AykeeSalazar commited on
Commit
7b08026
1 Parent(s): 7f0cc9e

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +90 -0
README.md ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - image_folder
7
+ metrics:
8
+ - accuracy
9
+ model-index:
10
+ - name: violation-classification-bantai-vit-v80ep
11
+ results:
12
+ - task:
13
+ name: Image Classification
14
+ type: image-classification
15
+ dataset:
16
+ name: image_folder
17
+ type: image_folder
18
+ args: default
19
+ metrics:
20
+ - name: Accuracy
21
+ type: accuracy
22
+ value: 0.9559725730783111
23
+ ---
24
+
25
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
26
+ should probably proofread and complete it, then remove this comment. -->
27
+
28
+ # violation-classification-bantai-vit-v80ep
29
+
30
+ 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 image_folder dataset.
31
+ It achieves the following results on the evaluation set:
32
+ - Loss: 0.1974
33
+ - Accuracy: 0.9560
34
+
35
+ ## Model description
36
+
37
+ More information needed
38
+
39
+ ## Intended uses & limitations
40
+
41
+ More information needed
42
+
43
+ ## Training and evaluation data
44
+
45
+ More information needed
46
+
47
+ ## Training procedure
48
+
49
+ ### Training hyperparameters
50
+
51
+ The following hyperparameters were used during training:
52
+ - learning_rate: 5e-05
53
+ - train_batch_size: 32
54
+ - eval_batch_size: 32
55
+ - seed: 42
56
+ - gradient_accumulation_steps: 4
57
+ - total_train_batch_size: 128
58
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
+ - lr_scheduler_type: linear
60
+ - lr_scheduler_warmup_ratio: 0.1
61
+ - num_epochs: 80
62
+
63
+ ### Training results
64
+
65
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
66
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
67
+ | 0.797 | 4.95 | 500 | 0.3926 | 0.8715 |
68
+ | 0.3095 | 9.9 | 1000 | 0.2597 | 0.9107 |
69
+ | 0.1726 | 14.85 | 1500 | 0.2157 | 0.9253 |
70
+ | 0.1259 | 19.8 | 2000 | 0.1870 | 0.9392 |
71
+ | 0.0959 | 24.75 | 2500 | 0.1797 | 0.9444 |
72
+ | 0.0835 | 29.7 | 3000 | 0.2293 | 0.9354 |
73
+ | 0.0722 | 34.65 | 3500 | 0.1921 | 0.9441 |
74
+ | 0.0628 | 39.6 | 4000 | 0.1897 | 0.9491 |
75
+ | 0.059 | 44.55 | 4500 | 0.1719 | 0.9520 |
76
+ | 0.0531 | 49.5 | 5000 | 0.1987 | 0.9513 |
77
+ | 0.046 | 54.45 | 5500 | 0.1713 | 0.9556 |
78
+ | 0.0444 | 59.4 | 6000 | 0.2016 | 0.9525 |
79
+ | 0.042 | 64.36 | 6500 | 0.1950 | 0.9525 |
80
+ | 0.0363 | 69.31 | 7000 | 0.2017 | 0.9549 |
81
+ | 0.037 | 74.26 | 7500 | 0.1943 | 0.9551 |
82
+ | 0.0343 | 79.21 | 8000 | 0.1974 | 0.9560 |
83
+
84
+
85
+ ### Framework versions
86
+
87
+ - Transformers 4.17.0
88
+ - Pytorch 1.10.0+cu111
89
+ - Datasets 2.0.0
90
+ - Tokenizers 0.11.6