Hugging Face's logo

stefan-it
/
hmbench-hipe2020-de-hmbert_64k-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2

Token Classification
Flair
PyTorch
TensorBoard
German
sequence-tagger-model
Model card Files Metrics Community
1
hmbench-hipe2020-de-hmbert_64k-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2 / training.log
stefan-it's picture
stefan-it
Upload ./training.log with huggingface_hub
c937ee9
raw history blame contribute delete
No virus
37 kB
2023-10-23 20:42:58,126 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,127 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(64001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(1): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(2): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(3): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(4): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(5): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(6): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(7): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(8): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(9): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(10): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(11): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=21, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-23 20:42:58,127 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 MultiCorpus: 3575 train + 1235 dev + 1266 test sentences
- NER_HIPE_2022 Corpus: 3575 train + 1235 dev + 1266 test sentences - /home/ubuntu/.flair/datasets/ner_hipe_2022/v2.1/hipe2020/de/with_doc_seperator
2023-10-23 20:42:58,128 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 Train: 3575 sentences
2023-10-23 20:42:58,128 (train_with_dev=False, train_with_test=False)
2023-10-23 20:42:58,128 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 Training Params:
2023-10-23 20:42:58,128 - learning_rate: "3e-05"
2023-10-23 20:42:58,128 - mini_batch_size: "4"
2023-10-23 20:42:58,128 - max_epochs: "10"
2023-10-23 20:42:58,128 - shuffle: "True"
2023-10-23 20:42:58,128 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 Plugins:
2023-10-23 20:42:58,128 - TensorboardLogger
2023-10-23 20:42:58,128 - LinearScheduler | warmup_fraction: '0.1'
2023-10-23 20:42:58,128 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 Final evaluation on model from best epoch (best-model.pt)
2023-10-23 20:42:58,128 - metric: "('micro avg', 'f1-score')"
2023-10-23 20:42:58,128 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,128 Computation:
2023-10-23 20:42:58,129 - compute on device: cuda:0
2023-10-23 20:42:58,129 - embedding storage: none
2023-10-23 20:42:58,129 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,129 Model training base path: "hmbench-hipe2020/de-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-23 20:42:58,129 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,129 ----------------------------------------------------------------------------------------------------
2023-10-23 20:42:58,129 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-23 20:43:03,640 epoch 1 - iter 89/894 - loss 2.48080723 - time (sec): 5.51 - samples/sec: 1445.17 - lr: 0.000003 - momentum: 0.000000
