File size: 4,811 Bytes
0541161
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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-tiny-patch16-224
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: hushem_5x_deit_tiny_sgd_001_fold2
  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: 0.3111111111111111
---

<!-- 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_5x_deit_tiny_sgd_001_fold2

This model is a fine-tuned version of [facebook/deit-tiny-patch16-224](https://huggingface.co/facebook/deit-tiny-patch16-224) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5326
- Accuracy: 0.3111

## 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: 0.001
- 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.5319        | 1.0   | 27   | 1.6147          | 0.1778   |
| 1.412         | 2.0   | 54   | 1.5641          | 0.2222   |
| 1.3305        | 3.0   | 81   | 1.5365          | 0.2      |
| 1.301         | 4.0   | 108  | 1.5559          | 0.2222   |
| 1.2455        | 5.0   | 135  | 1.5605          | 0.2444   |
| 1.184         | 6.0   | 162  | 1.5721          | 0.2444   |
| 1.1536        | 7.0   | 189  | 1.5847          | 0.2444   |
| 1.141         | 8.0   | 216  | 1.6070          | 0.2667   |
| 1.0813        | 9.0   | 243  | 1.6240          | 0.2667   |
| 1.0544        | 10.0  | 270  | 1.6212          | 0.2667   |
| 1.0306        | 11.0  | 297  | 1.6262          | 0.2667   |
| 0.9926        | 12.0  | 324  | 1.6270          | 0.2667   |
| 0.9991        | 13.0  | 351  | 1.6433          | 0.2444   |
| 0.9662        | 14.0  | 378  | 1.6269          | 0.2667   |
| 0.9752        | 15.0  | 405  | 1.6379          | 0.2444   |
| 0.9275        | 16.0  | 432  | 1.6386          | 0.2444   |
| 0.9112        | 17.0  | 459  | 1.6378          | 0.2667   |
| 0.8926        | 18.0  | 486  | 1.6345          | 0.2667   |
| 0.8698        | 19.0  | 513  | 1.6300          | 0.2444   |
| 0.8732        | 20.0  | 540  | 1.6217          | 0.2444   |
| 0.8587        | 21.0  | 567  | 1.6212          | 0.2667   |
| 0.8545        | 22.0  | 594  | 1.6207          | 0.2667   |
| 0.8339        | 23.0  | 621  | 1.6201          | 0.2444   |
| 0.8104        | 24.0  | 648  | 1.6072          | 0.2667   |
| 0.7957        | 25.0  | 675  | 1.6070          | 0.2667   |
| 0.8197        | 26.0  | 702  | 1.6043          | 0.2444   |
| 0.8076        | 27.0  | 729  | 1.6022          | 0.2667   |
| 0.7686        | 28.0  | 756  | 1.5925          | 0.2889   |
| 0.7691        | 29.0  | 783  | 1.5965          | 0.2889   |
| 0.7835        | 30.0  | 810  | 1.5836          | 0.2889   |
| 0.7441        | 31.0  | 837  | 1.5828          | 0.2889   |
| 0.7775        | 32.0  | 864  | 1.5709          | 0.2889   |
| 0.7317        | 33.0  | 891  | 1.5664          | 0.2889   |
| 0.7292        | 34.0  | 918  | 1.5626          | 0.2889   |
| 0.7179        | 35.0  | 945  | 1.5496          | 0.2667   |
| 0.7386        | 36.0  | 972  | 1.5502          | 0.2889   |
| 0.7342        | 37.0  | 999  | 1.5475          | 0.3111   |
| 0.734         | 38.0  | 1026 | 1.5457          | 0.3111   |
| 0.7069        | 39.0  | 1053 | 1.5425          | 0.3111   |
| 0.7143        | 40.0  | 1080 | 1.5429          | 0.3111   |
| 0.7105        | 41.0  | 1107 | 1.5401          | 0.3111   |
| 0.7189        | 42.0  | 1134 | 1.5394          | 0.3111   |
| 0.7216        | 43.0  | 1161 | 1.5376          | 0.3111   |
| 0.6896        | 44.0  | 1188 | 1.5358          | 0.3111   |
| 0.7099        | 45.0  | 1215 | 1.5345          | 0.3111   |
| 0.6751        | 46.0  | 1242 | 1.5331          | 0.3111   |
| 0.6824        | 47.0  | 1269 | 1.5327          | 0.3111   |
| 0.7027        | 48.0  | 1296 | 1.5326          | 0.3111   |
| 0.7357        | 49.0  | 1323 | 1.5326          | 0.3111   |
| 0.6799        | 50.0  | 1350 | 1.5326          | 0.3111   |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0