Hugging Face's logo

henryscheible
/
xlnet-base-cased_crows_pairs_finetuned

Text Classification
Transformers
PyTorch
xlnet
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
henryscheible commited on
Commit
f9d6f5f
•
1 Parent(s): c1bf57b

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +63 -26
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
- value: 0.44370860927152317
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the crows_pairs dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 0.6984
35
- - Accuracy: 0.4437
36
 
37
  ## Model description
38
 
@@ -51,38 +51,75 @@ 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 | 1.7260 | 0.4437 |
67
- | No log | 2.0 | 20 | 0.6869 | 0.5563 |
68
- | No log | 3.0 | 30 | 1.0103 | 0.4437 |
69
- | No log | 4.0 | 40 | 0.7207 | 0.5563 |
70
- | No log | 5.0 | 50 | 0.8402 | 0.4437 |
71
- | No log | 6.0 | 60 | 0.7060 | 0.5563 |
72
- | No log | 7.0 | 70 | 0.7714 | 0.4437 |
73
- | No log | 8.0 | 80 | 0.6924 | 0.5563 |
74
- | No log | 9.0 | 90 | 0.7429 | 0.4437 |
75
- | No log | 10.0 | 100 | 0.6886 | 0.5563 |
76
- | No log | 11.0 | 110 | 0.7110 | 0.4437 |
77
- | No log | 12.0 | 120 | 0.7309 | 0.4437 |
78
- | No log | 13.0 | 130 | 0.6999 | 0.5563 |
79
- | No log | 14.0 | 140 | 0.6962 | 0.4437 |
80
- | No log | 15.0 | 150 | 0.6869 | 0.5563 |
81
- | No log | 16.0 | 160 | 0.7281 | 0.4437 |
82
- | No log | 17.0 | 170 | 0.6870 | 0.5563 |
83
- | No log | 18.0 | 180 | 0.7582 | 0.4437 |
84
- | No log | 19.0 | 190 | 0.6998 | 0.4437 |
85
- | No log | 20.0 | 200 | 0.6984 | 0.4437 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
86
 
87
 
88
  ### Framework versions
 
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
+ value: 0.5066225165562914
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 [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the crows_pairs dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.6932
35
+ - Accuracy: 0.5066
36
 
37
  ## Model description
38
 
 
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
+ - learning_rate: 0.0005
55
+ - train_batch_size: 64
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: 30
61
 