2023-10-23 20:43:09,345 epoch 1 - iter 178/894 - loss 1.45885269 - time (sec): 11.22 - samples/sec: 1485.05 - lr: 0.000006 - momentum: 0.000000
2023-10-23 20:43:15,029 epoch 1 - iter 267/894 - loss 1.08472181 - time (sec): 16.90 - samples/sec: 1488.89 - lr: 0.000009 - momentum: 0.000000
2023-10-23 20:43:20,705 epoch 1 - iter 356/894 - loss 0.90566478 - time (sec): 22.58 - samples/sec: 1493.43 - lr: 0.000012 - momentum: 0.000000
2023-10-23 20:43:26,321 epoch 1 - iter 445/894 - loss 0.78795357 - time (sec): 28.19 - samples/sec: 1501.14 - lr: 0.000015 - momentum: 0.000000
2023-10-23 20:43:31,849 epoch 1 - iter 534/894 - loss 0.70900774 - time (sec): 33.72 - samples/sec: 1494.09 - lr: 0.000018 - momentum: 0.000000
2023-10-23 20:43:37,466 epoch 1 - iter 623/894 - loss 0.63979138 - time (sec): 39.34 - samples/sec: 1498.39 - lr: 0.000021 - momentum: 0.000000
2023-10-23 20:43:43,114 epoch 1 - iter 712/894 - loss 0.58612567 - time (sec): 44.98 - samples/sec: 1502.92 - lr: 0.000024 - momentum: 0.000000
2023-10-23 20:43:49,081 epoch 1 - iter 801/894 - loss 0.54155740 - time (sec): 50.95 - samples/sec: 1515.59 - lr: 0.000027 - momentum: 0.000000
2023-10-23 20:43:54,698 epoch 1 - iter 890/894 - loss 0.50622355 - time (sec): 56.57 - samples/sec: 1524.76 - lr: 0.000030 - momentum: 0.000000
2023-10-23 20:43:54,933 ----------------------------------------------------------------------------------------------------
2023-10-23 20:43:54,933 EPOCH 1 done: loss 0.5053 - lr: 0.000030
2023-10-23 20:43:59,782 DEV : loss 0.1566276252269745 - f1-score (micro avg) 0.6176
2023-10-23 20:43:59,803 saving best model
2023-10-23 20:44:00,374 ----------------------------------------------------------------------------------------------------
2023-10-23 20:44:05,850 epoch 2 - iter 89/894 - loss 0.16450987 - time (sec): 5.47 - samples/sec: 1511.62 - lr: 0.000030 - momentum: 0.000000
2023-10-23 20:44:11,563 epoch 2 - iter 178/894 - loss 0.15748841 - time (sec): 11.19 - samples/sec: 1519.78 - lr: 0.000029 - momentum: 0.000000
2023-10-23 20:44:17,274 epoch 2 - iter 267/894 - loss 0.15283243 - time (sec): 16.90 - samples/sec: 1533.58 - lr: 0.000029 - momentum: 0.000000
2023-10-23 20:44:23,049 epoch 2 - iter 356/894 - loss 0.15237744 - time (sec): 22.67 - samples/sec: 1528.63 - lr: 0.000029 - momentum: 0.000000
2023-10-23 20:44:28,594 epoch 2 - iter 445/894 - loss 0.14585054 - time (sec): 28.22 - samples/sec: 1505.44 - lr: 0.000028 - momentum: 0.000000
2023-10-23 20:44:34,379 epoch 2 - iter 534/894 - loss 0.15087654 - time (sec): 34.00 - samples/sec: 1513.75 - lr: 0.000028 - momentum: 0.000000
2023-10-23 20:44:40,096 epoch 2 - iter 623/894 - loss 0.14971394 - time (sec): 39.72 - samples/sec: 1522.56 - lr: 0.000028 - momentum: 0.000000
2023-10-23 20:44:45,566 epoch 2 - iter 712/894 - loss 0.15039161 - time (sec): 45.19 - samples/sec: 1511.99 - lr: 0.000027 - momentum: 0.000000
2023-10-23 20:44:51,516 epoch 2 - iter 801/894 - loss 0.14951375 - time (sec): 51.14 - samples/sec: 1523.17 - lr: 0.000027 - momentum: 0.000000
2023-10-23 20:44:57,059 epoch 2 - iter 890/894 - loss 0.14581873 - time (sec): 56.68 - samples/sec: 1518.95 - lr: 0.000027 - momentum: 0.000000
