Hugging Face's logo

Melo1512
/
vit-msn-small-wbc-blur-detector

Image Classification
Transformers
TensorBoard
Safetensors
vit_msn
Generated from Trainer
Eval Results
Model card Files
xet
Metrics Community
Melo1512 commited on
Commit
c179302
·
verified ·
1 Parent(s): cbd3b81

Model save

Browse files
Files changed (1) hide show
  1. README.md +12 -67
README.md CHANGED
@@ -23,7 +23,7 @@ model-index:
23
  metrics:
24
  - name: Accuracy
25
  type: accuracy
26
- value: 0.9565217391304348
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 [facebook/vit-msn-small](https://huggingface.co/facebook/vit-msn-small) on the imagefolder dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.1189
37
- - Accuracy: 0.9565
38
 
39
  ## Model description
40
 
@@ -54,80 +54,25 @@ More information needed
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: 60
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
- | No log | 1.0 | 3 | 0.6142 | 0.6304 |
72
- | No log | 2.0 | 6 | 0.3853 | 0.8696 |
73
- | No log | 3.0 | 9 | 0.4070 | 0.8261 |
74
- | 0.494 | 4.0 | 12 | 0.1461 | 0.9348 |
75
- | 0.494 | 5.0 | 15 | 0.1189 | 0.9565 |
76
- | 0.494 | 6.0 | 18 | 0.1527 | 0.9457 |
77
- | 0.2024 | 7.0 | 21 | 0.3323 | 0.9022 |
78
- | 0.2024 | 8.0 | 24 | 0.1520 | 0.9457 |
79
- | 0.2024 | 9.0 | 27 | 0.1572 | 0.9457 |
80
- | 0.1419 | 10.0 | 30 | 0.1814 | 0.9348 |
81
- | 0.1419 | 11.0 | 33 | 0.1778 | 0.9348 |
82
- | 0.1419 | 12.0 | 36 | 0.1505 | 0.9348 |
83
- | 0.1419 | 13.0 | 39 | 0.1891 | 0.9457 |
84
- | 0.1053 | 14.0 | 42 | 0.7274 | 0.7935 |
85
- | 0.1053 | 15.0 | 45 | 0.2669 | 0.9348 |
86
- | 0.1053 | 16.0 | 48 | 0.2240 | 0.9348 |
87
- | 0.3044 | 17.0 | 51 | 0.3497 | 0.8913 |
88
- | 0.3044 | 18.0 | 54 | 0.2208 | 0.9348 |
89
- | 0.3044 | 19.0 | 57 | 0.1733 | 0.9565 |
90
- | 0.151 | 20.0 | 60 | 0.2038 | 0.9239 |
91
- | 0.151 | 21.0 | 63 | 0.1282 | 0.9565 |
92
- | 0.151 | 22.0 | 66 | 0.3231 | 0.9239 |
93
- | 0.151 | 23.0 | 69 | 0.1565 | 0.9565 |
94
- | 0.0875 | 24.0 | 72 | 0.1981 | 0.9457 |
95
- | 0.0875 | 25.0 | 75 | 0.1974 | 0.9457 |
96
- | 0.0875 | 26.0 | 78 | 0.2045 | 0.9457 |
97
- | 0.0851 | 27.0 | 81 | 0.1841 | 0.9457 |
98
- | 0.0851 | 28.0 | 84 | 0.2061 | 0.9565 |
99
- | 0.0851 | 29.0 | 87 | 0.2077 | 0.9457 |
100
- | 0.046 | 30.0 | 90 | 0.2199 | 0.9565 |
101
- | 0.046 | 31.0 | 93 | 0.2038 | 0.9565 |
102
- | 0.046 | 32.0 | 96 | 0.2077 | 0.9457 |
103
- | 0.046 | 33.0 | 99 | 0.1877 | 0.9565 |
104
- | 0.0533 | 34.0 | 102 | 0.2383 | 0.9348 |
105
- | 0.0533 | 35.0 | 105 | 0.2571 | 0.9239 |
106
- | 0.0533 | 36.0 | 108 | 0.2330 | 0.9565 |
107
- | 0.0451 | 37.0 | 111 | 0.2420 | 0.9457 |
108
- | 0.0451 | 38.0 | 114 | 0.2882 | 0.9239 |
109
- | 0.0451 | 39.0 | 117 | 0.2386 | 0.9457 |
110
- | 0.0401 | 40.0 | 120 | 0.2513 | 0.9348 |
111
- | 0.0401 | 41.0 | 123 | 0.2672 | 0.9348 |
112
- | 0.0401 | 42.0 | 126 | 0.2950 | 0.9457 |
113
- | 0.0401 | 43.0 | 129 | 0.3232 | 0.9457 |
114
- | 0.0329 | 44.0 | 132 | 0.3712 | 0.9239 |
115
- | 0.0329 | 45.0 | 135 | 0.3529 | 0.9348 |
116
- | 0.0329 | 46.0 | 138 | 0.2905 | 0.9457 |
117
- | 0.0519 | 47.0 | 141 | 0.2670 | 0.9457 |
118
- | 0.0519 | 48.0 | 144 | 0.2629 | 0.9457 |
119
- | 0.0519 | 49.0 | 147 | 0.2761 | 0.9457 |
120
- | 0.0281 | 50.0 | 150 | 0.3040 | 0.9457 |
121
- | 0.0281 | 51.0 | 153 | 0.3191 | 0.9457 |
122
- | 0.0281 | 52.0 | 156 | 0.3214 | 0.9457 |
123
- | 0.0281 | 53.0 | 159 | 0.3132 | 0.9457 |
124
- | 0.028 | 54.0 | 162 | 0.3115 | 0.9457 |
125
- | 0.028 | 55.0 | 165 | 0.3116 | 0.9565 |
126
- | 0.028 | 56.0 | 168 | 0.3225 | 0.9457 |
127
- | 0.0361 | 57.0 | 171 | 0.3235 | 0.9457 |
128
- | 0.0361 | 58.0 | 174 | 0.3200 | 0.9457 |
129
- | 0.0361 | 59.0 | 177 | 0.3183 | 0.9457 |
130
- | 0.0312 | 60.0 | 180 | 0.3181 | 0.9457 |
131
 
132
 
133
  ### Framework versions
 
23
  metrics:
24
  - name: Accuracy
25
  type: accuracy
26
+ value: 0.92511454202441
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 [facebook/vit-msn-small](https://huggingface.co/facebook/vit-msn-small) on the imagefolder dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.2045
37
+ - Accuracy: 0.9251
38
 
39
  ## Model description
40
 
 
54
 
55
  The following hyperparameters were used during training:
56
  - learning_rate: 5e-05
57
+ - train_batch_size: 64
58
+ - eval_batch_size: 64
59
  - seed: 42
60
  - gradient_accumulation_steps: 4
61
+ - total_train_batch_size: 256
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: 5
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
+ | 0.3471 | 1.0 | 208 | 0.2960 | 0.8940 |
72
+ | 0.3113 | 2.0 | 416 | 0.2551 | 0.9088 |
73
+ | 0.3104 | 3.0 | 624 | 0.2106 | 0.9212 |
74
+ | 0.2855 | 4.0 | 832 | 0.2101 | 0.9221 |
75
+ | 0.2497 | 5.0 | 1040 | 0.2045 | 0.9251 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
76
 
77
 
78
  ### Framework versions