV : loss 0.19276094436645508 - f1-score (micro avg) 0.778
2023-10-10 21:00:56,904 ----------------------------------------------------------------------------------------------------
2023-10-10 21:01:47,781 epoch 10 - iter 89/893 - loss 0.01006358 - time (sec): 50.88 - samples/sec: 496.06 - lr: 0.000015 - momentum: 0.000000
2023-10-10 21:02:37,723 epoch 10 - iter 178/893 - loss 0.01182496 - time (sec): 100.82 - samples/sec: 488.56 - lr: 0.000013 - momentum: 0.000000
2023-10-10 21:03:26,107 epoch 10 - iter 267/893 - loss 0.01307392 - time (sec): 149.20 - samples/sec: 484.27 - lr: 0.000012 - momentum: 0.000000
2023-10-10 21:04:17,976 epoch 10 - iter 356/893 - loss 0.01229535 - time (sec): 201.07 - samples/sec: 490.49 - lr: 0.000010 - momentum: 0.000000
2023-10-10 21:05:09,179 epoch 10 - iter 445/893 - loss 0.01180387 - time (sec): 252.27 - samples/sec: 494.77 - lr: 0.000008 - momentum: 0.000000
2023-10-10 21:05:59,983 epoch 10 - iter 534/893 - loss 0.01212851 - time (sec): 303.08 - samples/sec: 490.61 - lr: 0.000007 - momentum: 0.000000
2023-10-10 21:06:51,056 epoch 10 - iter 623/893 - loss 0.01223689 - time (sec): 354.15 - samples/sec: 494.49 - lr: 0.000005 - momentum: 0.000000
2023-10-10 21:07:41,031 epoch 10 - iter 712/893 - loss 0.01282452 - time (sec): 404.12 - samples/sec: 491.58 - lr: 0.000004 - momentum: 0.000000
2023-10-10 21:08:33,478 epoch 10 - iter 801/893 - loss 0.01253621 - time (sec): 456.57 - samples/sec: 487.48 - lr: 0.000002 - momentum: 0.000000
2023-10-10 21:09:26,121 epoch 10 - iter 890/893 - loss 0.01243382 - time (sec): 509.21 - samples/sec: 487.12 - lr: 0.000000 - momentum: 0.000000
2023-10-10 21:09:27,717 ----------------------------------------------------------------------------------------------------
2023-10-10 21:09:27,718 EPOCH 10 done: loss 0.0124 - lr: 0.000000
2023-10-10 21:09:50,838 DEV : loss 0.19964276254177094 - f1-score (micro avg) 0.7778
2023-10-10 21:09:51,780 ----------------------------------------------------------------------------------------------------
2023-10-10 21:09:51,782 Loading model from best epoch ...
2023-10-10 21:09:57,984 SequenceTagger predicts: Dictionary with 17 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-10 21:11:09,160
Results:
- F-score (micro) 0.7086
- F-score (macro) 0.6249
- Accuracy 0.5641
By class:
precision recall f1-score support
LOC 0.7183 0.7406 0.7293 1095
PER 0.7906 0.7648 0.7775 1012
ORG 0.4558 0.5630 0.5038 357
HumanProd 0.3860 0.6667 0.4889 33
micro avg 0.6938 0.7241 0.7086 2497
macro avg 0.5877 0.6838 0.6249 2497
weighted avg 0.7057 0.7241 0.7134 2497
2023-10-10 21:11:09,161 ----------------------------------------------------------------------------------------------------
|