2023-10-23 20:44:57,322 ----------------------------------------------------------------------------------------------------
2023-10-23 20:44:57,322 EPOCH 2 done: loss 0.1455 - lr: 0.000027
2023-10-23 20:45:03,843 DEV : loss 0.16595827043056488 - f1-score (micro avg) 0.717
2023-10-23 20:45:03,863 saving best model
2023-10-23 20:45:04,578 ----------------------------------------------------------------------------------------------------
2023-10-23 20:45:10,343 epoch 3 - iter 89/894 - loss 0.09957272 - time (sec): 5.76 - samples/sec: 1554.16 - lr: 0.000026 - momentum: 0.000000
2023-10-23 20:45:16,075 epoch 3 - iter 178/894 - loss 0.09503941 - time (sec): 11.50 - samples/sec: 1536.73 - lr: 0.000026 - momentum: 0.000000
2023-10-23 20:45:21,827 epoch 3 - iter 267/894 - loss 0.08946847 - time (sec): 17.25 - samples/sec: 1556.44 - lr: 0.000026 - momentum: 0.000000
2023-10-23 20:45:27,441 epoch 3 - iter 356/894 - loss 0.08424858 - time (sec): 22.86 - samples/sec: 1525.11 - lr: 0.000025 - momentum: 0.000000
2023-10-23 20:45:33,050 epoch 3 - iter 445/894 - loss 0.08506845 - time (sec): 28.47 - samples/sec: 1528.17 - lr: 0.000025 - momentum: 0.000000
2023-10-23 20:45:38,640 epoch 3 - iter 534/894 - loss 0.08475705 - time (sec): 34.06 - samples/sec: 1522.51 - lr: 0.000025 - momentum: 0.000000
2023-10-23 20:45:44,182 epoch 3 - iter 623/894 - loss 0.08369846 - time (sec): 39.60 - samples/sec: 1513.84 - lr: 0.000024 - momentum: 0.000000
2023-10-23 20:45:50,076 epoch 3 - iter 712/894 - loss 0.08163382 - time (sec): 45.50 - samples/sec: 1521.35 - lr: 0.000024 - momentum: 0.000000
2023-10-23 20:45:55,687 epoch 3 - iter 801/894 - loss 0.08201642 - time (sec): 51.11 - samples/sec: 1515.67 - lr: 0.000024 - momentum: 0.000000
2023-10-23 20:46:01,448 epoch 3 - iter 890/894 - loss 0.08247904 - time (sec): 56.87 - samples/sec: 1516.36 - lr: 0.000023 - momentum: 0.000000
2023-10-23 20:46:01,681 ----------------------------------------------------------------------------------------------------
2023-10-23 20:46:01,681 EPOCH 3 done: loss 0.0827 - lr: 0.000023
2023-10-23 20:46:08,204 DEV : loss 0.17428651452064514 - f1-score (micro avg) 0.7264
2023-10-23 20:46:08,224 saving best model
2023-10-23 20:46:08,976 ----------------------------------------------------------------------------------------------------
2023-10-23 20:46:14,510 epoch 4 - iter 89/894 - loss 0.05262825 - time (sec): 5.53 - samples/sec: 1466.68 - lr: 0.000023 - momentum: 0.000000
2023-10-23 20:46:20,231 epoch 4 - iter 178/894 - loss 0.04468521 - time (sec): 11.25 - samples/sec: 1489.29 - lr: 0.000023 - momentum: 0.000000
2023-10-23 20:46:25,885 epoch 4 - iter 267/894 - loss 0.04312714 - time (sec): 16.91 - samples/sec: 1506.15 - lr: 0.000022 - momentum: 0.000000
2023-10-23 20:46:31,747 epoch 4 - iter 356/894 - loss 0.04469493 - time (sec): 22.77 - samples/sec: 1515.98 - lr: 0.000022 - momentum: 0.000000
2023-10-23 20:46:37,451 epoch 4 - iter 445/894 - loss 0.04758485 - time (sec): 28.47 - samples/sec: 1510.93 - lr: 0.000022 - momentum: 0.000000
2023-10-23 20:46:43,175 epoch 4 - iter 534/894 - loss 0.04927085 - time (sec): 34.20 - samples/sec: 1513.77 - lr: 0.000021 - momentum: 0.000000
2023-10-23 20:46:48,977 epoch 4 - iter 623/894 - loss 0.05000260 - time (sec): 40.00 - samples/sec: 1524.08 - lr: 0.000021 - momentum: 0.000000
