File size: 25,490 Bytes
fbe2877 |
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 |
2023-10-11 17:36:29,073 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,075 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()
)"
2023-10-11 17:36:29,075 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,075 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
2023-10-11 17:36:29,075 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,075 Train: 7142 sentences
2023-10-11 17:36:29,075 (train_with_dev=False, train_with_test=False)
2023-10-11 17:36:29,075 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,075 Training Params:
2023-10-11 17:36:29,076 - learning_rate: "0.00015"
2023-10-11 17:36:29,076 - mini_batch_size: "4"
2023-10-11 17:36:29,076 - max_epochs: "10"
2023-10-11 17:36:29,076 - shuffle: "True"
2023-10-11 17:36:29,076 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,076 Plugins:
2023-10-11 17:36:29,076 - TensorboardLogger
2023-10-11 17:36:29,076 - LinearScheduler | warmup_fraction: '0.1'
2023-10-11 17:36:29,076 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,076 Final evaluation on model from best epoch (best-model.pt)
2023-10-11 17:36:29,076 - metric: "('micro avg', 'f1-score')"
2023-10-11 17:36:29,076 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,076 Computation:
2023-10-11 17:36:29,076 - compute on device: cuda:0
2023-10-11 17:36:29,076 - embedding storage: none
2023-10-11 17:36:29,076 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,076 Model training base path: "hmbench-newseye/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-4"
2023-10-11 17:36:29,077 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,077 ----------------------------------------------------------------------------------------------------
2023-10-11 17:36:29,077 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-11 17:37:20,456 epoch 1 - iter 178/1786 - loss 2.81924723 - time (sec): 51.38 - samples/sec: 454.30 - lr: 0.000015 - momentum: 0.000000
2023-10-11 17:38:12,331 epoch 1 - iter 356/1786 - loss 2.64980608 - time (sec): 103.25 - samples/sec: 464.05 - lr: 0.000030 - momentum: 0.000000
2023-10-11 17:39:03,793 epoch 1 - iter 534/1786 - loss 2.37963726 - time (sec): 154.71 - samples/sec: 465.85 - lr: 0.000045 - momentum: 0.000000
2023-10-11 17:39:56,229 epoch 1 - iter 712/1786 - loss 2.08450845 - time (sec): 207.15 - samples/sec: 467.58 - lr: 0.000060 - momentum: 0.000000
2023-10-11 17:40:47,993 epoch 1 - iter 890/1786 - loss 1.81534257 - time (sec): 258.91 - samples/sec: 464.69 - lr: 0.000075 - momentum: 0.000000
2023-10-11 17:41:41,378 epoch 1 - iter 1068/1786 - loss 1.59589063 - time (sec): 312.30 - samples/sec: 468.60 - lr: 0.000090 - momentum: 0.000000
2023-10-11 17:42:35,379 epoch 1 - iter 1246/1786 - loss 1.41200758 - time (sec): 366.30 - samples/sec: 472.40 - lr: 0.000105 - momentum: 0.000000
2023-10-11 17:43:27,539 epoch 1 - iter 1424/1786 - loss 1.28012419 - time (sec): 418.46 - samples/sec: 473.51 - lr: 0.000120 - momentum: 0.000000
