Hugging Face's logo

Climate-TwitterBERT
/
Climate-TwitterBERT-step2

Text Classification
Transformers
PyTorch
bert
Generated from Trainer
Inference Endpoints
Model card Files Community
1
Climate-TwitterBERT commited on
Commit
9ec2681
·
1 Parent(s): b6634d0

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +17 -44
README.md CHANGED
@@ -19,12 +19,12 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  This model is a fine-tuned version of [digitalepidemiologylab/covid-twitter-bert-v2](https://huggingface.co/digitalepidemiologylab/covid-twitter-bert-v2) on an unknown dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 0.5380
23
- - Accuracy: 0.772
24
- - Precision: 0.0
25
- - Recall: 0.0
26
- - F1-weighted: 0.6727
27
- - F1: 0.0
28
 
29
  ## Model description
30
 
@@ -43,8 +43,8 @@ More information needed
43
  ### Training hyperparameters
44
 
45
  The following hyperparameters were used during training:
46
- - learning_rate: 0.001
47
- - train_batch_size: 4
48
  - eval_batch_size: 2
49
  - seed: 42
50
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
@@ -56,46 +56,19 @@ The following hyperparameters were used during training:
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1-weighted | F1 |
58
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:-----------:|:------:|
59
- | 0.6568 | 0.11 | 50 | 0.5513 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
60
- | 0.7992 | 0.23 | 100 | 0.5829 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
61
- | 1.0537 | 0.34 | 150 | 0.8565 | 0.228 | 0.228 | 1.0 | 0.0847 | 0.3713 |
62
- | 0.874 | 0.46 | 200 | 0.9819 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
63
- | 0.8665 | 0.57 | 250 | 0.6095 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
64
- | 0.6528 | 0.68 | 300 | 1.2397 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
65
- | 0.9204 | 0.8 | 350 | 0.6173 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
66
- | 0.6515 | 0.91 | 400 | 0.5916 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
67
- | 0.6148 | 1.03 | 450 | 0.6140 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
68
- | 0.7454 | 1.14 | 500 | 0.7652 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
69
- | 0.6223 | 1.26 | 550 | 0.5375 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
70
- | 0.6466 | 1.37 | 600 | 1.0611 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
71
- | 0.7892 | 1.48 | 650 | 0.7350 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
72
- | 0.602 | 1.6 | 700 | 0.7024 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
73
- | 0.8074 | 1.71 | 750 | 1.0565 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
74
- | 0.8569 | 1.83 | 800 | 0.6465 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
75
- | 0.6455 | 1.94 | 850 | 0.5401 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
76
- | 0.6624 | 2.05 | 900 | 0.6494 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
77
- | 0.5909 | 2.17 | 950 | 0.5444 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
78
- | 0.5943 | 2.28 | 1000 | 0.7197 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
79
- | 0.7885 | 2.4 | 1050 | 0.5382 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
80
- | 0.6684 | 2.51 | 1100 | 0.5417 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
81
- | 0.5723 | 2.63 | 1150 | 0.9149 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
82
- | 0.747 | 2.74 | 1200 | 0.6202 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
83
- | 0.6384 | 2.85 | 1250 | 0.5374 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
84
- | 0.634 | 2.97 | 1300 | 0.6424 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
85
- | 0.6256 | 3.08 | 1350 | 0.7462 | 0.228 | 0.228 | 1.0 | 0.0847 | 0.3713 |
86
- | 0.6282 | 3.2 | 1400 | 0.5424 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
87
- | 0.5496 | 3.31 | 1450 | 0.5402 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
88
- | 0.5245 | 3.42 | 1500 | 0.5399 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
89
- | 0.5226 | 3.54 | 1550 | 0.5699 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
90
- | 0.6251 | 3.65 | 1600 | 0.5496 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
91
- | 0.5547 | 3.77 | 1650 | 0.5435 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
92
- | 0.532 | 3.88 | 1700 | 0.5381 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
93
- | 0.6029 | 4.0 | 1750 | 0.5380 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
94
 
95
 
96
  ### Framework versions
97
 
98
  - Transformers 4.28.1
99
  - Pytorch 2.1.0+cu121
100
- - Datasets 2.16.0
101
  - Tokenizers 0.13.3
 
19
 
20
  This model is a fine-tuned version of [digitalepidemiologylab/covid-twitter-bert-v2](https://huggingface.co/digitalepidemiologylab/covid-twitter-bert-v2) on an unknown dataset.
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 0.3965
23
+ - Accuracy: 0.808
24
+ - Precision: 0.5584
25
+ - Recall: 0.7544
26
+ - F1-weighted: 0.8171
27
+ - F1: 0.6418
28
 
29
  ## Model description
30
 
 
43
  ### Training hyperparameters
44
 
45
  The following hyperparameters were used during training:
46
+ - learning_rate: 0.0001
47
+ - train_batch_size: 16
48
  - eval_batch_size: 2
49
  - seed: 42
50
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1-weighted | F1 |
58
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:-----------:|:------:|
59
+ | 0.5425 | 0.45 | 50 | 0.5093 | 0.772 | 0.0 | 0.0 | 0.6727 | 0.0 |
60
+ | 0.5154 | 0.91 | 100 | 0.5244 | 0.796 | 0.875 | 0.1228 | 0.7306 | 0.2154 |
61
+ | 0.5635 | 1.36 | 150 | 0.4300 | 0.848 | 0.8065 | 0.4386 | 0.8303 | 0.5682 |
62
+ | 0.4215 | 1.82 | 200 | 0.3679 | 0.844 | 0.625 | 0.7895 | 0.8499 | 0.6977 |
63
+ | 0.3491 | 2.27 | 250 | 0.3679 | 0.864 | 0.7170 | 0.6667 | 0.8622 | 0.6909 |
64
+ | 0.4113 | 2.73 | 300 | 0.3658 | 0.848 | 0.6418 | 0.7544 | 0.8521 | 0.6935 |
65
+ | 0.398 | 3.18 | 350 | 0.3998 | 0.78 | 0.5119 | 0.7544 | 0.7928 | 0.6099 |
66
+ | 0.315 | 3.64 | 400 | 0.3965 | 0.808 | 0.5584 | 0.7544 | 0.8171 | 0.6418 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67
 
68
 
69
  ### Framework versions
70
 
71
  - Transformers 4.28.1
72
  - Pytorch 2.1.0+cu121
73
+ - Datasets 2.16.1
74
  - Tokenizers 0.13.3