Hugging Face's logo

Aleksandar
/
bert-srb-ner

Token Classification
Transformers
PyTorch
Safetensors
bert
Generated from Trainer
Inference Endpoints
Model card Files Community
1
Aleksandar commited on
Commit
1774bdf
•
1 Parent(s): 12be85e

add model

Browse files
Files changed (4) hide show
  1. README.md +28 -18
  2. config.json +7 -7
  3. pytorch_model.bin +1 -1
  4. training_args.bin +1 -1
README.md CHANGED
@@ -21,7 +21,7 @@ model_index:
21
  metric:
22
  name: Accuracy
23
  type: accuracy
24
- value: 0.9542715764169646
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,11 +31,11 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model was trained from scratch on the wikiann dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 0.3045
35
- - Precision: 0.8922
36
- - Recall: 0.9050
37
- - F1: 0.8986
38
- - Accuracy: 0.9543
39
 
40
  ## Model description
41
 
@@ -55,27 +55,37 @@ More information needed
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 2e-05
58
- - train_batch_size: 16
59
  - eval_batch_size: 8
60
  - seed: 42
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
- - num_epochs: 10
64
 
65
  ### Training results
66
 
67
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
68
  |:-------------:|:-----:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
69
- | 0.276 | 1.0 | 1250 | 0.2359 | 0.8355 | 0.8334 | 0.8344 | 0.9276 |
70
- | 0.1722 | 2.0 | 2500 | 0.2016 | 0.8731 | 0.8685 | 0.8708 | 0.9426 |
71
- | 0.1155 | 3.0 | 3750 | 0.1897 | 0.8707 | 0.8860 | 0.8783 | 0.9463 |
72
- | 0.0849 | 4.0 | 5000 | 0.2151 | 0.8755 | 0.8980 | 0.8866 | 0.9494 |
73
- | 0.0554 | 5.0 | 6250 | 0.2373 | 0.8820 | 0.8923 | 0.8871 | 0.9495 |
74
- | 0.039 | 6.0 | 7500 | 0.2644 | 0.8808 | 0.8953 | 0.8880 | 0.9505 |
75
- | 0.0286 | 7.0 | 8750 | 0.2737 | 0.8915 | 0.8961 | 0.8938 | 0.9520 |
76
- | 0.018 | 8.0 | 10000 | 0.2879 | 0.8860 | 0.9039 | 0.8948 | 0.9526 |
77
- | 0.0116 | 9.0 | 11250 | 0.2973 | 0.8930 | 0.9032 | 0.8981 | 0.9542 |
78
- | 0.0079 | 10.0 | 12500 | 0.3045 | 0.8922 | 0.9050 | 0.8986 | 0.9543 |
 
 
 
 
 
 
 
 
 
 
79
 
80
 
81
  ### Framework versions
 
21
  metric:
22
  name: Accuracy
23
  type: accuracy
24
+ value: 0.9546696220907545
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model was trained from scratch on the wikiann dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.3561
35
+ - Precision: 0.8909
36
+ - Recall: 0.9082
37
+ - F1: 0.8995
38
+ - Accuracy: 0.9547
39
 
40
  ## Model description
41
 
 
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 2e-05
58
+ - train_batch_size: 32
59
  - eval_batch_size: 8
60
  - seed: 42
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
+ - num_epochs: 20
64
 