2023-10-11 17:44:19,457 epoch 1 - iter 1602/1786 - loss 1.17694462 - time (sec): 470.38 - samples/sec: 472.74 - lr: 0.000134 - momentum: 0.000000
2023-10-11 17:45:12,644 epoch 1 - iter 1780/1786 - loss 1.08357721 - time (sec): 523.57 - samples/sec: 473.74 - lr: 0.000149 - momentum: 0.000000
2023-10-11 17:45:14,235 ----------------------------------------------------------------------------------------------------
2023-10-11 17:45:14,235 EPOCH 1 done: loss 1.0810 - lr: 0.000149
2023-10-11 17:45:34,466 DEV : loss 0.20361904799938202 - f1-score (micro avg) 0.5388
2023-10-11 17:45:34,497 saving best model
2023-10-11 17:45:35,356 ----------------------------------------------------------------------------------------------------
2023-10-11 17:46:28,481 epoch 2 - iter 178/1786 - loss 0.21290432 - time (sec): 53.12 - samples/sec: 484.33 - lr: 0.000148 - momentum: 0.000000
2023-10-11 17:47:21,521 epoch 2 - iter 356/1786 - loss 0.20649008 - time (sec): 106.16 - samples/sec: 479.48 - lr: 0.000147 - momentum: 0.000000
2023-10-11 17:48:16,086 epoch 2 - iter 534/1786 - loss 0.19057159 - time (sec): 160.73 - samples/sec: 482.34 - lr: 0.000145 - momentum: 0.000000
2023-10-11 17:49:08,697 epoch 2 - iter 712/1786 - loss 0.17993568 - time (sec): 213.34 - samples/sec: 472.54 - lr: 0.000143 - momentum: 0.000000
2023-10-11 17:50:01,167 epoch 2 - iter 890/1786 - loss 0.17058695 - time (sec): 265.81 - samples/sec: 471.80 - lr: 0.000142 - momentum: 0.000000
2023-10-11 17:50:53,743 epoch 2 - iter 1068/1786 - loss 0.16082396 - time (sec): 318.38 - samples/sec: 470.38 - lr: 0.000140 - momentum: 0.000000
2023-10-11 17:51:45,421 epoch 2 - iter 1246/1786 - loss 0.15569068 - time (sec): 370.06 - samples/sec: 469.79 - lr: 0.000138 - momentum: 0.000000
2023-10-11 17:52:38,828 epoch 2 - iter 1424/1786 - loss 0.14996741 - time (sec): 423.47 - samples/sec: 468.27 - lr: 0.000137 - momentum: 0.000000
2023-10-11 17:53:34,073 epoch 2 - iter 1602/1786 - loss 0.14514985 - time (sec): 478.71 - samples/sec: 463.26 - lr: 0.000135 - momentum: 0.000000
2023-10-11 17:54:30,018 epoch 2 - iter 1780/1786 - loss 0.14157379 - time (sec): 534.66 - samples/sec: 463.16 - lr: 0.000133 - momentum: 0.000000
2023-10-11 17:54:32,031 ----------------------------------------------------------------------------------------------------
2023-10-11 17:54:32,032 EPOCH 2 done: loss 0.1412 - lr: 0.000133
2023-10-11 17:54:54,099 DEV : loss 0.10268282890319824 - f1-score (micro avg) 0.7698
2023-10-11 17:54:54,135 saving best model
2023-10-11 17:54:56,735 ----------------------------------------------------------------------------------------------------
2023-10-11 17:55:52,053 epoch 3 - iter 178/1786 - loss 0.07581422 - time (sec): 55.31 - samples/sec: 448.39 - lr: 0.000132 - momentum: 0.000000
2023-10-11 17:56:46,482 epoch 3 - iter 356/1786 - loss 0.08043397 - time (sec): 109.74 - samples/sec: 455.00 - lr: 0.000130 - momentum: 0.000000
2023-10-11 17:57:40,496 epoch 3 - iter 534/1786 - loss 0.07520227 - time (sec): 163.76 - samples/sec: 452.76 - lr: 0.000128 - momentum: 0.000000
