File size: 2,040 Bytes
e895f4f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9b303ce
 
e895f4f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9b303ce
 
 
 
 
 
 
 
 
 
 
e895f4f
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: vit-large-dataset-model-v3
  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. -->

# vit-large-dataset-model-v3

This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0630
- Accuracy: 0.9850

## 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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.0465        | 0.36  | 500  | 0.1289          | 0.9612   |
| 0.0253        | 0.71  | 1000 | 0.0983          | 0.9693   |
| 0.008         | 1.07  | 1500 | 0.0957          | 0.9728   |
| 0.0569        | 1.43  | 2000 | 0.0668          | 0.9793   |
| 0.035         | 1.79  | 2500 | 0.0865          | 0.9752   |
| 0.0034        | 2.14  | 3000 | 0.0748          | 0.9773   |
| 0.0638        | 2.5   | 3500 | 0.0708          | 0.9805   |
| 0.0195        | 2.86  | 4000 | 0.0782          | 0.9821   |
| 0.0012        | 3.21  | 4500 | 0.0739          | 0.9820   |
| 0.0013        | 3.57  | 5000 | 0.0680          | 0.9845   |
| 0.0417        | 3.93  | 5500 | 0.0630          | 0.9850   |


### Framework versions

- Transformers 4.22.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1