add Model

loss.tsv

@@ -0,0 +1,3 @@
+
+6	07:10:41	0	0.1000	0.07391205054453862	0.05534437298774719	0.8262	0.7999	0.8129	0.6949
+7	07:59:57	0	0.1000	0.07154660764968938	0.05505584925413132	0.8469	0.7901	0.8175	0.7019

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15da00307b7fec8958dafca847190ec19c2eaf49cfb1a3ac53d503c483e69152
+size 413753973

training.log

@@ -0,0 +1,96 @@
+2022-03-25 06:21:37,092 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,098 Model: "SequenceTagger(
+  (embeddings): StackedEmbeddings(
+    (list_embedding_0): WordEmbeddings(
+      'fa'
+      (embedding): Embedding(56850, 300)
+    )
+    (list_embedding_1): FlairEmbeddings(
+      (lm): LanguageModel(
+        (drop): Dropout(p=0.1, inplace=False)
+        (encoder): Embedding(5105, 100)
+        (rnn): LSTM(100, 2048)
+        (decoder): Linear(in_features=2048, out_features=5105, bias=True)
+      )
+    )
+    (list_embedding_2): FlairEmbeddings(
+      (lm): LanguageModel(
+        (drop): Dropout(p=0.1, inplace=False)
+        (encoder): Embedding(5105, 100)
+        (rnn): LSTM(100, 2048)
+        (decoder): Linear(in_features=2048, out_features=5105, bias=True)
+      )
+    )
+  )
+  (word_dropout): WordDropout(p=0.05)
+  (locked_dropout): LockedDropout(p=0.5)
+  (embedding2nn): Linear(in_features=4396, out_features=4396, bias=True)
+  (rnn): LSTM(4396, 256, batch_first=True, bidirectional=True)
+  (linear): Linear(in_features=512, out_features=18, bias=True)
+  (beta): 1.0
+  (weights): None
+  (weight_tensor) None
+)"
+2022-03-25 06:21:37,103 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,108 Corpus: "Corpus: 23060 train + 4070 dev + 4150 test sentences"
+2022-03-25 06:21:37,111 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,115 Parameters:
+2022-03-25 06:21:37,117  - learning_rate: "0.1"
+2022-03-25 06:21:37,119  - mini_batch_size: "4"
+2022-03-25 06:21:37,122  - patience: "3"
+2022-03-25 06:21:37,125  - anneal_factor: "0.5"
+2022-03-25 06:21:37,127  - max_epochs: "10"
+2022-03-25 06:21:37,129  - shuffle: "True"
+2022-03-25 06:21:37,132  - train_with_dev: "False"
+2022-03-25 06:21:37,135  - batch_growth_annealing: "False"
+2022-03-25 06:21:37,137 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,140 Model training base path: "/content/gdrive/MyDrive/project/data/ner/model"
+2022-03-25 06:21:37,142 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,145 Device: cuda:0
+2022-03-25 06:21:37,148 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:21:37,150 Embeddings storage mode: none
+2022-03-25 06:21:37,398 ----------------------------------------------------------------------------------------------------
+2022-03-25 06:25:43,993 epoch 6 - iter 576/5765 - loss 0.07042695 - samples/sec: 9.35 - lr: 0.100000
+2022-03-25 06:29:47,830 epoch 6 - iter 1152/5765 - loss 0.07287426 - samples/sec: 9.49 - lr: 0.100000
+2022-03-25 06:34:02,575 epoch 6 - iter 1728/5765 - loss 0.07379352 - samples/sec: 9.08 - lr: 0.100000
+2022-03-25 06:38:22,556 epoch 6 - iter 2304/5765 - loss 0.07346159 - samples/sec: 8.90 - lr: 0.100000
+2022-03-25 06:42:37,312 epoch 6 - iter 2880/5765 - loss 0.07318457 - samples/sec: 9.08 - lr: 0.100000
+2022-03-25 06:47:03,459 epoch 6 - iter 3456/5765 - loss 0.07343553 - samples/sec: 8.69 - lr: 0.100000
+2022-03-25 06:51:22,020 epoch 6 - iter 4032/5765 - loss 0.07360594 - samples/sec: 8.95 - lr: 0.100000
+2022-03-25 06:55:36,718 epoch 6 - iter 4608/5765 - loss 0.07332146 - samples/sec: 9.08 - lr: 0.100000
