File size: 4,856 Bytes
86339ef
 
d917ec4
86339ef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d917ec4
86339ef
 
 
 
 
 
 
d917ec4
86339ef
d917ec4
 
86339ef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d917ec4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
86339ef
 
 
 
 
 
 
 
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: smids_5x_deit_tiny_adamax_001_fold4
  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: 0.865
---

<!-- 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_adamax_001_fold4

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: 1.5310
- Accuracy: 0.865

## 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 |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.4168        | 1.0   | 375   | 0.3631          | 0.85     |
| 0.2785        | 2.0   | 750   | 0.4582          | 0.82     |
| 0.1977        | 3.0   | 1125  | 0.4757          | 0.845    |
| 0.2154        | 4.0   | 1500  | 0.4151          | 0.8567   |
| 0.2216        | 5.0   | 1875  | 0.4921          | 0.84     |
| 0.1277        | 6.0   | 2250  | 0.5208          | 0.84     |
| 0.1577        | 7.0   | 2625  | 0.6509          | 0.84     |
| 0.1043        | 8.0   | 3000  | 0.6131          | 0.8483   |
| 0.0606        | 9.0   | 3375  | 0.7321          | 0.85     |
| 0.0399        | 10.0  | 3750  | 0.7332          | 0.8483   |
| 0.0878        | 11.0  | 4125  | 0.7794          | 0.86     |
| 0.0753        | 12.0  | 4500  | 0.9361          | 0.855    |
| 0.0315        | 13.0  | 4875  | 0.7541          | 0.87     |
| 0.0322        | 14.0  | 5250  | 0.8827          | 0.855    |
| 0.0291        | 15.0  | 5625  | 0.8552          | 0.8667   |
| 0.0323        | 16.0  | 6000  | 1.0097          | 0.8533   |
| 0.0358        | 17.0  | 6375  | 1.0442          | 0.8367   |
| 0.0726        | 18.0  | 6750  | 1.0675          | 0.8533   |
| 0.0105        | 19.0  | 7125  | 1.0350          | 0.8567   |
| 0.0155        | 20.0  | 7500  | 1.0612          | 0.8467   |
| 0.0001        | 21.0  | 7875  | 1.1933          | 0.8467   |
| 0.001         | 22.0  | 8250  | 0.9964          | 0.86     |
| 0.0061        | 23.0  | 8625  | 1.0207          | 0.86     |
| 0.0139        | 24.0  | 9000  | 1.1598          | 0.8467   |
| 0.0232        | 25.0  | 9375  | 1.1652          | 0.8583   |
| 0.0001        | 26.0  | 9750  | 1.1454          | 0.8583   |
| 0.0011        | 27.0  | 10125 | 1.1331          | 0.865    |
| 0.0           | 28.0  | 10500 | 1.2646          | 0.8667   |
| 0.0           | 29.0  | 10875 | 1.1994          | 0.8683   |
| 0.0001        | 30.0  | 11250 | 1.2306          | 0.8533   |
| 0.004         | 31.0  | 11625 | 1.2452          | 0.8617   |
| 0.0           | 32.0  | 12000 | 1.2904          | 0.8633   |
| 0.0           | 33.0  | 12375 | 1.3971          | 0.86     |
| 0.0001        | 34.0  | 12750 | 1.2738          | 0.8633   |
| 0.0           | 35.0  | 13125 | 1.4099          | 0.865    |
| 0.0           | 36.0  | 13500 | 1.3138          | 0.8633   |
| 0.0           | 37.0  | 13875 | 1.3962          | 0.8617   |
| 0.0037        | 38.0  | 14250 | 1.4247          | 0.8633   |
| 0.0           | 39.0  | 14625 | 1.4177          | 0.865    |
| 0.0           | 40.0  | 15000 | 1.4033          | 0.8633   |
| 0.0           | 41.0  | 15375 | 1.4591          | 0.8633   |
| 0.0           | 42.0  | 15750 | 1.4725          | 0.8617   |
| 0.0           | 43.0  | 16125 | 1.4752          | 0.8633   |
| 0.0           | 44.0  | 16500 | 1.4834          | 0.8633   |
| 0.0           | 45.0  | 16875 | 1.4967          | 0.8633   |
| 0.0           | 46.0  | 17250 | 1.5039          | 0.8633   |
| 0.0           | 47.0  | 17625 | 1.5125          | 0.8633   |
| 0.0           | 48.0  | 18000 | 1.5211          | 0.8633   |
| 0.0           | 49.0  | 18375 | 1.5277          | 0.865    |
| 0.0           | 50.0  | 18750 | 1.5310          | 0.865    |


### Framework versions

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