File size: 4,861 Bytes
991a57d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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: smids_5x_deit_tiny_rms_001_fold5
  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: 0.7433333333333333
---

<!-- 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. -->

# smids_5x_deit_tiny_rms_001_fold5

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: 0.6094
- Accuracy: 0.7433

## 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 |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.8822        | 1.0   | 375   | 0.8719          | 0.52     |
| 0.8936        | 2.0   | 750   | 0.8470          | 0.535    |
| 0.8252        | 3.0   | 1125  | 0.8071          | 0.595    |
| 0.8333        | 4.0   | 1500  | 0.7970          | 0.6017   |
| 0.8046        | 5.0   | 1875  | 0.8070          | 0.5633   |
| 0.8082        | 6.0   | 2250  | 0.9208          | 0.5167   |
| 0.7481        | 7.0   | 2625  | 0.7984          | 0.5633   |
| 0.8409        | 8.0   | 3000  | 0.7900          | 0.5783   |
| 0.7673        | 9.0   | 3375  | 0.7551          | 0.62     |
| 0.7321        | 10.0  | 3750  | 0.7485          | 0.6133   |
| 0.8282        | 11.0  | 4125  | 0.7517          | 0.6083   |
| 0.7206        | 12.0  | 4500  | 0.7745          | 0.6      |
| 0.6841        | 13.0  | 4875  | 0.8307          | 0.5917   |
| 0.7738        | 14.0  | 5250  | 0.7274          | 0.6683   |
| 0.8416        | 15.0  | 5625  | 0.7353          | 0.67     |
| 0.704         | 16.0  | 6000  | 0.7258          | 0.65     |
| 0.6873        | 17.0  | 6375  | 0.7174          | 0.68     |
| 0.714         | 18.0  | 6750  | 0.7557          | 0.6483   |
| 0.7105        | 19.0  | 7125  | 0.6868          | 0.6917   |
| 0.6559        | 20.0  | 7500  | 0.6845          | 0.6783   |
| 0.6717        | 21.0  | 7875  | 0.7043          | 0.67     |
| 0.7139        | 22.0  | 8250  | 0.6944          | 0.68     |
| 0.6633        | 23.0  | 8625  | 0.7071          | 0.6667   |
| 0.6888        | 24.0  | 9000  | 0.6979          | 0.6883   |
| 0.6621        | 25.0  | 9375  | 0.6468          | 0.7117   |
| 0.6157        | 26.0  | 9750  | 0.6767          | 0.6833   |
| 0.6777        | 27.0  | 10125 | 0.7097          | 0.67     |
| 0.7108        | 28.0  | 10500 | 0.6811          | 0.6917   |
| 0.8139        | 29.0  | 10875 | 0.6750          | 0.7067   |
| 0.6291        | 30.0  | 11250 | 0.6415          | 0.7133   |
| 0.5725        | 31.0  | 11625 | 0.6769          | 0.6833   |
| 0.6243        | 32.0  | 12000 | 0.6733          | 0.7267   |
| 0.6311        | 33.0  | 12375 | 0.6227          | 0.7217   |
| 0.6254        | 34.0  | 12750 | 0.6222          | 0.72     |
| 0.567         | 35.0  | 13125 | 0.6040          | 0.735    |
| 0.5363        | 36.0  | 13500 | 0.5935          | 0.7533   |
| 0.6308        | 37.0  | 13875 | 0.6047          | 0.7267   |
| 0.5334        | 38.0  | 14250 | 0.6481          | 0.7217   |
| 0.5951        | 39.0  | 14625 | 0.6059          | 0.7317   |
| 0.6325        | 40.0  | 15000 | 0.6172          | 0.735    |
| 0.5905        | 41.0  | 15375 | 0.6255          | 0.7233   |
| 0.6095        | 42.0  | 15750 | 0.5896          | 0.7433   |
| 0.49          | 43.0  | 16125 | 0.5925          | 0.7367   |
| 0.4891        | 44.0  | 16500 | 0.5937          | 0.7367   |
| 0.4867        | 45.0  | 16875 | 0.5918          | 0.7583   |
| 0.5178        | 46.0  | 17250 | 0.6030          | 0.735    |
| 0.561         | 47.0  | 17625 | 0.6183          | 0.74     |
| 0.4632        | 48.0  | 18000 | 0.5943          | 0.7517   |
| 0.4666        | 49.0  | 18375 | 0.6107          | 0.7417   |
| 0.4901        | 50.0  | 18750 | 0.6094          | 0.7433   |


### Framework versions

- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2