fffffly commited on
Commit
0899fd3
1 Parent(s): 2ad06a8

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -25
README.md CHANGED
@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 0.3741
21
- - Accuracy: 0.8411
22
- - F1: 0.8015
23
 
24
  ## Model description
25
 
@@ -38,38 +38,33 @@ More information needed
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
- - learning_rate: 1e-06
42
  - train_batch_size: 16
43
  - eval_batch_size: 16
44
  - seed: 42
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
- - num_epochs: 20
48
 
49
  ### Training results
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
52
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
53
- | No log | 1.0 | 167 | 0.6611 | 0.6612 | 0.3722 |
54
- | No log | 2.0 | 334 | 0.5848 | 0.7421 | 0.6325 |
55
- | 0.6188 | 3.0 | 501 | 0.5176 | 0.7601 | 0.7080 |
56
- | 0.6188 | 4.0 | 668 | 0.4783 | 0.7706 | 0.7193 |
57
- | 0.6188 | 5.0 | 835 | 0.4566 | 0.7841 | 0.7273 |
58
- | 0.455 | 6.0 | 1002 | 0.4415 | 0.7946 | 0.7360 |
59
- | 0.455 | 7.0 | 1169 | 0.4250 | 0.8096 | 0.7703 |
60
- | 0.455 | 8.0 | 1336 | 0.4146 | 0.8186 | 0.7713 |
61
- | 0.3881 | 9.0 | 1503 | 0.4069 | 0.8261 | 0.7803 |
62
- | 0.3881 | 10.0 | 1670 | 0.3968 | 0.8321 | 0.7926 |
63
- | 0.3881 | 11.0 | 1837 | 0.3916 | 0.8351 | 0.7948 |
64
- | 0.3493 | 12.0 | 2004 | 0.3865 | 0.8426 | 0.8037 |
65
- | 0.3493 | 13.0 | 2171 | 0.3822 | 0.8426 | 0.8045 |
66
- | 0.3493 | 14.0 | 2338 | 0.3802 | 0.8456 | 0.8068 |
67
- | 0.3239 | 15.0 | 2505 | 0.3769 | 0.8471 | 0.8111 |
68
- | 0.3239 | 16.0 | 2672 | 0.3756 | 0.8441 | 0.8060 |
69
- | 0.3239 | 17.0 | 2839 | 0.3749 | 0.8411 | 0.8015 |
70
- | 0.3104 | 18.0 | 3006 | 0.3742 | 0.8396 | 0.8000 |
71
- | 0.3104 | 19.0 | 3173 | 0.3738 | 0.8396 | 0.8000 |
72
- | 0.3104 | 20.0 | 3340 | 0.3741 | 0.8411 | 0.8015 |
73
 
74
 
75
  ### Framework versions
 
17
 
18
  This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 1.0970
21
+ - Accuracy: 0.8681
22
+ - F1: 0.8376
23
 
24
  ## Model description
25
 
 
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
+ - learning_rate: 2e-05
42
  - train_batch_size: 16
43
  - eval_batch_size: 16
44
  - seed: 42
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
+ - num_epochs: 15
48
 
49
  ### Training results
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
52
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
53
+ | No log | 1.0 | 167 | 0.3828 | 0.8501 | 0.8031 |
54
+ | No log | 2.0 | 334 | 0.4787 | 0.8456 | 0.8275 |
55
+ | 0.2101 | 3.0 | 501 | 0.6186 | 0.8666 | 0.8367 |
56
+ | 0.2101 | 4.0 | 668 | 0.7201 | 0.8546 | 0.8265 |
57
+ | 0.2101 | 5.0 | 835 | 0.7675 | 0.8651 | 0.8346 |
58
+ | 0.0339 | 6.0 | 1002 | 0.8561 | 0.8681 | 0.8434 |
59
+ | 0.0339 | 7.0 | 1169 | 0.8898 | 0.8681 | 0.8382 |
60
+ | 0.0339 | 8.0 | 1336 | 0.9854 | 0.8711 | 0.8436 |
61
+ | 0.0069 | 9.0 | 1503 | 0.9919 | 0.8711 | 0.8407 |
62
+ | 0.0069 | 10.0 | 1670 | 1.0695 | 0.8561 | 0.8280 |
63
+ | 0.0069 | 11.0 | 1837 | 1.0542 | 0.8666 | 0.8349 |
64
+ | 0.0007 | 12.0 | 2004 | 1.0896 | 0.8681 | 0.8370 |
65
+ | 0.0007 | 13.0 | 2171 | 1.1001 | 0.8666 | 0.8349 |
66
+ | 0.0007 | 14.0 | 2338 | 1.0888 | 0.8606 | 0.8312 |
67
+ | 0.0012 | 15.0 | 2505 | 1.0970 | 0.8681 | 0.8376 |
 
 
 
 
 
68
 
69
 
70
  ### Framework versions