File size: 4,836 Bytes
e46fca7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5a58317
e46fca7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: vit-base-patch16-224-Trial006_007_008-YEL_STEM1
  results:
  - task:
      name: Image Classification
      type: image-classification
    dataset:
      name: imagefolder
      type: imagefolder
      config: default
      split: train
      args: default
    metrics:
    - name: Accuracy
      type: accuracy
      value: 1.0
---

<!-- 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-patch16-224-Trial006_007_008-YEL_STEM1

This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0453
- Accuracy: 1.0

## 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: 5e-05
- train_batch_size: 30
- eval_batch_size: 30
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 120
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.7064        | 1.0   | 5    | 0.6605          | 0.5373   |
| 0.5838        | 2.0   | 10   | 0.5289          | 0.8358   |
| 0.4909        | 3.0   | 15   | 0.3967          | 0.8209   |
| 0.3317        | 4.0   | 20   | 0.2759          | 0.9104   |
| 0.2813        | 5.0   | 25   | 0.1820          | 0.9403   |
| 0.2948        | 6.0   | 30   | 0.1286          | 0.9552   |
| 0.2253        | 7.0   | 35   | 0.0453          | 1.0      |
| 0.2125        | 8.0   | 40   | 0.0408          | 1.0      |
| 0.1288        | 9.0   | 45   | 0.0177          | 1.0      |
| 0.1084        | 10.0  | 50   | 0.0265          | 1.0      |
| 0.1642        | 11.0  | 55   | 0.0994          | 0.9403   |
| 0.149         | 12.0  | 60   | 0.0316          | 0.9851   |
| 0.1315        | 13.0  | 65   | 0.0325          | 0.9851   |
| 0.1101        | 14.0  | 70   | 0.1090          | 0.9701   |
| 0.1101        | 15.0  | 75   | 0.0094          | 1.0      |
| 0.0702        | 16.0  | 80   | 0.0070          | 1.0      |
| 0.1184        | 17.0  | 85   | 0.0634          | 0.9851   |
| 0.1506        | 18.0  | 90   | 0.0104          | 1.0      |
| 0.1027        | 19.0  | 95   | 0.0149          | 1.0      |
| 0.159         | 20.0  | 100  | 0.1021          | 0.9552   |
| 0.1205        | 21.0  | 105  | 0.0085          | 1.0      |
| 0.1511        | 22.0  | 110  | 0.0248          | 0.9851   |
| 0.2228        | 23.0  | 115  | 0.0993          | 0.9552   |
| 0.1431        | 24.0  | 120  | 0.0373          | 0.9851   |
| 0.1489        | 25.0  | 125  | 0.0161          | 1.0      |
| 0.0799        | 26.0  | 130  | 0.0382          | 0.9851   |
| 0.1411        | 27.0  | 135  | 0.0071          | 1.0      |
| 0.1457        | 28.0  | 140  | 0.0047          | 1.0      |
| 0.1434        | 29.0  | 145  | 0.0069          | 1.0      |
| 0.0913        | 30.0  | 150  | 0.0032          | 1.0      |
| 0.1354        | 31.0  | 155  | 0.0042          | 1.0      |
| 0.1253        | 32.0  | 160  | 0.0061          | 1.0      |
| 0.1065        | 33.0  | 165  | 0.0039          | 1.0      |
| 0.1199        | 34.0  | 170  | 0.0023          | 1.0      |
| 0.1274        | 35.0  | 175  | 0.0037          | 1.0      |
| 0.1118        | 36.0  | 180  | 0.0100          | 1.0      |
| 0.1237        | 37.0  | 185  | 0.0053          | 1.0      |
| 0.1311        | 38.0  | 190  | 0.0028          | 1.0      |
| 0.1833        | 39.0  | 195  | 0.0026          | 1.0      |
| 0.0858        | 40.0  | 200  | 0.0033          | 1.0      |
| 0.1503        | 41.0  | 205  | 0.0049          | 1.0      |
| 0.0547        | 42.0  | 210  | 0.0037          | 1.0      |
| 0.1647        | 43.0  | 215  | 0.0037          | 1.0      |
| 0.1066        | 44.0  | 220  | 0.0061          | 1.0      |
| 0.1277        | 45.0  | 225  | 0.0083          | 1.0      |
| 0.0885        | 46.0  | 230  | 0.0083          | 1.0      |
| 0.1339        | 47.0  | 235  | 0.0081          | 1.0      |
| 0.0904        | 48.0  | 240  | 0.0073          | 1.0      |
| 0.079         | 49.0  | 245  | 0.0080          | 1.0      |
| 0.0788        | 50.0  | 250  | 0.0086          | 1.0      |


### Framework versions

- Transformers 4.30.0.dev0
- Pytorch 1.12.1
- Datasets 2.12.0
- Tokenizers 0.13.1