File size: 4,864 Bytes
0035e5a
 
 
 
 
 
7356ed1
 
 
0035e5a
 
 
 
 
 
 
 
 
 
 
 
5c89330
 
 
 
 
0035e5a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
072f8cc
0035e5a
 
 
10e1454
0035e5a
 
 
 
7356ed1
 
5c89330
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0035e5a
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: vit-base-skin
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# vit-base-skin

This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7037
- Accuracy: 0.8549
- F1: 0.8534
- Precision: 0.8536
- Recall: 0.8549

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.7904        | 0.16  | 100  | 0.7231          | 0.7772   | 0.7372 | 0.7397    | 0.7772 |
| 0.7746        | 0.32  | 200  | 0.6372          | 0.7668   | 0.7147 | 0.7458    | 0.7668 |
| 0.611         | 0.48  | 300  | 0.6455          | 0.7409   | 0.6930 | 0.7665    | 0.7409 |
| 0.8406        | 0.64  | 400  | 0.6233          | 0.8031   | 0.8098 | 0.8433    | 0.8031 |
| 0.587         | 0.8   | 500  | 0.6028          | 0.7513   | 0.7163 | 0.7659    | 0.7513 |
| 0.5532        | 0.96  | 600  | 0.4689          | 0.8290   | 0.8090 | 0.8377    | 0.8290 |
| 0.4326        | 1.12  | 700  | 0.4968          | 0.8290   | 0.8200 | 0.8368    | 0.8290 |
| 0.4713        | 1.28  | 800  | 0.4973          | 0.8187   | 0.8222 | 0.8436    | 0.8187 |
| 0.4333        | 1.44  | 900  | 0.5500          | 0.7720   | 0.7615 | 0.7705    | 0.7720 |
| 0.441         | 1.6   | 1000 | 0.5518          | 0.8238   | 0.8398 | 0.8774    | 0.8238 |
| 0.4172        | 1.76  | 1100 | 0.5608          | 0.8031   | 0.7802 | 0.8260    | 0.8031 |
| 0.4062        | 1.92  | 1200 | 0.4730          | 0.8290   | 0.8312 | 0.8704    | 0.8290 |
| 0.271         | 2.08  | 1300 | 0.4893          | 0.8031   | 0.8018 | 0.8164    | 0.8031 |
| 0.2294        | 2.24  | 1400 | 0.4859          | 0.8342   | 0.8369 | 0.8442    | 0.8342 |
| 0.2687        | 2.4   | 1500 | 0.4805          | 0.8394   | 0.8391 | 0.8424    | 0.8394 |
| 0.2348        | 2.56  | 1600 | 0.4667          | 0.8497   | 0.8522 | 0.8567    | 0.8497 |
| 0.2038        | 2.72  | 1700 | 0.5050          | 0.8135   | 0.8148 | 0.8222    | 0.8135 |
| 0.2102        | 2.88  | 1800 | 0.4730          | 0.8497   | 0.8527 | 0.8695    | 0.8497 |
| 0.0978        | 3.04  | 1900 | 0.4673          | 0.8446   | 0.8450 | 0.8508    | 0.8446 |
| 0.104         | 3.19  | 2000 | 0.5348          | 0.8342   | 0.8274 | 0.8313    | 0.8342 |
| 0.0562        | 3.35  | 2100 | 0.5748          | 0.8342   | 0.8264 | 0.8299    | 0.8342 |
| 0.1443        | 3.51  | 2200 | 0.5903          | 0.8446   | 0.8432 | 0.8448    | 0.8446 |
| 0.1245        | 3.67  | 2300 | 0.5773          | 0.8601   | 0.8627 | 0.8779    | 0.8601 |
| 0.081         | 3.83  | 2400 | 0.6190          | 0.8394   | 0.8424 | 0.8487    | 0.8394 |
| 0.1314        | 3.99  | 2500 | 0.6078          | 0.8549   | 0.8509 | 0.8506    | 0.8549 |
| 0.0415        | 4.15  | 2600 | 0.7039          | 0.8290   | 0.8312 | 0.8358    | 0.8290 |
| 0.0402        | 4.31  | 2700 | 0.7477          | 0.8238   | 0.8166 | 0.8179    | 0.8238 |
| 0.0045        | 4.47  | 2800 | 0.7207          | 0.8497   | 0.8493 | 0.8539    | 0.8497 |
| 0.0608        | 4.63  | 2900 | 0.7339          | 0.8342   | 0.8370 | 0.8469    | 0.8342 |
| 0.0168        | 4.79  | 3000 | 0.7894          | 0.8290   | 0.8388 | 0.8539    | 0.8290 |
| 0.0042        | 4.95  | 3100 | 0.7268          | 0.8601   | 0.8628 | 0.8681    | 0.8601 |
| 0.0149        | 5.11  | 3200 | 0.7145          | 0.8601   | 0.8577 | 0.8600    | 0.8601 |
| 0.0074        | 5.27  | 3300 | 0.7424          | 0.8342   | 0.8354 | 0.8380    | 0.8342 |
| 0.0029        | 5.43  | 3400 | 0.7123          | 0.8653   | 0.8649 | 0.8686    | 0.8653 |
| 0.0123        | 5.59  | 3500 | 0.7052          | 0.8653   | 0.8633 | 0.8632    | 0.8653 |
| 0.0028        | 5.75  | 3600 | 0.7027          | 0.8601   | 0.8590 | 0.8601    | 0.8601 |
| 0.0029        | 5.91  | 3700 | 0.7037          | 0.8549   | 0.8534 | 0.8536    | 0.8549 |


### Framework versions

- Transformers 4.29.2
- Pytorch 1.13.1
- Datasets 2.14.5
- Tokenizers 0.13.3