File size: 4,808 Bytes
f813776
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: facebook/deit-small-patch16-224
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: hushem_1x_deit_small_adamax_00001_fold1
  results:
  - task:
      name: Image Classification
      type: image-classification
    dataset:
      name: imagefolder
      type: imagefolder
      config: default
      split: test
      args: default
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.6
---

<!-- 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. -->

# hushem_1x_deit_small_adamax_00001_fold1

This model is a fine-tuned version of [facebook/deit-small-patch16-224](https://huggingface.co/facebook/deit-small-patch16-224) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1270
- Accuracy: 0.6

## 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: 1e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log        | 1.0   | 6    | 1.3199          | 0.3333   |
| 1.3414        | 2.0   | 12   | 1.2923          | 0.4667   |
| 1.3414        | 3.0   | 18   | 1.2886          | 0.4667   |
| 1.0791        | 4.0   | 24   | 1.2761          | 0.4667   |
| 0.9244        | 5.0   | 30   | 1.2453          | 0.4889   |
| 0.9244        | 6.0   | 36   | 1.2252          | 0.4667   |
| 0.7694        | 7.0   | 42   | 1.2158          | 0.5111   |
| 0.7694        | 8.0   | 48   | 1.2163          | 0.4667   |
| 0.6552        | 9.0   | 54   | 1.2081          | 0.5111   |
| 0.5314        | 10.0  | 60   | 1.1883          | 0.5556   |
| 0.5314        | 11.0  | 66   | 1.1802          | 0.5556   |
| 0.4407        | 12.0  | 72   | 1.1737          | 0.5778   |
| 0.4407        | 13.0  | 78   | 1.1623          | 0.6222   |
| 0.3864        | 14.0  | 84   | 1.1625          | 0.6222   |
| 0.3093        | 15.0  | 90   | 1.1653          | 0.6222   |
| 0.3093        | 16.0  | 96   | 1.1658          | 0.6222   |
| 0.2597        | 17.0  | 102  | 1.1519          | 0.6444   |
| 0.2597        | 18.0  | 108  | 1.1466          | 0.6222   |
| 0.2099        | 19.0  | 114  | 1.1591          | 0.6      |
| 0.1766        | 20.0  | 120  | 1.1509          | 0.5778   |
| 0.1766        | 21.0  | 126  | 1.1488          | 0.5778   |
| 0.1537        | 22.0  | 132  | 1.1482          | 0.5778   |
| 0.1537        | 23.0  | 138  | 1.1427          | 0.6222   |
| 0.1244        | 24.0  | 144  | 1.1370          | 0.6      |
| 0.103         | 25.0  | 150  | 1.1285          | 0.6      |
| 0.103         | 26.0  | 156  | 1.1323          | 0.6      |
| 0.089         | 27.0  | 162  | 1.1268          | 0.6      |
| 0.089         | 28.0  | 168  | 1.1377          | 0.6      |
| 0.0777        | 29.0  | 174  | 1.1346          | 0.6      |
| 0.068         | 30.0  | 180  | 1.1274          | 0.6      |
| 0.068         | 31.0  | 186  | 1.1199          | 0.6      |
| 0.0597        | 32.0  | 192  | 1.1245          | 0.6      |
| 0.0597        | 33.0  | 198  | 1.1296          | 0.6      |
| 0.0547        | 34.0  | 204  | 1.1270          | 0.6      |
| 0.0493        | 35.0  | 210  | 1.1241          | 0.6      |
| 0.0493        | 36.0  | 216  | 1.1250          | 0.6      |
| 0.0441        | 37.0  | 222  | 1.1253          | 0.6      |
| 0.0441        | 38.0  | 228  | 1.1296          | 0.6      |
| 0.0409        | 39.0  | 234  | 1.1287          | 0.6      |
| 0.0405        | 40.0  | 240  | 1.1275          | 0.6      |
| 0.0405        | 41.0  | 246  | 1.1272          | 0.6      |
| 0.0391        | 42.0  | 252  | 1.1270          | 0.6      |
| 0.0391        | 43.0  | 258  | 1.1270          | 0.6      |
| 0.0395        | 44.0  | 264  | 1.1270          | 0.6      |
| 0.0377        | 45.0  | 270  | 1.1270          | 0.6      |
| 0.0377        | 46.0  | 276  | 1.1270          | 0.6      |
| 0.0388        | 47.0  | 282  | 1.1270          | 0.6      |
| 0.0388        | 48.0  | 288  | 1.1270          | 0.6      |
| 0.0366        | 49.0  | 294  | 1.1270          | 0.6      |
| 0.0396        | 50.0  | 300  | 1.1270          | 0.6      |


### Framework versions

- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1