2023-10-11 17:58:34,143 epoch 3 - iter 712/1786 - loss 0.07247742 - time (sec): 217.40 - samples/sec: 451.46 - lr: 0.000127 - momentum: 0.000000
2023-10-11 17:59:27,936 epoch 3 - iter 890/1786 - loss 0.07466025 - time (sec): 271.20 - samples/sec: 452.66 - lr: 0.000125 - momentum: 0.000000
2023-10-11 18:00:21,462 epoch 3 - iter 1068/1786 - loss 0.07516997 - time (sec): 324.72 - samples/sec: 454.15 - lr: 0.000123 - momentum: 0.000000
2023-10-11 18:01:15,676 epoch 3 - iter 1246/1786 - loss 0.07379608 - time (sec): 378.94 - samples/sec: 453.76 - lr: 0.000122 - momentum: 0.000000
2023-10-11 18:02:09,837 epoch 3 - iter 1424/1786 - loss 0.07445074 - time (sec): 433.10 - samples/sec: 455.26 - lr: 0.000120 - momentum: 0.000000
2023-10-11 18:03:03,839 epoch 3 - iter 1602/1786 - loss 0.07358598 - time (sec): 487.10 - samples/sec: 456.93 - lr: 0.000118 - momentum: 0.000000
2023-10-11 18:03:58,389 epoch 3 - iter 1780/1786 - loss 0.07423850 - time (sec): 541.65 - samples/sec: 457.42 - lr: 0.000117 - momentum: 0.000000
2023-10-11 18:04:00,209 ----------------------------------------------------------------------------------------------------
2023-10-11 18:04:00,209 EPOCH 3 done: loss 0.0743 - lr: 0.000117
2023-10-11 18:04:21,642 DEV : loss 0.13196961581707 - f1-score (micro avg) 0.7713
2023-10-11 18:04:21,673 saving best model
2023-10-11 18:04:24,252 ----------------------------------------------------------------------------------------------------
2023-10-11 18:05:16,538 epoch 4 - iter 178/1786 - loss 0.05605407 - time (sec): 52.28 - samples/sec: 471.55 - lr: 0.000115 - momentum: 0.000000
2023-10-11 18:06:08,745 epoch 4 - iter 356/1786 - loss 0.05000837 - time (sec): 104.49 - samples/sec: 475.50 - lr: 0.000113 - momentum: 0.000000
2023-10-11 18:07:01,886 epoch 4 - iter 534/1786 - loss 0.05222854 - time (sec): 157.63 - samples/sec: 481.86 - lr: 0.000112 - momentum: 0.000000
2023-10-11 18:07:54,105 epoch 4 - iter 712/1786 - loss 0.05172677 - time (sec): 209.85 - samples/sec: 479.63 - lr: 0.000110 - momentum: 0.000000
2023-10-11 18:08:45,957 epoch 4 - iter 890/1786 - loss 0.05101322 - time (sec): 261.70 - samples/sec: 477.03 - lr: 0.000108 - momentum: 0.000000
2023-10-11 18:09:40,097 epoch 4 - iter 1068/1786 - loss 0.05070108 - time (sec): 315.84 - samples/sec: 474.53 - lr: 0.000107 - momentum: 0.000000
2023-10-11 18:10:36,928 epoch 4 - iter 1246/1786 - loss 0.05208864 - time (sec): 372.67 - samples/sec: 473.16 - lr: 0.000105 - momentum: 0.000000
2023-10-11 18:11:31,138 epoch 4 - iter 1424/1786 - loss 0.05315052 - time (sec): 426.88 - samples/sec: 467.69 - lr: 0.000103 - momentum: 0.000000
2023-10-11 18:12:24,160 epoch 4 - iter 1602/1786 - loss 0.05314471 - time (sec): 479.91 - samples/sec: 466.16 - lr: 0.000102 - momentum: 0.000000
2023-10-11 18:13:18,131 epoch 4 - iter 1780/1786 - loss 0.05220709 - time (sec): 533.88 - samples/sec: 464.57 - lr: 0.000100 - momentum: 0.000000
2023-10-11 18:13:19,774 ----------------------------------------------------------------------------------------------------