2023-10-23 20:46:54,699 epoch 4 - iter 712/894 - loss 0.05005085 - time (sec): 45.72 - samples/sec: 1525.25 - lr: 0.000021 - momentum: 0.000000
2023-10-23 20:47:00,250 epoch 4 - iter 801/894 - loss 0.04974125 - time (sec): 51.27 - samples/sec: 1517.36 - lr: 0.000020 - momentum: 0.000000
2023-10-23 20:47:05,779 epoch 4 - iter 890/894 - loss 0.05150310 - time (sec): 56.80 - samples/sec: 1517.54 - lr: 0.000020 - momentum: 0.000000
2023-10-23 20:47:06,031 ----------------------------------------------------------------------------------------------------
2023-10-23 20:47:06,032 EPOCH 4 done: loss 0.0525 - lr: 0.000020
2023-10-23 20:47:12,556 DEV : loss 0.20298461616039276 - f1-score (micro avg) 0.7392
2023-10-23 20:47:12,576 saving best model
2023-10-23 20:47:13,313 ----------------------------------------------------------------------------------------------------
2023-10-23 20:47:18,999 epoch 5 - iter 89/894 - loss 0.03232253 - time (sec): 5.68 - samples/sec: 1546.64 - lr: 0.000020 - momentum: 0.000000
2023-10-23 20:47:24,694 epoch 5 - iter 178/894 - loss 0.03349854 - time (sec): 11.38 - samples/sec: 1498.69 - lr: 0.000019 - momentum: 0.000000
2023-10-23 20:47:30,205 epoch 5 - iter 267/894 - loss 0.03340786 - time (sec): 16.89 - samples/sec: 1484.26 - lr: 0.000019 - momentum: 0.000000
2023-10-23 20:47:36,073 epoch 5 - iter 356/894 - loss 0.03460467 - time (sec): 22.76 - samples/sec: 1518.03 - lr: 0.000019 - momentum: 0.000000
2023-10-23 20:47:41,657 epoch 5 - iter 445/894 - loss 0.03216847 - time (sec): 28.34 - samples/sec: 1504.36 - lr: 0.000018 - momentum: 0.000000
2023-10-23 20:47:47,217 epoch 5 - iter 534/894 - loss 0.03332744 - time (sec): 33.90 - samples/sec: 1500.36 - lr: 0.000018 - momentum: 0.000000
2023-10-23 20:47:53,166 epoch 5 - iter 623/894 - loss 0.03407386 - time (sec): 39.85 - samples/sec: 1514.43 - lr: 0.000018 - momentum: 0.000000
2023-10-23 20:47:58,849 epoch 5 - iter 712/894 - loss 0.03376706 - time (sec): 45.53 - samples/sec: 1516.40 - lr: 0.000017 - momentum: 0.000000
2023-10-23 20:48:04,460 epoch 5 - iter 801/894 - loss 0.03450533 - time (sec): 51.15 - samples/sec: 1523.39 - lr: 0.000017 - momentum: 0.000000
2023-10-23 20:48:10,007 epoch 5 - iter 890/894 - loss 0.03436077 - time (sec): 56.69 - samples/sec: 1519.59 - lr: 0.000017 - momentum: 0.000000
2023-10-23 20:48:10,262 ----------------------------------------------------------------------------------------------------
2023-10-23 20:48:10,262 EPOCH 5 done: loss 0.0344 - lr: 0.000017
2023-10-23 20:48:16,766 DEV : loss 0.24649250507354736 - f1-score (micro avg) 0.7686
2023-10-23 20:48:16,786 saving best model
2023-10-23 20:48:17,554 ----------------------------------------------------------------------------------------------------
2023-10-23 20:48:23,117 epoch 6 - iter 89/894 - loss 0.02315835 - time (sec): 5.56 - samples/sec: 1440.11 - lr: 0.000016 - momentum: 0.000000
2023-10-23 20:48:28,761 epoch 6 - iter 178/894 - loss 0.01840372 - time (sec): 11.21 - samples/sec: 1438.66 - lr: 0.000016 - momentum: 0.000000
2023-10-23 20:48:34,527 epoch 6 - iter 267/894 - loss 0.01933114 - time (sec): 16.97 - samples/sec: 1477.92 - lr: 0.000016 - momentum: 0.000000
