Hugging Face's logo

ToToKr
/
kobigbird-bert-base-finetuned-klue

Question Answering
Transformers
PyTorch
TensorBoard
big_bird
generated_from_trainer
Inference Endpoints
Model card Files Metrics Community
ToToKr commited on
Commit
518fbcf
•
1 Parent(s): 5667baf

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +34 -46
README.md CHANGED
@@ -13,7 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
13
 
14
  This model is a fine-tuned version of [monologg/kobigbird-bert-base](https://huggingface.co/monologg/kobigbird-bert-base) on the None dataset.
15
  It achieves the following results on the evaluation set:
16
- - Loss: 1.9124
17
 
18
  ## Model description
19
 
@@ -33,61 +33,49 @@ More information needed
33
 
34
  The following hyperparameters were used during training:
35
  - learning_rate: 5e-05
36
- - train_batch_size: 32
37
- - eval_batch_size: 32
38
  - seed: 42
39
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
40
  - lr_scheduler_type: linear
41
- - num_epochs: 60
42
 
43
  ### Training results
44
 
45
  | Training Loss | Epoch | Step | Validation Loss |
46
  |:-------------:|:-----:|:-----:|:---------------:|
47
- | 5.1951 | 0.69 | 500 | 3.9518 |
48
- | 3.7508 | 1.38 | 1000 | 3.4423 |
49
- | 3.4379 | 2.07 | 1500 | 3.1397 |
50
- | 3.1865 | 2.76 | 2000 | 2.8141 |
51
- | 2.9852 | 3.45 | 2500 | 2.6055 |
52
- | 2.7907 | 4.14 | 3000 | 2.4892 |
53
- | 2.6985 | 4.83 | 3500 | 2.4130 |
54
- | 2.63 | 5.52 | 4000 | 2.3479 |
55
- | 2.5616 | 6.21 | 4500 | 2.3134 |
56
- | 2.4898 | 6.9 | 5000 | 2.2526 |
57
- | 2.4167 | 7.59 | 5500 | 2.2422 |
58
- | 2.3861 | 8.28 | 6000 | 2.2138 |
59
- | 2.3401 | 8.97 | 6500 | 2.1762 |
60
- | 2.292 | 9.66 | 7000 | 2.1417 |
61
- | 2.2705 | 10.34 | 7500 | 2.1148 |
62
- | 2.2139 | 11.03 | 8000 | 2.1357 |
63
- | 2.187 | 11.72 | 8500 | 2.0995 |
64
- | 2.1518 | 12.41 | 9000 | 2.0554 |
65
- | 2.1179 | 13.1 | 9500 | 2.0467 |
66
- | 2.0781 | 13.79 | 10000 | 2.0418 |
67
- | 2.0457 | 14.48 | 10500 | 2.0057 |
68
- | 2.0326 | 15.17 | 11000 | 2.0022 |
69
- | 1.9976 | 15.86 | 11500 | 2.0137 |
70
- | 1.9602 | 16.55 | 12000 | 1.9703 |
71
- | 1.9435 | 17.24 | 12500 | 1.9920 |
72
- | 1.9092 | 17.93 | 13000 | 1.9504 |
73
- | 1.8714 | 18.62 | 13500 | 1.9731 |
74
- | 1.8715 | 19.31 | 14000 | 1.9471 |
75
- | 1.8824 | 20.0 | 14500 | 1.9462 |
76
- | 1.8261 | 20.69 | 15000 | 1.9412 |
77
- | 1.815 | 21.38 | 15500 | 1.9308 |
78
- | 1.7988 | 22.07 | 16000 | 1.9285 |
79
- | 1.7722 | 22.76 | 16500 | 1.9341 |
80
- | 1.7606 | 23.45 | 17000 | 1.9007 |
81
- | 1.7402 | 24.14 | 17500 | 1.9206 |
82
- | 1.7177 | 24.83 | 18000 | 1.9125 |
83
- | 1.7117 | 25.52 | 18500 | 1.9178 |
84
- | 1.6834 | 26.21 | 19000 | 1.9082 |
85
- | 1.6893 | 26.9 | 19500 | 1.9124 |
86
 
87
 
88
  ### Framework versions
89
 
90
- - Transformers 4.18.0
91
- - Pytorch 1.11.0.dev20211231+cu113
92
- - Datasets 2.1.0
93
  - Tokenizers 0.12.1
13
 
14
  This model is a fine-tuned version of [monologg/kobigbird-bert-base](https://huggingface.co/monologg/kobigbird-bert-base) on the None dataset.
15
  It achieves the following results on the evaluation set:
16
+ - Loss: 0.8347
17
 
18
  ## Model description
19
 
33
 
34
  The following hyperparameters were used during training:
35
  - learning_rate: 5e-05
36
+ - train_batch_size: 16
37
+ - eval_batch_size: 16
38
  - seed: 42
39
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
40
  - lr_scheduler_type: linear
41
+ - num_epochs: 20
42
 
43
  ### Training results
44
 
45
  | Training Loss | Epoch | Step | Validation Loss |
46
  |:-------------:|:-----:|:-----:|:---------------:|
47
+ | 5.3957 | 0.13 | 500 | 3.7603 |
48
+ | 3.2242 | 0.26 | 1000 | 2.3961 |
49
+ | 2.0812 | 0.4 | 1500 | 1.5552 |
50
+ | 1.6198 | 0.53 | 2000 | 1.3609 |
51
+ | 1.447 | 0.66 | 2500 | 1.2270 |
52
+ | 1.3438 | 0.79 | 3000 | 1.1321 |
53
+ | 1.2399 | 0.93 | 3500 | 1.0973 |
54
+ | 1.1976 | 1.06 | 4000 | 1.0418 |
55
+ | 1.1177 | 1.19 | 4500 | 1.0301 |
56
+ | 1.0811 | 1.32 | 5000 | 1.0232 |
57
+ | 1.0506 | 1.45 | 5500 | 0.9971 |
58
+ | 1.0293 | 1.59 | 6000 | 0.9580 |
59
+ | 1.0196 | 1.72 | 6500 | 0.9551 |
60
+ | 0.9846 | 1.85 | 7000 | 0.9274 |
61
+ | 0.9702 | 1.98 | 7500 | 0.9286 |
62
+ | 0.9224 | 2.11 | 8000 | 0.8961 |
63
+ | 0.8867 | 2.25 | 8500 | 0.9193 |
64
+ | 0.8711 | 2.38 | 9000 | 0.8727 |
65
+ | 0.883 | 2.51 | 9500 | 0.8790 |
66
+ | 0.8513 | 2.64 | 10000 | 0.8830 |
67
+ | 0.8709 | 2.78 | 10500 | 0.8604 |
68
+ | 0.8766 | 2.91 | 11000 | 0.8260 |
69
+ | 0.7976 | 3.04 | 11500 | 0.8401 |
70
+ | 0.7724 | 3.17 | 12000 | 0.8617 |
71
+ | 0.78 | 3.3 | 12500 | 0.8601 |
72
+ | 0.7566 | 3.44 | 13000 | 0.8657 |
73
+ | 0.7407 | 3.57 | 13500 | 0.8347 |
 
 
 
 
 
 
 
 
 
 
 
 
74
 
75
 
76
  ### Framework versions
77
 
78
+ - Transformers 4.19.2
79
+ - Pytorch 1.11.0+cu113
80
+ - Datasets 2.2.2
81
  - Tokenizers 0.12.1