2023-10-11 18:13:19,775 EPOCH 4 done: loss 0.0522 - lr: 0.000100
2023-10-11 18:13:42,131 DEV : loss 0.15180714428424835 - f1-score (micro avg) 0.7815
2023-10-11 18:13:42,166 saving best model
2023-10-11 18:13:46,961 ----------------------------------------------------------------------------------------------------
2023-10-11 18:14:40,307 epoch 5 - iter 178/1786 - loss 0.04155092 - time (sec): 53.34 - samples/sec: 445.72 - lr: 0.000098 - momentum: 0.000000
2023-10-11 18:15:33,137 epoch 5 - iter 356/1786 - loss 0.03973759 - time (sec): 106.17 - samples/sec: 441.09 - lr: 0.000097 - momentum: 0.000000
2023-10-11 18:16:28,231 epoch 5 - iter 534/1786 - loss 0.03869343 - time (sec): 161.27 - samples/sec: 455.20 - lr: 0.000095 - momentum: 0.000000
2023-10-11 18:17:22,718 epoch 5 - iter 712/1786 - loss 0.03913412 - time (sec): 215.75 - samples/sec: 457.97 - lr: 0.000093 - momentum: 0.000000
2023-10-11 18:18:15,097 epoch 5 - iter 890/1786 - loss 0.03875654 - time (sec): 268.13 - samples/sec: 458.77 - lr: 0.000092 - momentum: 0.000000
2023-10-11 18:19:07,851 epoch 5 - iter 1068/1786 - loss 0.03759774 - time (sec): 320.89 - samples/sec: 456.30 - lr: 0.000090 - momentum: 0.000000
2023-10-11 18:20:03,982 epoch 5 - iter 1246/1786 - loss 0.03756197 - time (sec): 377.02 - samples/sec: 458.06 - lr: 0.000088 - momentum: 0.000000
2023-10-11 18:20:58,516 epoch 5 - iter 1424/1786 - loss 0.03737673 - time (sec): 431.55 - samples/sec: 455.81 - lr: 0.000087 - momentum: 0.000000
2023-10-11 18:21:52,617 epoch 5 - iter 1602/1786 - loss 0.03776158 - time (sec): 485.65 - samples/sec: 457.43 - lr: 0.000085 - momentum: 0.000000
2023-10-11 18:22:49,426 epoch 5 - iter 1780/1786 - loss 0.03807107 - time (sec): 542.46 - samples/sec: 456.73 - lr: 0.000083 - momentum: 0.000000
2023-10-11 18:22:51,427 ----------------------------------------------------------------------------------------------------
2023-10-11 18:22:51,427 EPOCH 5 done: loss 0.0380 - lr: 0.000083
2023-10-11 18:23:14,760 DEV : loss 0.16010259091854095 - f1-score (micro avg) 0.8083
2023-10-11 18:23:14,792 saving best model
2023-10-11 18:23:17,472 ----------------------------------------------------------------------------------------------------
2023-10-11 18:24:12,145 epoch 6 - iter 178/1786 - loss 0.03238792 - time (sec): 54.67 - samples/sec: 459.55 - lr: 0.000082 - momentum: 0.000000
2023-10-11 18:25:05,682 epoch 6 - iter 356/1786 - loss 0.02941863 - time (sec): 108.20 - samples/sec: 458.47 - lr: 0.000080 - momentum: 0.000000
2023-10-11 18:25:58,332 epoch 6 - iter 534/1786 - loss 0.02729745 - time (sec): 160.85 - samples/sec: 457.92 - lr: 0.000078 - momentum: 0.000000
2023-10-11 18:26:52,267 epoch 6 - iter 712/1786 - loss 0.02776539 - time (sec): 214.79 - samples/sec: 459.33 - lr: 0.000077 - momentum: 0.000000
2023-10-11 18:27:47,349 epoch 6 - iter 890/1786 - loss 0.02725920 - time (sec): 269.87 - samples/sec: 455.75 - lr: 0.000075 - momentum: 0.000000
2023-10-11 18:28:44,263 epoch 6 - iter 1068/1786 - loss 0.02841412 - time (sec): 326.79 - samples/sec: 454.06 - lr: 0.000073 - momentum: 0.000000