62
  ### Training results
63
 
64
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
65
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
66
+ | 0.9006 | 0.53 | 10 | 0.7493 | 0.4934 |
67
+ | 0.7565 | 1.05 | 20 | 0.7202 | 0.4934 |
68
+ | 0.7303 | 1.58 | 30 | 0.6968 | 0.4934 |
69
+ | 0.7495 | 2.11 | 40 | 0.7210 | 0.5066 |
70
+ | 0.8008 | 2.63 | 50 | 0.6944 | 0.5066 |
71
+ | 0.7251 | 3.16 | 60 | 0.6982 | 0.5066 |
72
+ | 0.7193 | 3.68 | 70 | 0.7032 | 0.5066 |
73
+ | 0.7118 | 4.21 | 80 | 0.6975 | 0.5066 |
74
+ | 0.7419 | 4.74 | 90 | 0.7311 | 0.5066 |
75
+ | 0.7175 | 5.26 | 100 | 0.6946 | 0.5066 |
76
+ | 0.7293 | 5.79 | 110 | 0.7008 | 0.4934 |
77
+ | 0.7208 | 6.32 | 120 | 0.6940 | 0.4934 |
78
+ | 0.7101 | 6.84 | 130 | 0.6975 | 0.5066 |
79
+ | 0.7138 | 7.37 | 140 | 0.7065 | 0.4934 |
80
+ | 0.7112 | 7.89 | 150 | 0.6931 | 0.5066 |
81
+ | 0.7093 | 8.42 | 160 | 0.6931 | 0.5066 |
82
+ | 0.6996 | 8.95 | 170 | 0.6931 | 0.5066 |
83
+ | 0.6948 | 9.47 | 180 | 0.7050 | 0.4934 |
84
+ | 0.7118 | 10.0 | 190 | 0.6935 | 0.4934 |
85
+ | 0.7015 | 10.53 | 200 | 0.6993 | 0.5066 |
86
+ | 0.6985 | 11.05 | 210 | 0.6941 | 0.4934 |
87
+ | 0.6983 | 11.58 | 220 | 0.7118 | 0.4934 |
88
+ | 0.7031 | 12.11 | 230 | 0.7110 | 0.5066 |
89
+ | 0.6987 | 12.63 | 240 | 0.7643 | 0.4934 |
90
+ | 0.7483 | 13.16 | 250 | 0.7019 | 0.5066 |
91
+ | 0.7065 | 13.68 | 260 | 0.7018 | 0.4934 |
92
+ | 0.7008 | 14.21 | 270 | 0.6931 | 0.5066 |
93
+ | 0.7074 | 14.74 | 280 | 0.6932 | 0.4934 |
94
+ | 0.7097 | 15.26 | 290 | 0.6931 | 0.5066 |
95
+ | 0.7284 | 15.79 | 300 | 0.6956 | 0.4934 |
96
+ | 0.7045 | 16.32 | 310 | 0.6948 | 0.5066 |
97
+ | 0.7041 | 16.84 | 320 | 0.7176 | 0.4934 |
98
+ | 0.7118 | 17.37 | 330 | 0.6941 | 0.5066 |
99
+ | 0.7044 | 17.89 | 340 | 0.6931 | 0.5066 |
100
+ | 0.7034 | 18.42 | 350 | 0.6938 | 0.4934 |
101
+ | 0.683 | 18.95 | 360 | 0.6984 | 0.4934 |
102
+ | 0.7024 | 19.47 | 370 | 0.7009 | 0.4934 |
103
+ | 0.6988 | 20.0 | 380 | 0.6999 | 0.5066 |
104
+ | 0.6977 | 20.53 | 390 | 0.6974 | 0.4934 |
105
+ | 0.709 | 21.05 | 400 | 0.6932 | 0.5066 |
106
+ | 0.6991 | 21.58 | 410 | 0.6940 | 0.4934 |
107
+ | 0.7058 | 22.11 | 420 | 0.6931 | 0.5066 |
108
+ | 0.7101 | 22.63 | 430 | 0.6934 | 0.4934 |
109
+ | 0.7086 | 23.16 | 440 | 0.6956 | 0.4934 |
110
+ | 0.6973 | 23.68 | 450 | 0.6970 | 0.5066 |
111
+ | 0.7059 | 24.21 | 460 | 0.6931 | 0.5066 |
112
+ | 0.7021 | 24.74 | 470 | 0.6988 | 0.4934 |
113
+ | 0.6996 | 25.26 | 480 | 0.7006 | 0.4934 |
114
+ | 0.6963 | 25.79 | 490 | 0.6931 | 0.5066 |
115
+ | 0.6962 | 26.32 | 500 | 0.6932 | 0.5066 |
116
+ | 0.691 | 26.84 | 510 | 0.6944 | 0.4934 |
117
+ | 0.7003 | 27.37 | 520 | 0.6933 | 0.4934 |
118
+ | 0.6944 | 27.89 | 530 | 0.6934 | 0.4934 |
119
+ | 0.6988 | 28.42 | 540 | 0.6931 | 0.5066 |
120
+ | 0.7009 | 28.95 | 550 | 0.6931 | 0.5066 |
121
+ | 0.699 | 29.47 | 560 | 0.6933 | 0.5066 |
122
+ | 0.696 | 30.0 | 570 | 0.6932 | 0.5066 |
123
 
124
 
125
  ### Framework versions