Update model

loss.tsv

@@ -0,0 +1,11 @@
+EPOCH	TIMESTAMP	BAD_EPOCHS	LEARNING_RATE	TRAIN_LOSS
+1	15:21:48	0	0.1000	0.27953692015655984
+2	15:31:22	0	0.1000	0.15365227826273328
+3	15:41:06	0	0.1000	0.12001519515322241
+4	15:50:54	0	0.1000	0.10328522111398844
+5	16:00:45	0	0.1000	0.09241386713466632
+6	16:10:29	0	0.1000	0.08505490679055881
+7	16:20:25	0	0.1000	0.07861811519301767
+8	16:30:21	0	0.1000	0.07341135664633389
+9	16:40:13	0	0.1000	0.06911533349940868
+10	16:50:01	0	0.1000	0.06593435410093888

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:698a7ca2b501a853c807de4defc42901968d932393a86d6a636d5ff4346dc54a
+size 494428971

training.log

@@ -0,0 +1,522 @@
+2022-08-06 15:12:29,180 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,182 Model: "SequenceTagger(
+  (embeddings): TransformerWordEmbeddings(
+    (model): BertModel(
+      (embeddings): BertEmbeddings(
+        (word_embeddings): Embedding(42000, 768, padding_idx=0)
+        (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()
+      )
+    )
+  )
+  (word_dropout): WordDropout(p=0.05)
+  (locked_dropout): LockedDropout(p=0.5)
+  (rnn): LSTM(768, 512, batch_first=True, bidirectional=True)
+  (linear): Linear(in_features=1024, out_features=30, bias=True)
+  (beta): 1.0
+  (weights): None
+  (weight_tensor) None
+)"
+2022-08-06 15:12:29,182 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,183 Corpus: "Corpus: 24000 train + 3000 dev + 3000 test sentences"
+2022-08-06 15:12:29,183 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,183 Parameters:
+2022-08-06 15:12:29,183  - learning_rate: "0.1"
+2022-08-06 15:12:29,183  - mini_batch_size: "8"
+2022-08-06 15:12:29,183  - patience: "3"
+2022-08-06 15:12:29,183  - anneal_factor: "0.5"
+2022-08-06 15:12:29,183  - max_epochs: "10"
+2022-08-06 15:12:29,183  - shuffle: "True"
+2022-08-06 15:12:29,183  - train_with_dev: "True"
+2022-08-06 15:12:29,183  - batch_growth_annealing: "False"
+2022-08-06 15:12:29,183 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,183 Model training base path: "data/pos-Uppsala/model"
+2022-08-06 15:12:29,183 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,183 Device: cuda:0
+2022-08-06 15:12:29,183 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:12:29,184 Embeddings storage mode: gpu
+2022-08-06 15:12:29,185 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:13:18,972 epoch 1 - iter 337/3375 - loss 0.74289984 - samples/sec: 54.18 - lr: 0.100000
+2022-08-06 15:14:15,036 epoch 1 - iter 674/3375 - loss 0.53599298 - samples/sec: 48.11 - lr: 0.100000
+2022-08-06 15:15:12,610 epoch 1 - iter 1011/3375 - loss 0.45754038 - samples/sec: 46.85 - lr: 0.100000
+2022-08-06 15:16:09,043 epoch 1 - iter 1348/3375 - loss 0.40111208 - samples/sec: 47.79 - lr: 0.100000
+2022-08-06 15:17:04,137 epoch 1 - iter 1685/3375 - loss 0.36712663 - samples/sec: 48.96 - lr: 0.100000
+2022-08-06 15:17:58,402 epoch 1 - iter 2022/3375 - loss 0.34049225 - samples/sec: 49.70 - lr: 0.100000