2023-10-11 18:29:41,292 epoch 6 - iter 1246/1786 - loss 0.02811007 - time (sec): 383.82 - samples/sec: 451.55 - lr: 0.000072 - momentum: 0.000000
2023-10-11 18:30:36,642 epoch 6 - iter 1424/1786 - loss 0.02811630 - time (sec): 439.16 - samples/sec: 451.16 - lr: 0.000070 - momentum: 0.000000
2023-10-11 18:31:31,328 epoch 6 - iter 1602/1786 - loss 0.02770252 - time (sec): 493.85 - samples/sec: 451.21 - lr: 0.000068 - momentum: 0.000000
2023-10-11 18:32:25,480 epoch 6 - iter 1780/1786 - loss 0.02744603 - time (sec): 548.00 - samples/sec: 452.48 - lr: 0.000067 - momentum: 0.000000
2023-10-11 18:32:27,134 ----------------------------------------------------------------------------------------------------
2023-10-11 18:32:27,135 EPOCH 6 done: loss 0.0275 - lr: 0.000067
2023-10-11 18:32:49,419 DEV : loss 0.1975564956665039 - f1-score (micro avg) 0.7922
2023-10-11 18:32:49,452 ----------------------------------------------------------------------------------------------------
2023-10-11 18:33:42,384 epoch 7 - iter 178/1786 - loss 0.02364687 - time (sec): 52.93 - samples/sec: 456.89 - lr: 0.000065 - momentum: 0.000000
2023-10-11 18:34:36,361 epoch 7 - iter 356/1786 - loss 0.02566367 - time (sec): 106.91 - samples/sec: 456.18 - lr: 0.000063 - momentum: 0.000000
2023-10-11 18:35:30,916 epoch 7 - iter 534/1786 - loss 0.02607659 - time (sec): 161.46 - samples/sec: 451.34 - lr: 0.000062 - momentum: 0.000000
2023-10-11 18:36:25,221 epoch 7 - iter 712/1786 - loss 0.02345986 - time (sec): 215.77 - samples/sec: 456.83 - lr: 0.000060 - momentum: 0.000000
2023-10-11 18:37:20,581 epoch 7 - iter 890/1786 - loss 0.02360187 - time (sec): 271.13 - samples/sec: 456.96 - lr: 0.000058 - momentum: 0.000000
2023-10-11 18:38:15,580 epoch 7 - iter 1068/1786 - loss 0.02238154 - time (sec): 326.13 - samples/sec: 455.70 - lr: 0.000057 - momentum: 0.000000
2023-10-11 18:39:10,005 epoch 7 - iter 1246/1786 - loss 0.02133457 - time (sec): 380.55 - samples/sec: 454.89 - lr: 0.000055 - momentum: 0.000000
2023-10-11 18:40:05,662 epoch 7 - iter 1424/1786 - loss 0.02172766 - time (sec): 436.21 - samples/sec: 455.44 - lr: 0.000053 - momentum: 0.000000
2023-10-11 18:41:01,549 epoch 7 - iter 1602/1786 - loss 0.02138396 - time (sec): 492.10 - samples/sec: 454.50 - lr: 0.000052 - momentum: 0.000000
2023-10-11 18:41:54,273 epoch 7 - iter 1780/1786 - loss 0.02081679 - time (sec): 544.82 - samples/sec: 454.51 - lr: 0.000050 - momentum: 0.000000
2023-10-11 18:41:56,288 ----------------------------------------------------------------------------------------------------
2023-10-11 18:41:56,288 EPOCH 7 done: loss 0.0207 - lr: 0.000050
2023-10-11 18:42:18,651 DEV : loss 0.2033122181892395 - f1-score (micro avg) 0.7906
2023-10-11 18:42:18,684 ----------------------------------------------------------------------------------------------------
2023-10-11 18:43:14,073 epoch 8 - iter 178/1786 - loss 0.01534023 - time (sec): 55.39 - samples/sec: 446.17 - lr: 0.000048 - momentum: 0.000000