2023-10-23 20:48:40,231 epoch 6 - iter 356/894 - loss 0.01924240 - time (sec): 22.68 - samples/sec: 1518.61 - lr: 0.000015 - momentum: 0.000000
2023-10-23 20:48:45,870 epoch 6 - iter 445/894 - loss 0.01840968 - time (sec): 28.31 - samples/sec: 1521.80 - lr: 0.000015 - momentum: 0.000000
2023-10-23 20:48:51,569 epoch 6 - iter 534/894 - loss 0.01732048 - time (sec): 34.01 - samples/sec: 1511.68 - lr: 0.000015 - momentum: 0.000000
2023-10-23 20:48:57,092 epoch 6 - iter 623/894 - loss 0.01729568 - time (sec): 39.54 - samples/sec: 1511.72 - lr: 0.000014 - momentum: 0.000000
2023-10-23 20:49:02,783 epoch 6 - iter 712/894 - loss 0.02022187 - time (sec): 45.23 - samples/sec: 1519.40 - lr: 0.000014 - momentum: 0.000000
2023-10-23 20:49:08,622 epoch 6 - iter 801/894 - loss 0.02054773 - time (sec): 51.07 - samples/sec: 1513.92 - lr: 0.000014 - momentum: 0.000000
2023-10-23 20:49:14,229 epoch 6 - iter 890/894 - loss 0.02089905 - time (sec): 56.67 - samples/sec: 1521.70 - lr: 0.000013 - momentum: 0.000000
2023-10-23 20:49:14,474 ----------------------------------------------------------------------------------------------------
2023-10-23 20:49:14,474 EPOCH 6 done: loss 0.0210 - lr: 0.000013
2023-10-23 20:49:21,009 DEV : loss 0.22474254667758942 - f1-score (micro avg) 0.7671
2023-10-23 20:49:21,030 ----------------------------------------------------------------------------------------------------
2023-10-23 20:49:26,585 epoch 7 - iter 89/894 - loss 0.00932732 - time (sec): 5.55 - samples/sec: 1525.93 - lr: 0.000013 - momentum: 0.000000
2023-10-23 20:49:32,387 epoch 7 - iter 178/894 - loss 0.01432918 - time (sec): 11.36 - samples/sec: 1517.03 - lr: 0.000013 - momentum: 0.000000
2023-10-23 20:49:38,406 epoch 7 - iter 267/894 - loss 0.01700073 - time (sec): 17.38 - samples/sec: 1533.79 - lr: 0.000012 - momentum: 0.000000
2023-10-23 20:49:44,028 epoch 7 - iter 356/894 - loss 0.01537951 - time (sec): 23.00 - samples/sec: 1525.27 - lr: 0.000012 - momentum: 0.000000
2023-10-23 20:49:49,679 epoch 7 - iter 445/894 - loss 0.01506681 - time (sec): 28.65 - samples/sec: 1519.99 - lr: 0.000012 - momentum: 0.000000
2023-10-23 20:49:55,356 epoch 7 - iter 534/894 - loss 0.01557746 - time (sec): 34.33 - samples/sec: 1523.54 - lr: 0.000011 - momentum: 0.000000
2023-10-23 20:50:01,041 epoch 7 - iter 623/894 - loss 0.01611775 - time (sec): 40.01 - samples/sec: 1520.46 - lr: 0.000011 - momentum: 0.000000
2023-10-23 20:50:06,612 epoch 7 - iter 712/894 - loss 0.01564183 - time (sec): 45.58 - samples/sec: 1517.64 - lr: 0.000011 - momentum: 0.000000
2023-10-23 20:50:12,188 epoch 7 - iter 801/894 - loss 0.01618850 - time (sec): 51.16 - samples/sec: 1520.34 - lr: 0.000010 - momentum: 0.000000
2023-10-23 20:50:17,797 epoch 7 - iter 890/894 - loss 0.01525803 - time (sec): 56.77 - samples/sec: 1519.19 - lr: 0.000010 - momentum: 0.000000
2023-10-23 20:50:18,038 ----------------------------------------------------------------------------------------------------
2023-10-23 20:50:18,038 EPOCH 7 done: loss 0.0152 - lr: 0.000010
2023-10-23 20:50:24,588 DEV : loss 0.25017356872558594 - f1-score (micro avg) 0.7642
2023-10-23 20:50:24,609 ----------------------------------------------------------------------------------------------------