+2022-08-06 15:18:55,276 epoch 1 - iter 2359/3375 - loss 0.32076226 - samples/sec: 47.42 - lr: 0.100000
+2022-08-06 15:19:49,979 epoch 1 - iter 2696/3375 - loss 0.31015506 - samples/sec: 49.31 - lr: 0.100000
+2022-08-06 15:20:48,410 epoch 1 - iter 3033/3375 - loss 0.29391699 - samples/sec: 46.16 - lr: 0.100000
+2022-08-06 15:21:47,572 epoch 1 - iter 3370/3375 - loss 0.27989028 - samples/sec: 45.59 - lr: 0.100000
+2022-08-06 15:21:48,555 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:21:48,555 EPOCH 1 done: loss 0.2795 - lr 0.1000000
+2022-08-06 15:21:48,555 BAD EPOCHS (no improvement): 0
+2022-08-06 15:21:48,555 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:22:45,590 epoch 2 - iter 337/3375 - loss 0.18085661 - samples/sec: 47.29 - lr: 0.100000
+2022-08-06 15:23:42,698 epoch 2 - iter 674/3375 - loss 0.17216272 - samples/sec: 47.23 - lr: 0.100000
+2022-08-06 15:24:38,534 epoch 2 - iter 1011/3375 - loss 0.16694117 - samples/sec: 48.31 - lr: 0.100000
+2022-08-06 15:25:36,464 epoch 2 - iter 1348/3375 - loss 0.16500505 - samples/sec: 46.56 - lr: 0.100000
+2022-08-06 15:26:32,174 epoch 2 - iter 1685/3375 - loss 0.16167195 - samples/sec: 48.42 - lr: 0.100000
+2022-08-06 15:27:28,418 epoch 2 - iter 2022/3375 - loss 0.15991464 - samples/sec: 47.96 - lr: 0.100000
+2022-08-06 15:28:30,730 epoch 2 - iter 2359/3375 - loss 0.15942296 - samples/sec: 43.29 - lr: 0.100000
+2022-08-06 15:29:27,444 epoch 2 - iter 2696/3375 - loss 0.15779417 - samples/sec: 47.56 - lr: 0.100000
+2022-08-06 15:30:25,187 epoch 2 - iter 3033/3375 - loss 0.15553239 - samples/sec: 46.71 - lr: 0.100000
+2022-08-06 15:31:21,714 epoch 2 - iter 3370/3375 - loss 0.15352182 - samples/sec: 47.72 - lr: 0.100000
+2022-08-06 15:31:22,712 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:31:22,712 EPOCH 2 done: loss 0.1537 - lr 0.1000000
+2022-08-06 15:31:22,712 BAD EPOCHS (no improvement): 0
+2022-08-06 15:31:22,712 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:32:23,790 epoch 3 - iter 337/3375 - loss 0.11867195 - samples/sec: 44.16 - lr: 0.100000
+2022-08-06 15:33:21,161 epoch 3 - iter 674/3375 - loss 0.11878234 - samples/sec: 47.02 - lr: 0.100000
+2022-08-06 15:34:20,702 epoch 3 - iter 1011/3375 - loss 0.11942785 - samples/sec: 45.31 - lr: 0.100000
+2022-08-06 15:35:18,259 epoch 3 - iter 1348/3375 - loss 0.11958903 - samples/sec: 46.86 - lr: 0.100000
+2022-08-06 15:36:16,967 epoch 3 - iter 1685/3375 - loss 0.11914369 - samples/sec: 45.94 - lr: 0.100000
+2022-08-06 15:37:13,560 epoch 3 - iter 2022/3375 - loss 0.11916365 - samples/sec: 47.66 - lr: 0.100000
+2022-08-06 15:38:10,624 epoch 3 - iter 2359/3375 - loss 0.12096981 - samples/sec: 47.27 - lr: 0.100000
+2022-08-06 15:39:10,034 epoch 3 - iter 2696/3375 - loss 0.11987245 - samples/sec: 45.40 - lr: 0.100000
+2022-08-06 15:40:07,877 epoch 3 - iter 3033/3375 - loss 0.11973164 - samples/sec: 46.63 - lr: 0.100000
