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 ----------------------------------------------------------------------------------------------------
|