File size: 1,580 Bytes
742ef58
 
 
 
 
 
 
 
 
 
 
 
 
8e178d9
742ef58
8e178d9
742ef58
df14da9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
742ef58
 
 
 
 
 
df14da9
 
 
 
 
742ef58
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: microsoft/resnet-18
tags:
- generated_from_trainer
model-index:
- name: resnet18-food-classifier
  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. -->

# Model description

This model is a fine-tuned version of [microsoft/resnet-18](https://huggingface.co/microsoft/resnet-18) on an [custom](https://www.kaggle.com/datasets/faldoae/padangfood) dataset. This model was built using the "Padang Cuisine (Indonesian Food Image Classification)" dataset obtained from Kaggle. During the model building process, this was done using the Pytorch framework with pre-trained Resnet-18. The method used during the process of building this classification model is fine-tuning with the dataset.

## Training results

| Epoch | Accuracy |
|:-----:|:--------:|
| 1.0   | 0.6030   |
| 2.0   | 0.8342   |
| 3.0   | 0.8442   |
| 4.0   | 0.8191   |
| 5.0   | 0.8693   |
| 6.0   | 0.8643   |
| 7.0   | 0.8744   |
| 8.0   | 0.8643   |
| 9.0   | 0.8744   |
| 10.0  | 0.8744   |
| 11.0  | 0.8794   |
| 12.0  | 0.8744   |
| 13.0  | 0.8894   |
| 14.0  | 0.8794   |
| 15.0  | 0.8945   |

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- loss_function = CrossEntropyLoss
- optimizer = AdamW
- learning_rate: 0.00001
- batch_size: 16
- num_epochs: 15

### Framework versions

- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1