+2022-08-06 15:41:05,610 epoch 3 - iter 3370/3375 - loss 0.12003917 - samples/sec: 46.72 - lr: 0.100000
+2022-08-06 15:41:06,450 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:41:06,450 EPOCH 3 done: loss 0.1200 - lr 0.1000000
+2022-08-06 15:41:06,450 BAD EPOCHS (no improvement): 0
+2022-08-06 15:41:06,451 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:42:04,442 epoch 4 - iter 337/3375 - loss 0.09805702 - samples/sec: 46.51 - lr: 0.100000
+2022-08-06 15:43:05,164 epoch 4 - iter 674/3375 - loss 0.09888569 - samples/sec: 44.42 - lr: 0.100000
+2022-08-06 15:44:02,546 epoch 4 - iter 1011/3375 - loss 0.10053644 - samples/sec: 47.01 - lr: 0.100000
+2022-08-06 15:45:01,384 epoch 4 - iter 1348/3375 - loss 0.10119574 - samples/sec: 45.84 - lr: 0.100000
+2022-08-06 15:46:00,229 epoch 4 - iter 1685/3375 - loss 0.10374826 - samples/sec: 45.84 - lr: 0.100000
+2022-08-06 15:46:59,791 epoch 4 - iter 2022/3375 - loss 0.10405522 - samples/sec: 45.28 - lr: 0.100000
+2022-08-06 15:47:57,607 epoch 4 - iter 2359/3375 - loss 0.10411718 - samples/sec: 46.65 - lr: 0.100000
+2022-08-06 15:48:55,410 epoch 4 - iter 2696/3375 - loss 0.10394934 - samples/sec: 46.66 - lr: 0.100000
+2022-08-06 15:49:56,783 epoch 4 - iter 3033/3375 - loss 0.10374714 - samples/sec: 43.95 - lr: 0.100000
+2022-08-06 15:50:54,113 epoch 4 - iter 3370/3375 - loss 0.10333066 - samples/sec: 47.05 - lr: 0.100000
+2022-08-06 15:50:54,961 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:50:54,961 EPOCH 4 done: loss 0.1033 - lr 0.1000000
+2022-08-06 15:50:54,961 BAD EPOCHS (no improvement): 0
+2022-08-06 15:50:54,961 ----------------------------------------------------------------------------------------------------
+2022-08-06 15:51:52,151 epoch 5 - iter 337/3375 - loss 0.08744228 - samples/sec: 47.17 - lr: 0.100000
+2022-08-06 15:52:49,910 epoch 5 - iter 674/3375 - loss 0.08896766 - samples/sec: 46.70 - lr: 0.100000
+2022-08-06 15:53:50,861 epoch 5 - iter 1011/3375 - loss 0.09000325 - samples/sec: 44.25 - lr: 0.100000
+2022-08-06 15:54:48,357 epoch 5 - iter 1348/3375 - loss 0.09103779 - samples/sec: 46.91 - lr: 0.100000
+2022-08-06 15:55:48,122 epoch 5 - iter 1685/3375 - loss 0.09107958 - samples/sec: 45.13 - lr: 0.100000
+2022-08-06 15:56:49,324 epoch 5 - iter 2022/3375 - loss 0.09135469 - samples/sec: 44.07 - lr: 0.100000
+2022-08-06 15:57:47,393 epoch 5 - iter 2359/3375 - loss 0.09172710 - samples/sec: 46.45 - lr: 0.100000
+2022-08-06 15:58:45,694 epoch 5 - iter 2696/3375 - loss 0.09238154 - samples/sec: 46.27 - lr: 0.100000
+2022-08-06 15:59:42,885 epoch 5 - iter 3033/3375 - loss 0.09253470 - samples/sec: 47.16 - lr: 0.100000
+2022-08-06 16:00:44,492 epoch 5 - iter 3370/3375 - loss 0.09240350 - samples/sec: 43.78 - lr: 0.100000
+2022-08-06 16:00:45,327 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:00:45,328 EPOCH 5 done: loss 0.0924 - lr 0.1000000
+2022-08-06 16:00:45,328 BAD EPOCHS (no improvement): 0
