James-kc-min commited on
Commit
c91574e
1 Parent(s): 7a6f838

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +120 -0
README.md ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - accuracy
7
+ - f1
8
+ - precision
9
+ - recall
10
+ model-index:
11
+ - name: L_Roberta3
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
+ # L_Roberta3
19
+
20
+ This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.2095
23
+ - Accuracy: 0.9555
24
+ - F1: 0.9555
25
+ - Precision: 0.9555
26
+ - Recall: 0.9555
27
+ - C Report: precision recall f1-score support
28
+
29
+ 0 0.97 0.95 0.96 876
30
+ 1 0.94 0.97 0.95 696
31
+
32
+ accuracy 0.96 1572
33
+ macro avg 0.95 0.96 0.96 1572
34
+ weighted avg 0.96 0.96 0.96 1572
35
+
36
+ - C Matrix: None
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 5e-05
56
+ - train_batch_size: 32
57
+ - eval_batch_size: 32
58
+ - seed: 42
59
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
+ - lr_scheduler_type: linear
61
+ - num_epochs: 5
62
+ - mixed_precision_training: Native AMP
63
+
64
+ ### Training results
65
+
66
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | C Report | C Matrix |
67
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|:--------:|
68
+ | 0.2674 | 1.0 | 329 | 0.2436 | 0.9389 | 0.9389 | 0.9389 | 0.9389 | precision recall f1-score support
69
+
70
+ 0 0.94 0.95 0.95 876
71
+ 1 0.94 0.92 0.93 696
72
+
73
+ accuracy 0.94 1572
74
+ macro avg 0.94 0.94 0.94 1572
75
+ weighted avg 0.94 0.94 0.94 1572
76
+ | None |
77
+ | 0.1377 | 2.0 | 658 | 0.1506 | 0.9408 | 0.9408 | 0.9408 | 0.9408 | precision recall f1-score support
78
+
79
+ 0 0.97 0.92 0.95 876
80
+ 1 0.91 0.96 0.94 696
81
+
82
+ accuracy 0.94 1572
83
+ macro avg 0.94 0.94 0.94 1572
84
+ weighted avg 0.94 0.94 0.94 1572
85
+ | None |
86
+ | 0.0898 | 3.0 | 987 | 0.1491 | 0.9548 | 0.9548 | 0.9548 | 0.9548 | precision recall f1-score support
87
+
88
+ 0 0.96 0.96 0.96 876
89
+ 1 0.95 0.95 0.95 696
90
+
91
+ accuracy 0.95 1572
92
+ macro avg 0.95 0.95 0.95 1572
93
+ weighted avg 0.95 0.95 0.95 1572
94
+ | None |
95
+ | 0.0543 | 4.0 | 1316 | 0.1831 | 0.9561 | 0.9561 | 0.9561 | 0.9561 | precision recall f1-score support
96
+
97
+ 0 0.97 0.95 0.96 876
98
+ 1 0.94 0.96 0.95 696
99
+
100
+ accuracy 0.96 1572
101
+ macro avg 0.95 0.96 0.96 1572
102
+ weighted avg 0.96 0.96 0.96 1572
103
+ | None |
104
+ | 0.0394 | 5.0 | 1645 | 0.2095 | 0.9555 | 0.9555 | 0.9555 | 0.9555 | precision recall f1-score support
105
+
106
+ 0 0.97 0.95 0.96 876
107
+ 1 0.94 0.97 0.95 696
108
+
109
+ accuracy 0.96 1572
110
+ macro avg 0.95 0.96 0.96 1572
111
+ weighted avg 0.96 0.96 0.96 1572
112
+ | None |
113
+
114
+
115
+ ### Framework versions
116
+
117
+ - Transformers 4.18.0
118
+ - Pytorch 1.10.2+cu102
119
+ - Datasets 2.2.2
120
+ - Tokenizers 0.12.1