File size: 6,658 Bytes
ae7dad9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: facebook/deit-base-distilled-patch16-224
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: deit-cvc-drop-aug
  results: []
---

<!-- 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. -->

# deit-cvc-drop-aug

This model is a fine-tuned version of [facebook/deit-base-distilled-patch16-224](https://huggingface.co/facebook/deit-base-distilled-patch16-224) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4769
- Precision: 0.8894
- Recall: 0.8064
- F1: 0.8458
- Accuracy: 0.8489

## 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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 17
- gradient_accumulation_steps: 6
- total_train_batch_size: 96
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 15
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.5453        | 0.27  | 100  | 0.4824          | 0.7776    | 0.7726 | 0.7751 | 0.7696   |
| 0.4324        | 0.54  | 200  | 0.4796          | 0.8033    | 0.7279 | 0.7637 | 0.7686   |
| 0.4042        | 0.82  | 300  | 0.3790          | 0.7697    | 0.9608 | 0.8547 | 0.8321   |
| 0.3849        | 1.09  | 400  | 0.4100          | 0.8125    | 0.8198 | 0.8161 | 0.8102   |
| 0.3621        | 1.36  | 500  | 0.3689          | 0.8099    | 0.8967 | 0.8511 | 0.8387   |
| 0.3457        | 1.63  | 600  | 0.3313          | 0.7896    | 0.9543 | 0.8642 | 0.8459   |
| 0.3443        | 1.9   | 700  | 0.3424          | 0.7836    | 0.9528 | 0.8600 | 0.8405   |
| 0.3287        | 2.18  | 800  | 0.3308          | 0.8206    | 0.8947 | 0.8561 | 0.8454   |
| 0.3224        | 2.45  | 900  | 0.4546          | 0.8481    | 0.6624 | 0.7438 | 0.7655   |
| 0.3096        | 2.72  | 1000 | 0.3402          | 0.8300    | 0.8754 | 0.8521 | 0.8438   |
| 0.3095        | 2.99  | 1100 | 0.3691          | 0.8035    | 0.9076 | 0.8524 | 0.8385   |
| 0.2901        | 3.27  | 1200 | 0.3643          | 0.8008    | 0.8982 | 0.8467 | 0.8329   |
| 0.2939        | 3.54  | 1300 | 0.3021          | 0.8047    | 0.9613 | 0.8760 | 0.8602   |
| 0.2946        | 3.81  | 1400 | 0.3617          | 0.8363    | 0.8322 | 0.8342 | 0.8301   |
| 0.2856        | 4.08  | 1500 | 0.4884          | 0.8401    | 0.7850 | 0.8116 | 0.8127   |
| 0.2683        | 4.35  | 1600 | 0.3540          | 0.841     | 0.8352 | 0.8381 | 0.8341   |
| 0.2724        | 4.63  | 1700 | 0.3078          | 0.8391    | 0.8957 | 0.8665 | 0.8581   |
| 0.2685        | 4.9   | 1800 | 0.2913          | 0.8455    | 0.8967 | 0.8704 | 0.8627   |
| 0.2449        | 5.17  | 1900 | 0.3866          | 0.8465    | 0.8515 | 0.8490 | 0.8443   |
| 0.2468        | 5.44  | 2000 | 0.3072          | 0.8406    | 0.8952 | 0.8670 | 0.8589   |
| 0.2557        | 5.71  | 2100 | 0.3735          | 0.8595    | 0.7900 | 0.8233 | 0.8257   |
| 0.25          | 5.99  | 2200 | 0.3117          | 0.8755    | 0.8376 | 0.8561 | 0.8553   |
| 0.2256        | 6.26  | 2300 | 0.3264          | 0.8407    | 0.8913 | 0.8653 | 0.8574   |