65
  ### Training results
66
 
67
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
68
  |:-------------:|:-----:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
69
+ | 0.3907 | 1.0 | 625 | 0.2316 | 0.8255 | 0.8314 | 0.8285 | 0.9259 |
70
+ | 0.2091 | 2.0 | 1250 | 0.1920 | 0.8598 | 0.8731 | 0.8664 | 0.9420 |
71
+ | 0.1562 | 3.0 | 1875 | 0.1833 | 0.8608 | 0.8820 | 0.8713 | 0.9441 |
72
+ | 0.0919 | 4.0 | 2500 | 0.1985 | 0.8712 | 0.8886 | 0.8798 | 0.9476 |
73
+ | 0.0625 | 5.0 | 3125 | 0.2195 | 0.8762 | 0.8923 | 0.8842 | 0.9485 |
74
+ | 0.0545 | 6.0 | 3750 | 0.2320 | 0.8706 | 0.9004 | 0.8852 | 0.9495 |
75
+ | 0.0403 | 7.0 | 4375 | 0.2459 | 0.8817 | 0.8957 | 0.8887 | 0.9505 |
76
+ | 0.0269 | 8.0 | 5000 | 0.2603 | 0.8813 | 0.9021 | 0.8916 | 0.9516 |
77
+ | 0.0193 | 9.0 | 5625 | 0.2916 | 0.8812 | 0.8949 | 0.8880 | 0.9500 |
78
+ | 0.0162 | 10.0 | 6250 | 0.2938 | 0.8814 | 0.9025 | 0.8918 | 0.9520 |
79
+ | 0.0134 | 11.0 | 6875 | 0.3330 | 0.8809 | 0.8961 | 0.8885 | 0.9497 |
80
+ | 0.0076 | 12.0 | 7500 | 0.3141 | 0.8840 | 0.9025 | 0.8932 | 0.9524 |
81
+ | 0.0069 | 13.0 | 8125 | 0.3292 | 0.8819 | 0.9065 | 0.8940 | 0.9535 |
82
+ | 0.0053 | 14.0 | 8750 | 0.3454 | 0.8844 | 0.9018 | 0.8930 | 0.9523 |
83
+ | 0.0038 | 15.0 | 9375 | 0.3519 | 0.8912 | 0.9061 | 0.8986 | 0.9539 |
84
+ | 0.0034 | 16.0 | 10000 | 0.3437 | 0.8894 | 0.9038 | 0.8965 | 0.9539 |
85
+ | 0.0024 | 17.0 | 10625 | 0.3518 | 0.8896 | 0.9072 | 0.8983 | 0.9543 |
86
+ | 0.0018 | 18.0 | 11250 | 0.3572 | 0.8877 | 0.9072 | 0.8973 | 0.9543 |
87
+ | 0.0015 | 19.0 | 11875 | 0.3554 | 0.8910 | 0.9081 | 0.8994 | 0.9549 |
88
+ | 0.0011 | 20.0 | 12500 | 0.3561 | 0.8909 | 0.9082 | 0.8995 | 0.9547 |
89
 
90
 
91
  ### Framework versions
config.json CHANGED
@@ -9,13 +9,13 @@
9
  "hidden_dropout_prob": 0.1,
10
  "hidden_size": 768,
11
  "id2label": {
12
- "0": "LABEL_0",
13
- "1": "LABEL_1",
14
- "2": "LABEL_2",
15
- "3": "LABEL_3",
16
- "4": "LABEL_4",
17
- "5": "LABEL_5",
18
- "6": "LABEL_6"
19
  },
20
  "initializer_range": 0.02,
21
  "intermediate_size": 3072,
 
9
  "hidden_dropout_prob": 0.1,
10
  "hidden_size": 768,
11
  "id2label": {
12
+ "0": "O",
13
+ "1": "B-per",
14
+ "2": "I-per",
15
+ "3": "B-org",
16
+ "4": "I-org",
17
+ "5": "B-loc",
18
+ "6": "I-loc"
19
  },
20
  "initializer_range": 0.02,
21
  "intermediate_size": 3072,
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8c32dc26e67d4fa77db3bc506b926488345693fe3c58fa1540780fc65fddbe82
3
  size 430983665
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b79cffddee930f8c3389f33b87bee0d41808c6ce898b45bc8ef2c10596480d5a
3
  size 430983665
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:75f1b23b7155993c980bae67c150064f0ae25e706d1369893636b31f3daf6136
3
  size 2671
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c159f436f728689563a5600d9aaa0888ea1263b539721c8b26016f1a3c2912a2
3
  size 2671