2023-10-23 20:50:30,199 epoch 8 - iter 89/894 - loss 0.01612093 - time (sec): 5.59 - samples/sec: 1518.14 - lr: 0.000010 - momentum: 0.000000
2023-10-23 20:50:35,787 epoch 8 - iter 178/894 - loss 0.01639135 - time (sec): 11.18 - samples/sec: 1523.56 - lr: 0.000009 - momentum: 0.000000
2023-10-23 20:50:41,367 epoch 8 - iter 267/894 - loss 0.01252323 - time (sec): 16.76 - samples/sec: 1488.68 - lr: 0.000009 - momentum: 0.000000
2023-10-23 20:50:47,473 epoch 8 - iter 356/894 - loss 0.01262541 - time (sec): 22.86 - samples/sec: 1534.17 - lr: 0.000009 - momentum: 0.000000
2023-10-23 20:50:53,120 epoch 8 - iter 445/894 - loss 0.01149455 - time (sec): 28.51 - samples/sec: 1537.53 - lr: 0.000008 - momentum: 0.000000
2023-10-23 20:50:58,782 epoch 8 - iter 534/894 - loss 0.01152579 - time (sec): 34.17 - samples/sec: 1517.00 - lr: 0.000008 - momentum: 0.000000
2023-10-23 20:51:04,387 epoch 8 - iter 623/894 - loss 0.01078718 - time (sec): 39.78 - samples/sec: 1515.93 - lr: 0.000008 - momentum: 0.000000
2023-10-23 20:51:10,043 epoch 8 - iter 712/894 - loss 0.01121208 - time (sec): 45.43 - samples/sec: 1514.39 - lr: 0.000007 - momentum: 0.000000
2023-10-23 20:51:15,934 epoch 8 - iter 801/894 - loss 0.01081237 - time (sec): 51.32 - samples/sec: 1519.75 - lr: 0.000007 - momentum: 0.000000
2023-10-23 20:51:21,479 epoch 8 - iter 890/894 - loss 0.01053265 - time (sec): 56.87 - samples/sec: 1516.06 - lr: 0.000007 - momentum: 0.000000
2023-10-23 20:51:21,720 ----------------------------------------------------------------------------------------------------
2023-10-23 20:51:21,720 EPOCH 8 done: loss 0.0106 - lr: 0.000007
2023-10-23 20:51:27,969 DEV : loss 0.29330340027809143 - f1-score (micro avg) 0.7596
2023-10-23 20:51:27,990 ----------------------------------------------------------------------------------------------------
2023-10-23 20:51:33,729 epoch 9 - iter 89/894 - loss 0.00425727 - time (sec): 5.74 - samples/sec: 1410.23 - lr: 0.000006 - momentum: 0.000000
2023-10-23 20:51:39,392 epoch 9 - iter 178/894 - loss 0.00701630 - time (sec): 11.40 - samples/sec: 1444.35 - lr: 0.000006 - momentum: 0.000000
2023-10-23 20:51:45,246 epoch 9 - iter 267/894 - loss 0.00717471 - time (sec): 17.26 - samples/sec: 1473.49 - lr: 0.000006 - momentum: 0.000000
2023-10-23 20:51:50,879 epoch 9 - iter 356/894 - loss 0.00646948 - time (sec): 22.89 - samples/sec: 1489.65 - lr: 0.000005 - momentum: 0.000000
2023-10-23 20:51:56,536 epoch 9 - iter 445/894 - loss 0.00613541 - time (sec): 28.54 - samples/sec: 1503.56 - lr: 0.000005 - momentum: 0.000000
2023-10-23 20:52:02,312 epoch 9 - iter 534/894 - loss 0.00574876 - time (sec): 34.32 - samples/sec: 1509.25 - lr: 0.000005 - momentum: 0.000000
2023-10-23 20:52:08,201 epoch 9 - iter 623/894 - loss 0.00558351 - time (sec): 40.21 - samples/sec: 1520.38 - lr: 0.000004 - momentum: 0.000000
2023-10-23 20:52:13,789 epoch 9 - iter 712/894 - loss 0.00596474 - time (sec): 45.80 - samples/sec: 1514.27 - lr: 0.000004 - momentum: 0.000000
2023-10-23 20:52:19,295 epoch 9 - iter 801/894 - loss 0.00595036 - time (sec): 51.30 - samples/sec: 1512.79 - lr: 0.000004 - momentum: 0.000000
