Hugging Face's logo

henryscheible
/
gpt2_crows_pairs_finetuned

Text Classification
Transformers
PyTorch
gpt2
generated_from_trainer
Eval Results
Inference Endpoints
text-generation-inference
Model card Files Community
1
henryscheible commited on
Commit
b2ff2c6
•
1 Parent(s): ce18bdf

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +40 -28
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
- value: 0.5033112582781457
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,12 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the crows_pairs dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 2.6133
35
- - Accuracy: 0.5033
 
 
 
 
36
 
37
  ## Model description
38
 
@@ -51,38 +55,46 @@ More information needed
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
- - learning_rate: 0.01
55
- - train_batch_size: 128
56
  - eval_batch_size: 64
57
  - seed: 42
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: linear
60
- - num_epochs: 20
61
 
62
  ### Training results
63
 
64
- | Training Loss | Epoch | Step | Validation Loss | Accuracy |
65
- |:-------------:|:-----:|:----:|:---------------:|:--------:|
66
- | No log | 1.0 | 10 | 1.2278 | 0.5033 |
67
- | No log | 2.0 | 20 | 0.7814 | 0.4967 |
68
- | No log | 3.0 | 30 | 0.6958 | 0.5033 |
69
- | No log | 4.0 | 40 | 0.6935 | 0.5066 |
70
- | No log | 5.0 | 50 | 0.6903 | 0.5099 |
71
- | No log | 6.0 | 60 | 0.6912 | 0.5166 |
72
- | No log | 7.0 | 70 | 0.7034 | 0.4967 |
73
- | No log | 8.0 | 80 | 0.7621 | 0.5397 |
74
- | No log | 9.0 | 90 | 0.7317 | 0.5232 |
75
- | No log | 10.0 | 100 | 0.8999 | 0.5331 |
76
- | No log | 11.0 | 110 | 1.1728 | 0.5 |
77
- | No log | 12.0 | 120 | 1.4566 | 0.4934 |
78
- | No log | 13.0 | 130 | 1.2100 | 0.5099 |
79
- | No log | 14.0 | 140 | 1.3960 | 0.5 |
80
- | No log | 15.0 | 150 | 1.7366 | 0.4934 |
81
- | No log | 16.0 | 160 | 2.1969 | 0.5331 |
82
- | No log | 17.0 | 170 | 2.5608 | 0.5132 |
83
- | No log | 18.0 | 180 | 2.7295 | 0.5530 |
84
- | No log | 19.0 | 190 | 2.7093 | 0.5232 |
85
- | No log | 20.0 | 200 | 2.6133 | 0.5033 |
 
 
 
 
 
 
 
 
86
 
87
 
88
  ### Framework versions
 
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
+ value: 0.6423841059602649
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the crows_pairs dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 3.4639
35
+ - Accuracy: 0.6424
36
+ - Tp: 0.3344
37
+ - Tn: 0.3079
38
+ - Fp: 0.1656
39
+ - Fn: 0.1921
40
 
41
  ## Model description
42
 
 
55
  ### Training hyperparameters
56
 
57
  The following hyperparameters were used during training:
58
+ - learning_rate: 5e-05
59
+ - train_batch_size: 64
60
  - eval_batch_size: 64
61
  - seed: 42
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
+ - num_epochs: 30
65
 