2023-10-11 18:44:09,481 epoch 8 - iter 356/1786 - loss 0.01688005 - time (sec): 110.79 - samples/sec: 455.36 - lr: 0.000047 - momentum: 0.000000
2023-10-11 18:45:04,364 epoch 8 - iter 534/1786 - loss 0.01455467 - time (sec): 165.68 - samples/sec: 454.79 - lr: 0.000045 - momentum: 0.000000
2023-10-11 18:46:00,082 epoch 8 - iter 712/1786 - loss 0.01641305 - time (sec): 221.40 - samples/sec: 453.36 - lr: 0.000043 - momentum: 0.000000
2023-10-11 18:46:56,812 epoch 8 - iter 890/1786 - loss 0.01612635 - time (sec): 278.13 - samples/sec: 446.87 - lr: 0.000042 - momentum: 0.000000
2023-10-11 18:47:53,561 epoch 8 - iter 1068/1786 - loss 0.01499776 - time (sec): 334.87 - samples/sec: 443.86 - lr: 0.000040 - momentum: 0.000000
2023-10-11 18:48:50,595 epoch 8 - iter 1246/1786 - loss 0.01544156 - time (sec): 391.91 - samples/sec: 444.25 - lr: 0.000038 - momentum: 0.000000
2023-10-11 18:49:45,028 epoch 8 - iter 1424/1786 - loss 0.01554454 - time (sec): 446.34 - samples/sec: 440.67 - lr: 0.000037 - momentum: 0.000000
2023-10-11 18:50:39,377 epoch 8 - iter 1602/1786 - loss 0.01519250 - time (sec): 500.69 - samples/sec: 444.59 - lr: 0.000035 - momentum: 0.000000
2023-10-11 18:51:34,428 epoch 8 - iter 1780/1786 - loss 0.01551751 - time (sec): 555.74 - samples/sec: 446.37 - lr: 0.000033 - momentum: 0.000000
2023-10-11 18:51:36,048 ----------------------------------------------------------------------------------------------------
2023-10-11 18:51:36,048 EPOCH 8 done: loss 0.0155 - lr: 0.000033
2023-10-11 18:51:59,372 DEV : loss 0.2236306071281433 - f1-score (micro avg) 0.7925
2023-10-11 18:51:59,405 ----------------------------------------------------------------------------------------------------
2023-10-11 18:52:53,303 epoch 9 - iter 178/1786 - loss 0.00534514 - time (sec): 53.90 - samples/sec: 438.90 - lr: 0.000032 - momentum: 0.000000
2023-10-11 18:53:48,375 epoch 9 - iter 356/1786 - loss 0.01015032 - time (sec): 108.97 - samples/sec: 453.68 - lr: 0.000030 - momentum: 0.000000
2023-10-11 18:54:44,141 epoch 9 - iter 534/1786 - loss 0.01306851 - time (sec): 164.73 - samples/sec: 459.92 - lr: 0.000028 - momentum: 0.000000
2023-10-11 18:55:39,984 epoch 9 - iter 712/1786 - loss 0.01202527 - time (sec): 220.58 - samples/sec: 456.11 - lr: 0.000027 - momentum: 0.000000
2023-10-11 18:56:34,979 epoch 9 - iter 890/1786 - loss 0.01142279 - time (sec): 275.57 - samples/sec: 452.74 - lr: 0.000025 - momentum: 0.000000
2023-10-11 18:57:31,235 epoch 9 - iter 1068/1786 - loss 0.01115001 - time (sec): 331.83 - samples/sec: 451.22 - lr: 0.000023 - momentum: 0.000000
2023-10-11 18:58:26,472 epoch 9 - iter 1246/1786 - loss 0.01140760 - time (sec): 387.06 - samples/sec: 449.42 - lr: 0.000022 - momentum: 0.000000
2023-10-11 18:59:22,703 epoch 9 - iter 1424/1786 - loss 0.01133329 - time (sec): 443.30 - samples/sec: 447.16 - lr: 0.000020 - momentum: 0.000000
2023-10-11 19:00:18,219 epoch 9 - iter 1602/1786 - loss 0.01110225 - time (sec): 498.81 - samples/sec: 446.00 - lr: 0.000018 - momentum: 0.000000