2023-10-23 20:52:25,015 epoch 9 - iter 890/894 - loss 0.00559881 - time (sec): 57.02 - samples/sec: 1513.03 - lr: 0.000003 - momentum: 0.000000
2023-10-23 20:52:25,245 ----------------------------------------------------------------------------------------------------
2023-10-23 20:52:25,245 EPOCH 9 done: loss 0.0056 - lr: 0.000003
2023-10-23 20:52:31,486 DEV : loss 0.27675920724868774 - f1-score (micro avg) 0.7733
2023-10-23 20:52:31,507 saving best model
2023-10-23 20:52:32,180 ----------------------------------------------------------------------------------------------------
2023-10-23 20:52:37,758 epoch 10 - iter 89/894 - loss 0.00098253 - time (sec): 5.58 - samples/sec: 1545.49 - lr: 0.000003 - momentum: 0.000000
2023-10-23 20:52:43,517 epoch 10 - iter 178/894 - loss 0.00095969 - time (sec): 11.34 - samples/sec: 1560.78 - lr: 0.000003 - momentum: 0.000000
2023-10-23 20:52:49,428 epoch 10 - iter 267/894 - loss 0.00128812 - time (sec): 17.25 - samples/sec: 1508.63 - lr: 0.000002 - momentum: 0.000000
2023-10-23 20:52:54,932 epoch 10 - iter 356/894 - loss 0.00278304 - time (sec): 22.75 - samples/sec: 1509.09 - lr: 0.000002 - momentum: 0.000000
2023-10-23 20:53:00,826 epoch 10 - iter 445/894 - loss 0.00345471 - time (sec): 28.64 - samples/sec: 1527.64 - lr: 0.000002 - momentum: 0.000000
2023-10-23 20:53:06,392 epoch 10 - iter 534/894 - loss 0.00389630 - time (sec): 34.21 - samples/sec: 1511.09 - lr: 0.000001 - momentum: 0.000000
2023-10-23 20:53:11,957 epoch 10 - iter 623/894 - loss 0.00368630 - time (sec): 39.78 - samples/sec: 1515.44 - lr: 0.000001 - momentum: 0.000000
2023-10-23 20:53:17,683 epoch 10 - iter 712/894 - loss 0.00344338 - time (sec): 45.50 - samples/sec: 1513.16 - lr: 0.000001 - momentum: 0.000000
2023-10-23 20:53:23,567 epoch 10 - iter 801/894 - loss 0.00365553 - time (sec): 51.39 - samples/sec: 1510.67 - lr: 0.000000 - momentum: 0.000000
2023-10-23 20:53:29,280 epoch 10 - iter 890/894 - loss 0.00368854 - time (sec): 57.10 - samples/sec: 1509.37 - lr: 0.000000 - momentum: 0.000000
2023-10-23 20:53:29,517 ----------------------------------------------------------------------------------------------------
2023-10-23 20:53:29,517 EPOCH 10 done: loss 0.0039 - lr: 0.000000
2023-10-23 20:53:35,745 DEV : loss 0.27931466698646545 - f1-score (micro avg) 0.7767
2023-10-23 20:53:35,765 saving best model
2023-10-23 20:53:37,037 ----------------------------------------------------------------------------------------------------
2023-10-23 20:53:37,038 Loading model from best epoch ...
2023-10-23 20:53:38,841 SequenceTagger predicts: Dictionary with 21 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org, S-prod, B-prod, E-prod, I-prod, S-time, B-time, E-time, I-time
2023-10-23 20:53:43,691
Results:
- F-score (micro) 0.7553
- F-score (macro) 0.6752
- Accuracy 0.6225
By class:
precision recall f1-score support
loc 0.8232 0.8674 0.8448 596
pers 0.6800 0.7658 0.7203 333
org 0.5379 0.5379 0.5379 132
prod 0.6316 0.5455 0.5854 66
time 0.7021 0.6735 0.6875 49
micro avg 0.7361 0.7755 0.7553 1176
macro avg 0.6750 0.6780 0.6752 1176
weighted avg 0.7349 0.7755 0.7540 1176
2023-10-23 20:53:43,691 ----------------------------------------------------------------------------------------------------