Hugging Face's logo

MakAIHealthLab
/
vit-base-patch16-224-in21k-finetuned-papsmear

Image Classification
Transformers
TensorBoard
Safetensors
vit
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
wcosmas commited on
Commit
bdf8fc6
•
1 Parent(s): 5f0716b

Model save

Browse files
Files changed (2) hide show
  1. README.md +54 -37
  2. model.safetensors +1 -1
README.md CHANGED
@@ -23,7 +23,7 @@ model-index:
23
  metrics:
24
  - name: Accuracy
25
  type: accuracy
26
- value: 0.9411764705882353
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
33
 
34
  This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.2523
37
- - Accuracy: 0.9412
38
 
39
  ## Model description
40
 
@@ -54,50 +54,67 @@ More information needed
54
 
55
  The following hyperparameters were used during training:
56
  - learning_rate: 5e-05
57
- - train_batch_size: 8
58
- - eval_batch_size: 8
59
  - seed: 42
60
  - gradient_accumulation_steps: 4
61
- - total_train_batch_size: 32
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
  - lr_scheduler_warmup_ratio: 0.1
65
- - num_epochs: 30
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-------:|:----:|:---------------:|:--------:|
71
- | 1.6954 | 0.9935 | 38 | 1.6106 | 0.3456 |
72
- | 1.2818 | 1.9869 | 76 | 1.2412 | 0.5735 |
73
- | 1.0023 | 2.9804 | 114 | 0.9875 | 0.7132 |
74
- | 0.7163 | 4.0 | 153 | 0.8399 | 0.6912 |
75
- | 0.5173 | 4.9935 | 191 | 0.6546 | 0.8162 |
76
- | 0.5057 | 5.9869 | 229 | 0.6251 | 0.8309 |
77
- | 0.4313 | 6.9804 | 267 | 0.5696 | 0.8309 |
78
- | 0.325 | 8.0 | 306 | 0.5507 | 0.8309 |
79
- | 0.3811 | 8.9935 | 344 | 0.4429 | 0.8676 |
80
- | 0.2341 | 9.9869 | 382 | 0.4222 | 0.875 |
81
- | 0.3082 | 10.9804 | 420 | 0.6573 | 0.7721 |
82
- | 0.2571 | 12.0 | 459 | 0.4229 | 0.8897 |
83
- | 0.2374 | 12.9935 | 497 | 0.4233 | 0.875 |
84
- | 0.128 | 13.9869 | 535 | 0.3671 | 0.8971 |
85
- | 0.1718 | 14.9804 | 573 | 0.3430 | 0.8971 |
86
- | 0.16 | 16.0 | 612 | 0.4104 | 0.875 |
87
- | 0.1096 | 16.9935 | 650 | 0.2920 | 0.9118 |
88
- | 0.1408 | 17.9869 | 688 | 0.2630 | 0.9044 |
89
- | 0.113 | 18.9804 | 726 | 0.3084 | 0.8824 |
90
- | 0.1272 | 20.0 | 765 | 0.2523 | 0.9412 |
91
- | 0.119 | 20.9935 | 803 | 0.4254 | 0.8824 |
92
- | 0.1068 | 21.9869 | 841 | 0.3519 | 0.8971 |
93
- | 0.0723 | 22.9804 | 879 | 0.3293 | 0.9191 |
94
- | 0.0769 | 24.0 | 918 | 0.2613 | 0.9265 |
95
- | 0.095 | 24.9935 | 956 | 0.2609 | 0.9412 |
96
- | 0.0863 | 25.9869 | 994 | 0.2650 | 0.9265 |
97
- | 0.0795 | 26.9804 | 1032 | 0.2978 | 0.9118 |
98
- | 0.0564 | 28.0 | 1071 | 0.2737 | 0.9191 |
99
- | 0.0562 | 28.9935 | 1109 | 0.2941 | 0.9191 |
100
- | 0.0751 | 29.8039 | 1140 | 0.3111 | 0.9191 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
101
 
102
 
103
  ### Framework versions
 
23
  metrics:
24
  - name: Accuracy
25
  type: accuracy
26
+ value: 0.9338235294117647
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
33
 