+2022-08-06 16:00:45,328 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:01:42,167 epoch 6 - iter 337/3375 - loss 0.08075428 - samples/sec: 47.46 - lr: 0.100000
+2022-08-06 16:02:39,509 epoch 6 - iter 674/3375 - loss 0.08099115 - samples/sec: 47.04 - lr: 0.100000
+2022-08-06 16:03:37,688 epoch 6 - iter 1011/3375 - loss 0.08140463 - samples/sec: 46.36 - lr: 0.100000
+2022-08-06 16:04:38,640 epoch 6 - iter 1348/3375 - loss 0.08175190 - samples/sec: 44.25 - lr: 0.100000
+2022-08-06 16:05:35,459 epoch 6 - iter 1685/3375 - loss 0.08233525 - samples/sec: 47.47 - lr: 0.100000
+2022-08-06 16:06:33,941 epoch 6 - iter 2022/3375 - loss 0.08333964 - samples/sec: 46.12 - lr: 0.100000
+2022-08-06 16:07:34,247 epoch 6 - iter 2359/3375 - loss 0.08370656 - samples/sec: 44.73 - lr: 0.100000
+2022-08-06 16:08:32,546 epoch 6 - iter 2696/3375 - loss 0.08503503 - samples/sec: 46.27 - lr: 0.100000
+2022-08-06 16:09:30,447 epoch 6 - iter 3033/3375 - loss 0.08526801 - samples/sec: 46.58 - lr: 0.100000
+2022-08-06 16:10:29,216 epoch 6 - iter 3370/3375 - loss 0.08506276 - samples/sec: 45.90 - lr: 0.100000
+2022-08-06 16:10:29,946 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:10:29,947 EPOCH 6 done: loss 0.0851 - lr 0.1000000
+2022-08-06 16:10:29,947 BAD EPOCHS (no improvement): 0
+2022-08-06 16:10:29,947 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:11:31,042 epoch 7 - iter 337/3375 - loss 0.07328964 - samples/sec: 44.15 - lr: 0.100000
+2022-08-06 16:12:31,218 epoch 7 - iter 674/3375 - loss 0.07556648 - samples/sec: 44.82 - lr: 0.100000
+2022-08-06 16:13:28,468 epoch 7 - iter 1011/3375 - loss 0.07578294 - samples/sec: 47.11 - lr: 0.100000
+2022-08-06 16:14:28,318 epoch 7 - iter 1348/3375 - loss 0.07581855 - samples/sec: 45.07 - lr: 0.100000
+2022-08-06 16:15:27,119 epoch 7 - iter 1685/3375 - loss 0.07674717 - samples/sec: 45.87 - lr: 0.100000
+2022-08-06 16:16:25,205 epoch 7 - iter 2022/3375 - loss 0.07800463 - samples/sec: 46.44 - lr: 0.100000
+2022-08-06 16:17:25,635 epoch 7 - iter 2359/3375 - loss 0.07788540 - samples/sec: 44.64 - lr: 0.100000
+2022-08-06 16:18:25,934 epoch 7 - iter 2696/3375 - loss 0.07823310 - samples/sec: 44.73 - lr: 0.100000
+2022-08-06 16:19:25,742 epoch 7 - iter 3033/3375 - loss 0.07862489 - samples/sec: 45.10 - lr: 0.100000
+2022-08-06 16:20:24,514 epoch 7 - iter 3370/3375 - loss 0.07864779 - samples/sec: 45.89 - lr: 0.100000
+2022-08-06 16:20:25,316 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:20:25,317 EPOCH 7 done: loss 0.0786 - lr 0.1000000
+2022-08-06 16:20:25,317 BAD EPOCHS (no improvement): 0
+2022-08-06 16:20:25,317 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:21:23,040 epoch 8 - iter 337/3375 - loss 0.06876001 - samples/sec: 46.73 - lr: 0.100000
+2022-08-06 16:22:25,028 epoch 8 - iter 674/3375 - loss 0.06867038 - samples/sec: 43.51 - lr: 0.100000
+2022-08-06 16:23:25,046 epoch 8 - iter 1011/3375 - loss 0.07011779 - samples/sec: 44.94 - lr: 0.100000
