easwar03 commited on
Commit
1ad38bb
1 Parent(s): fcbab06

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +108 -0
README.md ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-sa-4.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - accuracy
7
+ - precision
8
+ - recall
9
+ - f1
10
+ model-index:
11
+ - name: legal-bert-small
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # legal-bert-small
19
+
20
+ This model is a fine-tuned version of [nlpaueb/legal-bert-small-uncased](https://huggingface.co/nlpaueb/legal-bert-small-uncased) on the None dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.1837
23
+ - Accuracy: 0.9548
24
+ - Precision: 0.7491
25
+ - Recall: 0.7882
26
+ - F1: 0.7682
27
+ - Classification Report: precision recall f1-score support
28
+
29
+ LOC 0.84 0.86 0.85 1668
30
+ MISC 0.59 0.63 0.61 702
31
+ ORG 0.64 0.67 0.66 1661
32
+ PER 0.83 0.90 0.87 1617
33
+
34
+ micro avg 0.75 0.79 0.77 5648
35
+ macro avg 0.73 0.77 0.75 5648
36
+ weighted avg 0.75 0.79 0.77 5648
37
+
38
+
39
+ ## Model description
40
+
41
+ More information needed
42
+
43
+ ## Intended uses & limitations
44
+
45
+ More information needed
46
+
47
+ ## Training and evaluation data
48
+
49
+ More information needed
50
+
51
+ ## Training procedure
52
+
53
+ ### Training hyperparameters
54
+
55
+ The following hyperparameters were used during training:
56
+ - learning_rate: 2e-05
57
+ - train_batch_size: 16
58
+ - eval_batch_size: 32
59
+ - seed: 42
60
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
+ - lr_scheduler_type: linear
62
+ - num_epochs: 3
63
+
64
+ ### Training results
65
+
66
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Classification Report |
67
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|
68
+ | 0.1443 | 1.0 | 434 | 0.1949 | 0.9462 | 0.6982 | 0.7466 | 0.7216 | precision recall f1-score support
69
+
70
+ LOC 0.83 0.80 0.81 1668
71
+ MISC 0.59 0.62 0.60 702
72
+ ORG 0.56 0.58 0.57 1661
73
+ PER 0.75 0.92 0.83 1617
74
+
75
+ micro avg 0.70 0.75 0.72 5648
76
+ macro avg 0.68 0.73 0.70 5648
77
+ weighted avg 0.70 0.75 0.72 5648
78
+ |
79
+ | 0.071 | 2.0 | 868 | 0.1764 | 0.9551 | 0.7548 | 0.7702 | 0.7624 | precision recall f1-score support
80
+
81
+ LOC 0.81 0.88 0.84 1668
82
+ MISC 0.59 0.64 0.61 702
83
+ ORG 0.68 0.59 0.63 1661
84
+ PER 0.83 0.90 0.86 1617
85
+
86
+ micro avg 0.75 0.77 0.76 5648
87
+ macro avg 0.73 0.75 0.74 5648
88
+ weighted avg 0.75 0.77 0.76 5648
89
+ |
90
+ | 0.0713 | 3.0 | 1302 | 0.1837 | 0.9548 | 0.7491 | 0.7882 | 0.7682 | precision recall f1-score support
91
+
92
+ LOC 0.84 0.86 0.85 1668
93
+ MISC 0.59 0.63 0.61 702
94
+ ORG 0.64 0.67 0.66 1661
95
+ PER 0.83 0.90 0.87 1617
96
+
97
+ micro avg 0.75 0.79 0.77 5648
98
+ macro avg 0.73 0.77 0.75 5648
99
+ weighted avg 0.75 0.79 0.77 5648
100
+ |
101
+
102
+
103
+ ### Framework versions
104
+
105
+ - Transformers 4.30.2
106
+ - Pytorch 2.0.0
107
+ - Datasets 2.1.0
108
+ - Tokenizers 0.13.3