Hugging Face's logo

tiennvcs
/
bert-base-uncased-finetuned-vi-infovqa

Question Answering
Transformers
PyTorch
TensorBoard
bert
generated_from_trainer
Inference Endpoints
Model card Files Metrics Community
1
tiennvcs commited on
Commit
2fa675b
•
1 Parent(s): d8859ff

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -53
README.md CHANGED
@@ -14,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
14
 
15
  This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
- - Loss: 13.9895
18
 
19
  ## Model description
20
 
@@ -33,69 +33,32 @@ More information needed
33
  ### Training hyperparameters
34
 
35
  The following hyperparameters were used during training:
36
- - learning_rate: 0.0001
37
  - train_batch_size: 4
38
  - eval_batch_size: 4
39
  - seed: 250500
40
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
41
  - lr_scheduler_type: linear
42
- - num_epochs: 50
43
 
44
  ### Training results
45
 
46
- | Training Loss | Epoch | Step | Validation Loss |
47
- |:-------------:|:-----:|:-----:|:---------------:|
48
- | 4.3635 | 1.07 | 500 | 4.5286 |
49
- | 3.8961 | 2.13 | 1000 | 4.4055 |
50
- | 3.3012 | 3.2 | 1500 | 5.0429 |
51
- | 2.5998 | 4.26 | 2000 | 5.4774 |
52
- | 1.8851 | 5.33 | 2500 | 6.4376 |
53
- | 1.3968 | 6.4 | 3000 | 6.8778 |
54
- | 1.1244 | 7.46 | 3500 | 7.1008 |
55
- | 0.8851 | 8.53 | 4000 | 6.8303 |
56
- | 0.7654 | 9.59 | 4500 | 8.6628 |
57
- | 0.6895 | 10.66 | 5000 | 6.0228 |
58
- | 0.5989 | 11.73 | 5500 | 8.8191 |
59
- | 0.6033 | 12.79 | 6000 | 9.7359 |
60
- | 0.6005 | 13.86 | 6500 | 7.6668 |
61
- | 0.522 | 14.93 | 7000 | 8.7185 |
62
- | 0.464 | 15.99 | 7500 | 10.1035 |
63
- | 0.4112 | 17.06 | 8000 | 8.7928 |
64
- | 0.3501 | 18.12 | 8500 | 9.9157 |
65
- | 0.3401 | 19.19 | 9000 | 12.1013 |
66
- | 0.3233 | 20.26 | 9500 | 10.2730 |
67
- | 0.2513 | 21.32 | 10000 | 8.4839 |
68
- | 0.2319 | 22.39 | 10500 | 10.9367 |
69
- | 0.2269 | 23.45 | 11000 | 9.6821 |
70
- | 0.237 | 24.52 | 11500 | 10.2357 |
71
- | 0.1868 | 25.59 | 12000 | 9.8762 |
72
- | 0.1655 | 26.65 | 12500 | 10.7398 |
73
- | 0.1561 | 27.72 | 13000 | 11.8157 |
74
- | 0.1714 | 28.78 | 13500 | 10.8686 |
75
- | 0.1098 | 29.85 | 14000 | 13.1537 |
76
- | 0.1222 | 30.92 | 14500 | 14.5398 |
77
- | 0.1325 | 31.98 | 15000 | 12.6095 |
78
- | 0.1088 | 33.05 | 15500 | 12.0747 |
79
- | 0.0855 | 34.12 | 16000 | 12.4450 |
80
- | 0.0832 | 35.18 | 16500 | 12.9436 |
81
- | 0.0687 | 36.25 | 17000 | 12.5902 |
82
- | 0.0804 | 37.31 | 17500 | 11.9873 |
83
- | 0.0427 | 38.38 | 18000 | 12.6357 |
84
- | 0.0529 | 39.45 | 18500 | 11.5851 |
85
- | 0.0478 | 40.51 | 19000 | 13.1796 |
86
- | 0.0513 | 41.58 | 19500 | 13.8597 |
87
- | 0.0449 | 42.64 | 20000 | 13.6467 |
88
- | 0.0327 | 43.71 | 20500 | 13.0391 |
89
- | 0.0329 | 44.78 | 21000 | 13.1984 |
90
- | 0.0266 | 45.84 | 21500 | 13.9690 |
91
- | 0.0301 | 46.91 | 22000 | 13.9516 |
92
- | 0.0193 | 47.97 | 22500 | 13.9104 |
93
- | 0.0273 | 49.04 | 23000 | 13.9895 |
94
 
95
 
96
  ### Framework versions
97
 
98
- - Transformers 4.14.1
99
  - Pytorch 1.10.0+cu111
100
- - Datasets 1.16.1
101
  - Tokenizers 0.10.3
14
 
15
  This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Loss: 3.5470
18
 
19
  ## Model description
20
 
33
  ### Training hyperparameters
34
 
35
  The following hyperparameters were used during training:
36
+ - learning_rate: 5e-05
37
  - train_batch_size: 4
38
  - eval_batch_size: 4
39
  - seed: 250500
40
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
41
  - lr_scheduler_type: linear
42
+ - num_epochs: 2
43
 
44
  ### Training results
45
 
46
+ | Training Loss | Epoch | Step | Validation Loss |
47
+ |:-------------:|:-----:|:----:|:---------------:|
48
+ | No log | 0.21 | 100 | 4.2058 |
49
+ | No log | 0.43 | 200 | 4.0210 |
50
+ | No log | 0.64 | 300 | 4.0454 |
51
+ | No log | 0.85 | 400 | 3.7557 |
52
+ | 4.04 | 1.07 | 500 | 3.8257 |
53
+ | 4.04 | 1.28 | 600 | 3.7713 |
54
+ | 4.04 | 1.49 | 700 | 3.6075 |
55
+ | 4.04 | 1.71 | 800 | 3.6155 |
56
+ | 4.04 | 1.92 | 900 | 3.5470 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
57
 
58
 
59
  ### Framework versions
60
 
61
+ - Transformers 4.15.0
62
  - Pytorch 1.10.0+cu111
63
+ - Datasets 1.17.0
64
  - Tokenizers 0.10.3