34
  This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.2870
37
+ - Accuracy: 0.9338
38
 
39
  ## Model description
40
 
 
54
 
55
  The following hyperparameters were used during training:
56
  - learning_rate: 5e-05
57
+ - train_batch_size: 32
58
+ - eval_batch_size: 32
59
  - seed: 42
60
  - gradient_accumulation_steps: 4
61
+ - total_train_batch_size: 128
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
  - lr_scheduler_warmup_ratio: 0.1
65
+ - num_epochs: 50
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-------:|:----:|:---------------:|:--------:|
71
+ | No log | 0.9231 | 9 | 1.7879 | 0.2059 |
72
+ | 1.8037 | 1.9487 | 19 | 1.6485 | 0.4044 |
73
+ | 1.6961 | 2.9744 | 29 | 1.4882 | 0.3971 |
74
+ | 1.5407 | 4.0 | 39 | 1.3069 | 0.5221 |
75
+ | 1.3308 | 4.9231 | 48 | 1.1339 | 0.6029 |
76
+ | 1.1074 | 5.9487 | 58 | 0.9396 | 0.75 |
77
+ | 0.9162 | 6.9744 | 68 | 0.8551 | 0.7647 |
78
+ | 0.8174 | 8.0 | 78 | 0.8291 | 0.7574 |
79
+ | 0.7135 | 8.9231 | 87 | 0.7505 | 0.7941 |
80
+ | 0.6222 | 9.9487 | 97 | 0.6434 | 0.8456 |
81
+ | 0.5445 | 10.9744 | 107 | 0.5996 | 0.8529 |
82
+ | 0.4935 | 12.0 | 117 | 0.5514 | 0.8529 |
83
+ | 0.4131 | 12.9231 | 126 | 0.5029 | 0.8603 |
84
+ | 0.4012 | 13.9487 | 136 | 0.5566 | 0.8382 |
85
+ | 0.3689 | 14.9744 | 146 | 0.5533 | 0.8382 |
86
+ | 0.3533 | 16.0 | 156 | 0.4232 | 0.8971 |
87
+ | 0.2954 | 16.9231 | 165 | 0.4589 | 0.8897 |
88
+ | 0.2907 | 17.9487 | 175 | 0.4223 | 0.8971 |
89
+ | 0.2804 | 18.9744 | 185 | 0.4056 | 0.8971 |
90
+ | 0.2469 | 20.0 | 195 | 0.3904 | 0.9118 |
91
+ | 0.2643 | 20.9231 | 204 | 0.3866 | 0.9044 |
92
+ | 0.2212 | 21.9487 | 214 | 0.4173 | 0.875 |
93
+ | 0.2476 | 22.9744 | 224 | 0.6001 | 0.8015 |
94
+ | 0.2347 | 24.0 | 234 | 0.3900 | 0.9044 |
95
+ | 0.207 | 24.9231 | 243 | 0.4033 | 0.8897 |
96
+ | 0.1803 | 25.9487 | 253 | 0.3510 | 0.9265 |
97
+ | 0.1979 | 26.9744 | 263 | 0.3723 | 0.9191 |
98
+ | 0.1821 | 28.0 | 273 | 0.4320 | 0.8824 |
99
+ | 0.1992 | 28.9231 | 282 | 0.3557 | 0.9118 |
100
+ | 0.2154 | 29.9487 | 292 | 0.3362 | 0.9191 |
101
+ | 0.1801 | 30.9744 | 302 | 0.4358 | 0.875 |
102
+ | 0.1794 | 32.0 | 312 | 0.3500 | 0.9191 |
103
+ | 0.1566 | 32.9231 | 321 | 0.3046 | 0.9265 |
104
+ | 0.1432 | 33.9487 | 331 | 0.3239 | 0.9265 |
105
+ | 0.145 | 34.9744 | 341 | 0.3311 | 0.9338 |
106
+ | 0.1578 | 36.0 | 351 | 0.3029 | 0.9338 |
107
+ | 0.1511 | 36.9231 | 360 | 0.3010 | 0.9338 |
108
+ | 0.139 | 37.9487 | 370 | 0.2982 | 0.9265 |
109
+ | 0.1294 | 38.9744 | 380 | 0.3261 | 0.9191 |
110
+ | 0.1263 | 40.0 | 390 | 0.2932 | 0.9338 |
111
+ | 0.1263 | 40.9231 | 399 | 0.2944 | 0.9338 |
112
+ | 0.1216 | 41.9487 | 409 | 0.2867 | 0.9338 |
113
+ | 0.1199 | 42.9744 | 419 | 0.2887 | 0.9338 |
114
+ | 0.128 | 44.0 | 429 | 0.2825 | 0.9338 |
115
+ | 0.1115 | 44.9231 | 438 | 0.2880 | 0.9338 |
116
+ | 0.1179 | 45.9487 | 448 | 0.2871 | 0.9338 |
117
+ | 0.12 | 46.1538 | 450 | 0.2870 | 0.9338 |
118
 
119
 
120
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:db1ce91d4b2d0899f8a624b33d3703e144587de2c0abef5e567dcbc93e5384d6
3
  size 343236280
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d25291798151095ab4d777e075e01885447ccfae7dbb6cf288cea854163acd2
3
  size 343236280