+2022-08-06 16:24:23,287 epoch 8 - iter 1348/3375 - loss 0.07118411 - samples/sec: 46.31 - lr: 0.100000
+2022-08-06 16:25:24,939 epoch 8 - iter 1685/3375 - loss 0.07159055 - samples/sec: 43.75 - lr: 0.100000
+2022-08-06 16:26:23,316 epoch 8 - iter 2022/3375 - loss 0.07167687 - samples/sec: 46.21 - lr: 0.100000
+2022-08-06 16:27:22,234 epoch 8 - iter 2359/3375 - loss 0.07190781 - samples/sec: 45.78 - lr: 0.100000
+2022-08-06 16:28:20,921 epoch 8 - iter 2696/3375 - loss 0.07263123 - samples/sec: 45.96 - lr: 0.100000
+2022-08-06 16:29:21,637 epoch 8 - iter 3033/3375 - loss 0.07345723 - samples/sec: 44.42 - lr: 0.100000
+2022-08-06 16:30:20,403 epoch 8 - iter 3370/3375 - loss 0.07338627 - samples/sec: 45.90 - lr: 0.100000
+2022-08-06 16:30:21,375 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:30:21,375 EPOCH 8 done: loss 0.0734 - lr 0.1000000
+2022-08-06 16:30:21,375 BAD EPOCHS (no improvement): 0
+2022-08-06 16:30:21,376 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:31:18,803 epoch 9 - iter 337/3375 - loss 0.06314787 - samples/sec: 46.97 - lr: 0.100000
+2022-08-06 16:32:16,661 epoch 9 - iter 674/3375 - loss 0.06638022 - samples/sec: 46.62 - lr: 0.100000
+2022-08-06 16:33:15,745 epoch 9 - iter 1011/3375 - loss 0.06547021 - samples/sec: 45.65 - lr: 0.100000
+2022-08-06 16:34:14,632 epoch 9 - iter 1348/3375 - loss 0.06593581 - samples/sec: 45.81 - lr: 0.100000
+2022-08-06 16:35:13,668 epoch 9 - iter 1685/3375 - loss 0.06772817 - samples/sec: 45.69 - lr: 0.100000
+2022-08-06 16:36:15,567 epoch 9 - iter 2022/3375 - loss 0.06808051 - samples/sec: 43.58 - lr: 0.100000
+2022-08-06 16:37:16,651 epoch 9 - iter 2359/3375 - loss 0.06796916 - samples/sec: 44.16 - lr: 0.100000
+2022-08-06 16:38:14,513 epoch 9 - iter 2696/3375 - loss 0.06906572 - samples/sec: 46.62 - lr: 0.100000
+2022-08-06 16:39:13,107 epoch 9 - iter 3033/3375 - loss 0.06917054 - samples/sec: 46.03 - lr: 0.100000
+2022-08-06 16:40:12,475 epoch 9 - iter 3370/3375 - loss 0.06913866 - samples/sec: 45.43 - lr: 0.100000
+2022-08-06 16:40:13,344 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:40:13,344 EPOCH 9 done: loss 0.0691 - lr 0.1000000
+2022-08-06 16:40:13,344 BAD EPOCHS (no improvement): 0
+2022-08-06 16:40:13,345 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:41:11,629 epoch 10 - iter 337/3375 - loss 0.05727560 - samples/sec: 46.28 - lr: 0.100000
+2022-08-06 16:42:09,047 epoch 10 - iter 674/3375 - loss 0.06063155 - samples/sec: 46.98 - lr: 0.100000
+2022-08-06 16:43:09,515 epoch 10 - iter 1011/3375 - loss 0.06369582 - samples/sec: 44.61 - lr: 0.100000
+2022-08-06 16:44:07,978 epoch 10 - iter 1348/3375 - loss 0.06421773 - samples/sec: 46.14 - lr: 0.100000
+2022-08-06 16:45:07,015 epoch 10 - iter 1685/3375 - loss 0.06397856 - samples/sec: 45.69 - lr: 0.100000
+2022-08-06 16:46:05,736 epoch 10 - iter 2022/3375 - loss 0.06424947 - samples/sec: 45.93 - lr: 0.100000