| 0.234         | 6.53  | 2400 | 0.3617          | 0.8950    | 0.7572 | 0.8203 | 0.8295   |
| 0.2259        | 6.8   | 2500 | 0.3284          | 0.8476    | 0.8893 | 0.8679 | 0.8609   |
| 0.2261        | 7.07  | 2600 | 0.3486          | 0.9034    | 0.7805 | 0.8375 | 0.8443   |
| 0.2087        | 7.35  | 2700 | 0.3971          | 0.8628    | 0.8118 | 0.8365 | 0.8369   |
| 0.2035        | 7.62  | 2800 | 0.3106          | 0.8722    | 0.8942 | 0.8831 | 0.8783   |
| 0.2116        | 7.89  | 2900 | 0.3734          | 0.8805    | 0.8083 | 0.8429 | 0.8451   |
| 0.1956        | 8.16  | 3000 | 0.3443          | 0.8612    | 0.8654 | 0.8633 | 0.8591   |
| 0.1826        | 8.44  | 3100 | 0.3795          | 0.8908    | 0.7900 | 0.8374 | 0.8423   |
| 0.1918        | 8.71  | 3200 | 0.3362          | 0.8894    | 0.8267 | 0.8569 | 0.8581   |
| 0.1886        | 8.98  | 3300 | 0.3259          | 0.8589    | 0.8798 | 0.8693 | 0.8640   |
| 0.1716        | 9.25  | 3400 | 0.4269          | 0.8621    | 0.8347 | 0.8481 | 0.8464   |
| 0.1654        | 9.52  | 3500 | 0.4066          | 0.8881    | 0.8317 | 0.8590 | 0.8597   |
| 0.1625        | 9.8   | 3600 | 0.3927          | 0.8882    | 0.8128 | 0.8488 | 0.8512   |
| 0.1659        | 10.07 | 3700 | 0.3797          | 0.8895    | 0.8193 | 0.8529 | 0.8548   |
| 0.1519        | 10.34 | 3800 | 0.4089          | 0.8808    | 0.8217 | 0.8502 | 0.8512   |
| 0.1484        | 10.61 | 3900 | 0.3865          | 0.8853    | 0.8237 | 0.8534 | 0.8546   |
| 0.1427        | 10.88 | 4000 | 0.4347          | 0.8892    | 0.8009 | 0.8427 | 0.8464   |
| 0.1375        | 11.16 | 4100 | 0.4688          | 0.8878    | 0.8213 | 0.8532 | 0.8548   |
| 0.1276        | 11.43 | 4200 | 0.4687          | 0.8932    | 0.7974 | 0.8426 | 0.8469   |
| 0.1275        | 11.7  | 4300 | 0.4493          | 0.8936    | 0.8009 | 0.8447 | 0.8487   |
| 0.1349        | 11.97 | 4400 | 0.4618          | 0.8975    | 0.7825 | 0.8361 | 0.8423   |
| 0.1217        | 12.24 | 4500 | 0.4636          | 0.8987    | 0.7974 | 0.8450 | 0.8497   |
| 0.1211        | 12.52 | 4600 | 0.4527          | 0.8815    | 0.8307 | 0.8553 | 0.8556   |
| 0.1164        | 12.79 | 4700 | 0.4669          | 0.8950    | 0.8123 | 0.8516 | 0.8546   |
| 0.1119        | 13.06 | 4800 | 0.4617          | 0.8875    | 0.8148 | 0.8496 | 0.8517   |
| 0.11          | 13.33 | 4900 | 0.4718          | 0.8894    | 0.8103 | 0.8480 | 0.8507   |
| 0.1138        | 13.61 | 5000 | 0.4892          | 0.8939    | 0.7989 | 0.8437 | 0.8479   |
| 0.1058        | 13.88 | 5100 | 0.4725          | 0.8875    | 0.8108 | 0.8474 | 0.8500   |
| 0.1042        | 14.15 | 5200 | 0.4788          | 0.8908    | 0.8064 | 0.8465 | 0.8497   |
| 0.107         | 14.42 | 5300 | 0.4759          | 0.8900    | 0.8073 | 0.8467 | 0.8497   |
| 0.1047        | 14.69 | 5400 | 0.4767          | 0.8894    | 0.8064 | 0.8458 | 0.8489   |
| 0.1085        | 14.97 | 5500 | 0.4769          | 0.8894    | 0.8064 | 0.8458 | 0.8489   |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.16.1
- Tokenizers 0.15.0