2023-10-11 19:01:15,055 epoch 9 - iter 1780/1786 - loss 0.01103208 - time (sec): 555.65 - samples/sec: 446.12 - lr: 0.000017 - momentum: 0.000000
2023-10-11 19:01:16,897 ----------------------------------------------------------------------------------------------------
2023-10-11 19:01:16,897 EPOCH 9 done: loss 0.0110 - lr: 0.000017
2023-10-11 19:01:39,717 DEV : loss 0.23193296790122986 - f1-score (micro avg) 0.7974
2023-10-11 19:01:39,750 ----------------------------------------------------------------------------------------------------
2023-10-11 19:02:37,006 epoch 10 - iter 178/1786 - loss 0.00698458 - time (sec): 57.25 - samples/sec: 439.62 - lr: 0.000015 - momentum: 0.000000
2023-10-11 19:03:34,910 epoch 10 - iter 356/1786 - loss 0.00817366 - time (sec): 115.16 - samples/sec: 445.19 - lr: 0.000013 - momentum: 0.000000
2023-10-11 19:04:32,347 epoch 10 - iter 534/1786 - loss 0.00869250 - time (sec): 172.59 - samples/sec: 432.02 - lr: 0.000012 - momentum: 0.000000
2023-10-11 19:05:30,527 epoch 10 - iter 712/1786 - loss 0.00881744 - time (sec): 230.77 - samples/sec: 430.16 - lr: 0.000010 - momentum: 0.000000
2023-10-11 19:06:24,815 epoch 10 - iter 890/1786 - loss 0.00835789 - time (sec): 285.06 - samples/sec: 429.36 - lr: 0.000008 - momentum: 0.000000
2023-10-11 19:07:21,724 epoch 10 - iter 1068/1786 - loss 0.00867803 - time (sec): 341.97 - samples/sec: 436.06 - lr: 0.000007 - momentum: 0.000000
2023-10-11 19:08:17,370 epoch 10 - iter 1246/1786 - loss 0.00883856 - time (sec): 397.62 - samples/sec: 434.74 - lr: 0.000005 - momentum: 0.000000
2023-10-11 19:09:14,246 epoch 10 - iter 1424/1786 - loss 0.00953875 - time (sec): 454.49 - samples/sec: 435.60 - lr: 0.000003 - momentum: 0.000000
2023-10-11 19:10:12,234 epoch 10 - iter 1602/1786 - loss 0.00908507 - time (sec): 512.48 - samples/sec: 435.95 - lr: 0.000002 - momentum: 0.000000
2023-10-11 19:11:08,898 epoch 10 - iter 1780/1786 - loss 0.00899664 - time (sec): 569.15 - samples/sec: 436.09 - lr: 0.000000 - momentum: 0.000000
2023-10-11 19:11:10,476 ----------------------------------------------------------------------------------------------------
2023-10-11 19:11:10,477 EPOCH 10 done: loss 0.0090 - lr: 0.000000
2023-10-11 19:11:34,051 DEV : loss 0.2363290935754776 - f1-score (micro avg) 0.7909
2023-10-11 19:11:35,015 ----------------------------------------------------------------------------------------------------
2023-10-11 19:11:35,017 Loading model from best epoch ...
2023-10-11 19:11:41,051 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
2023-10-11 19:12:53,145
Results:
- F-score (micro) 0.6861
- F-score (macro) 0.6058
- Accuracy 0.5407
By class:
precision recall f1-score support
LOC 0.7063 0.7050 0.7057 1095
PER 0.7809 0.7500 0.7651 1012
ORG 0.4030 0.5994 0.4820 357
HumanProd 0.3846 0.6061 0.4706 33
micro avg 0.6665 0.7068 0.6861 2497
macro avg 0.5687 0.6651 0.6058 2497
weighted avg 0.6889 0.7068 0.6947 2497
2023-10-11 19:12:53,145 ----------------------------------------------------------------------------------------------------
|