+2022-08-06 16:47:06,945 epoch 10 - iter 2359/3375 - loss 0.06511606 - samples/sec: 44.07 - lr: 0.100000
+2022-08-06 16:48:05,819 epoch 10 - iter 2696/3375 - loss 0.06574495 - samples/sec: 45.82 - lr: 0.100000
+2022-08-06 16:49:03,924 epoch 10 - iter 3033/3375 - loss 0.06552271 - samples/sec: 46.42 - lr: 0.100000
+2022-08-06 16:50:00,641 epoch 10 - iter 3370/3375 - loss 0.06594147 - samples/sec: 47.56 - lr: 0.100000
+2022-08-06 16:50:01,493 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:50:01,493 EPOCH 10 done: loss 0.0659 - lr 0.1000000
+2022-08-06 16:50:01,493 BAD EPOCHS (no improvement): 0
+2022-08-06 16:50:02,708 ----------------------------------------------------------------------------------------------------
+2022-08-06 16:50:02,709 Testing using last state of model ...
+2022-08-06 16:53:40,214 0.9632	0.9632	0.9632	0.9632
+2022-08-06 16:53:40,215 
+Results:
+- F-score (micro) 0.9632
+- F-score (macro) 0.9031
+- Accuracy 0.9632
+
+By class:
+              precision    recall  f1-score   support
+
+      N_SING     0.9691    0.9565    0.9627     30553
+           P     0.9560    0.9937    0.9745      9951
+        DELM     0.9936    0.9906    0.9921      8122
+         ADJ     0.9205    0.9152    0.9179      7466
+         CON     0.9892    0.9799    0.9845      6823
+        N_PL     0.9476    0.9642    0.9558      5163
+        V_PA     0.9729    0.9746    0.9737      2873
+       V_PRS     0.9825    0.9898    0.9861      2841
+         PRO     0.9656    0.9455    0.9555      2258
+         NUM     0.9937    0.9933    0.9935      2232
+         DET     0.9423    0.9698    0.9559      1853
+      CLITIC     0.9992    1.0000    0.9996      1259
+        V_PP     0.9699    0.9741    0.9720      1158
+       V_SUB     0.9620    0.9573    0.9596      1031
+         ADV     0.7784    0.8182    0.7978       880
+    ADV_TIME     0.9126    0.9611    0.9363       489
+       V_AUX     0.9869    0.9974    0.9921       379
+     ADJ_SUP     0.9851    0.9815    0.9833       270
+    ADJ_CMPR     0.9246    0.9534    0.9388       193
+     ADJ_INO     0.7294    0.7381    0.7337       168
+     ADV_NEG     0.9034    0.8792    0.8912       149
+       ADV_I     0.8926    0.7714    0.8276       140
+          FW     0.6893    0.5772    0.6283       123
+    ADV_COMP     0.8267    0.8158    0.8212        76
+     ADV_LOC     0.9722    0.9589    0.9655        73
+       V_IMP     0.7292    0.6250    0.6731        56
+        PREV     0.9286    0.8125    0.8667        32
+         INT     0.9231    0.5000    0.6486        24
+
+   micro avg     0.9632    0.9632    0.9632     86635
+   macro avg     0.9195    0.8926    0.9031     86635
+weighted avg     0.9633    0.9632    0.9631     86635
+ samples avg     0.9632    0.9632    0.9632     86635
+
+2022-08-06 16:53:40,215 ----------------------------------------------------------------------------------------------------