66
  ### Training results
67
 
68
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy | Tp | Tn | Fp | Fn |
69
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:------:|:------:|:------:|
70
+ | 0.7226 | 1.05 | 20 | 0.7030 | 0.5265 | 0.5265 | 0.0 | 0.4735 | 0.0 |
71
+ | 0.691 | 2.11 | 40 | 0.7176 | 0.5199 | 0.4503 | 0.0695 | 0.4040 | 0.0762 |
72
+ | 0.7288 | 3.16 | 60 | 0.7409 | 0.5331 | 0.3146 | 0.2185 | 0.2550 | 0.2119 |
73
+ | 0.403 | 4.21 | 80 | 0.8119 | 0.5563 | 0.2616 | 0.2947 | 0.1788 | 0.2649 |
74
+ | 0.2029 | 5.26 | 100 | 1.2245 | 0.5960 | 0.4205 | 0.1755 | 0.2980 | 0.1060 |
75
+ | 0.1727 | 6.32 | 120 | 1.3791 | 0.5795 | 0.4437 | 0.1358 | 0.3377 | 0.0828 |
76
+ | 0.0907 | 7.37 | 140 | 1.6621 | 0.5960 | 0.3907 | 0.2053 | 0.2682 | 0.1358 |
77
+ | 0.0566 | 8.42 | 160 | 2.7341 | 0.5960 | 0.3444 | 0.2517 | 0.2219 | 0.1821 |
78
+ | 0.0616 | 9.47 | 180 | 2.4315 | 0.5861 | 0.2947 | 0.2914 | 0.1821 | 0.2318 |
79
+ | 0.0794 | 10.53 | 200 | 2.3228 | 0.5795 | 0.4371 | 0.1424 | 0.3311 | 0.0894 |
80
+ | 0.0329 | 11.58 | 220 | 2.4578 | 0.6026 | 0.2318 | 0.3709 | 0.1026 | 0.2947 |
81
+ | 0.0788 | 12.63 | 240 | 2.5637 | 0.6026 | 0.3311 | 0.2715 | 0.2020 | 0.1954 |
82
+ | 0.0164 | 13.68 | 260 | 2.4828 | 0.6159 | 0.3510 | 0.2649 | 0.2086 | 0.1755 |
83
+ | 0.02 | 14.74 | 280 | 3.1787 | 0.5894 | 0.3046 | 0.2848 | 0.1887 | 0.2219 |
84
+ | 0.0256 | 15.79 | 300 | 3.4015 | 0.6093 | 0.3212 | 0.2881 | 0.1854 | 0.2053 |
85
+ | 0.0006 | 16.84 | 320 | 3.1505 | 0.6358 | 0.3411 | 0.2947 | 0.1788 | 0.1854 |
86
+ | 0.0004 | 17.89 | 340 | 3.3157 | 0.6126 | 0.3642 | 0.2483 | 0.2252 | 0.1623 |
87
+ | 0.0224 | 18.95 | 360 | 3.2259 | 0.6258 | 0.3113 | 0.3146 | 0.1589 | 0.2152 |
88
+ | 0.0001 | 20.0 | 380 | 3.3545 | 0.6325 | 0.3576 | 0.2748 | 0.1987 | 0.1689 |
89
+ | 0.0033 | 21.05 | 400 | 3.2960 | 0.6358 | 0.3377 | 0.2980 | 0.1755 | 0.1887 |
90
+ | 0.0039 | 22.11 | 420 | 3.4235 | 0.6358 | 0.3642 | 0.2715 | 0.2020 | 0.1623 |
91
+ | 0.0001 | 23.16 | 440 | 3.3855 | 0.6325 | 0.3311 | 0.3013 | 0.1722 | 0.1954 |
92
+ | 0.0001 | 24.21 | 460 | 3.5120 | 0.6291 | 0.3709 | 0.2583 | 0.2152 | 0.1556 |
93
+ | 0.0222 | 25.26 | 480 | 3.5530 | 0.6457 | 0.3278 | 0.3179 | 0.1556 | 0.1987 |
94
+ | 0.0003 | 26.32 | 500 | 3.4705 | 0.6325 | 0.3477 | 0.2848 | 0.1887 | 0.1788 |
95
+ | 0.0013 | 27.37 | 520 | 3.5134 | 0.6424 | 0.3411 | 0.3013 | 0.1722 | 0.1854 |
96
+ | 0.017 | 28.42 | 540 | 3.4743 | 0.6424 | 0.3411 | 0.3013 | 0.1722 | 0.1854 |
97
+ | 0.0069 | 29.47 | 560 | 3.4639 | 0.6424 | 0.3344 | 0.3079 | 0.1656 | 0.1921 |
98
 
99
 
100
  ### Framework versions