+2022-03-25 07:00:02,036 epoch 6 - iter 5184/5765 - loss 0.07376939 - samples/sec: 8.72 - lr: 0.100000
+2022-03-25 07:04:32,247 epoch 6 - iter 5760/5765 - loss 0.07393306 - samples/sec: 8.56 - lr: 0.100000
+2022-03-25 07:04:35,408 ----------------------------------------------------------------------------------------------------
+2022-03-25 07:04:35,411 EPOCH 6 done: loss 0.0739 - lr 0.1000000
+2022-03-25 07:10:41,676 DEV : loss 0.05534437298774719 - f1-score (micro avg)  0.8129
+2022-03-25 07:10:41,758 BAD EPOCHS (no improvement): 0
+2022-03-25 07:10:43,386 saving best model
+2022-03-25 07:10:45,085 ----------------------------------------------------------------------------------------------------
+2022-03-25 07:15:08,362 epoch 7 - iter 576/5765 - loss 0.06846625 - samples/sec: 8.75 - lr: 0.100000
+2022-03-25 07:19:20,901 epoch 7 - iter 1152/5765 - loss 0.07066517 - samples/sec: 9.16 - lr: 0.100000
+2022-03-25 07:23:45,054 epoch 7 - iter 1728/5765 - loss 0.07063719 - samples/sec: 8.76 - lr: 0.100000
+2022-03-25 07:27:58,256 epoch 7 - iter 2304/5765 - loss 0.07101257 - samples/sec: 9.14 - lr: 0.100000
+2022-03-25 07:32:05,224 epoch 7 - iter 2880/5765 - loss 0.07072532 - samples/sec: 9.37 - lr: 0.100000
+2022-03-25 07:36:19,489 epoch 7 - iter 3456/5765 - loss 0.07040446 - samples/sec: 9.10 - lr: 0.100000
+2022-03-25 07:40:49,459 epoch 7 - iter 4032/5765 - loss 0.07117669 - samples/sec: 8.57 - lr: 0.100000
+2022-03-25 07:45:06,879 epoch 7 - iter 4608/5765 - loss 0.07147140 - samples/sec: 8.99 - lr: 0.100000
+2022-03-25 07:49:20,561 epoch 7 - iter 5184/5765 - loss 0.07151126 - samples/sec: 9.12 - lr: 0.100000
+2022-03-25 07:53:46,941 epoch 7 - iter 5760/5765 - loss 0.07156780 - samples/sec: 8.69 - lr: 0.100000
+2022-03-25 07:53:49,751 ----------------------------------------------------------------------------------------------------
+2022-03-25 07:53:49,759 EPOCH 7 done: loss 0.0715 - lr 0.1000000
+2022-03-25 07:59:57,729 DEV : loss 0.05505584925413132 - f1-score (micro avg)  0.8175
+2022-03-25 07:59:57,813 BAD EPOCHS (no improvement): 0
+2022-03-25 07:59:59,910 saving best model
+2022-03-25 08:00:01,383 ----------------------------------------------------------------------------------------------------
+2022-03-25 08:04:20,017 epoch 8 - iter 576/5765 - loss 0.06595992 - samples/sec: 8.91 - lr: 0.100000
+2022-03-25 08:08:34,362 epoch 8 - iter 1152/5765 - loss 0.06695618 - samples/sec: 9.10 - lr: 0.100000
+2022-03-25 08:13:01,311 epoch 8 - iter 1728/5765 - loss 0.06868385 - samples/sec: 8.66 - lr: 0.100000
+2022-03-25 08:17:19,699 epoch 8 - iter 2304/5765 - loss 0.06855573 - samples/sec: 8.95 - lr: 0.100000
+2022-03-25 08:21:39,417 epoch 8 - iter 2880/5765 - loss 0.06828534 - samples/sec: 8.91 - lr: 0.100000
+2022-03-25 08:25:58,656 epoch 8 - iter 3456/5765 - loss 0.06920992 - samples/sec: 8.92 - lr: 0.100000
+2022-03-25 08:30:19,059 epoch 8 - iter 4032/5765 - loss 0.06966214 - samples/sec: 8.88 - lr: 0.100000
+2022-03-25 08:34:32,114 epoch 8 - iter 4608/5765 - loss 0.06999527 - samples/sec: 9.14 - lr: 0.100000
+2022-03-25 08:38:45,063 epoch 8 - iter 5184/5765 - loss 0.07041313 - samples/sec: 9.15 - lr: 0.100000
+2022-03-25 08:42:53,891 epoch 8 - iter 5760/5765 - loss 0.07067043 - samples/sec: 9.30 - lr: 0.100000
+2022-03-25 08:42:56,995 ----------------------------------------------------------------------------------------------------
+2022-03-25 08:42:56,998 EPOCH 8 done: loss 0.0707 - lr 0.1000000