File size: 3,618 Bytes
ddacdf7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ddde838
ddacdf7
 
 
 
 
 
 
 
 
ddde838
 
ddacdf7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ddde838
 
ddacdf7
 
ddde838
ddacdf7
 
 
d52e137
ddacdf7
 
 
 
 
ddde838
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ddacdf7
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: microsoft/resnet-152
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: resnet-152-finetuned-cassava-leaf-disease
  results:
  - task:
      name: Image Classification
      type: image-classification
    dataset:
      name: imagefolder
      type: imagefolder
      config: default
      split: train
      args: default
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.7397196261682243
---

<!-- 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. -->

# resnet-152-finetuned-cassava-leaf-disease

This model is a fine-tuned version of [microsoft/resnet-152](https://huggingface.co/microsoft/resnet-152) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7961
- Accuracy: 0.7397

## 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: 5e-05
- train_batch_size: 480
- eval_batch_size: 480
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 1920
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 7.309         | 0.98  | 10   | 7.0088          | 0.0028   |
| 6.9946        | 1.95  | 20   | 6.4363          | 0.0061   |
| 6.4082        | 2.93  | 30   | 5.5840          | 0.0673   |
| 5.6018        | 4.0   | 41   | 4.1884          | 0.3687   |
| 4.5652        | 4.98  | 51   | 3.3123          | 0.4640   |
| 3.6106        | 5.95  | 61   | 2.7918          | 0.5136   |
| 2.9184        | 6.93  | 71   | 2.3762          | 0.5636   |
| 2.3775        | 8.0   | 82   | 1.9163          | 0.6084   |
| 2.0119        | 8.98  | 92   | 1.7038          | 0.6299   |
| 1.7519        | 9.95  | 102  | 1.5220          | 0.6411   |
| 1.4995        | 10.93 | 112  | 1.3828          | 0.6575   |
| 1.3648        | 12.0  | 123  | 1.2715          | 0.6668   |
| 1.2357        | 12.98 | 133  | 1.2040          | 0.6692   |
| 1.1606        | 13.95 | 143  | 1.1249          | 0.6785   |
| 1.0793        | 14.93 | 153  | 1.0600          | 0.6897   |
| 1.0332        | 16.0  | 164  | 1.0160          | 0.6935   |
| 0.9724        | 16.98 | 174  | 0.9706          | 0.7047   |
| 0.9349        | 17.95 | 184  | 0.9524          | 0.7075   |
| 0.895         | 18.93 | 194  | 0.9210          | 0.7093   |
| 0.8913        | 20.0  | 205  | 0.9007          | 0.7168   |
| 0.8519        | 20.98 | 215  | 0.8672          | 0.7229   |
| 0.8434        | 21.95 | 225  | 0.8432          | 0.7252   |
| 0.8346        | 22.93 | 235  | 0.8307          | 0.7304   |
| 0.8019        | 24.0  | 246  | 0.8154          | 0.7308   |
| 0.8001        | 24.98 | 256  | 0.8121          | 0.7327   |
| 0.7813        | 25.95 | 266  | 0.8036          | 0.7341   |
| 0.7845        | 26.93 | 276  | 0.8025          | 0.7383   |
| 0.7635        | 28.0  | 287  | 0.7934          | 0.7444   |
| 0.7782        | 28.98 | 297  | 0.7910          | 0.7421   |
| 0.7634        | 29.27 | 300  | 0.7961          | 0.7397   |


### Framework versions

- Transformers 4.37.2
- Pytorch 2.2.1
- Datasets 2.18.0
- Tokenizers 0.15.1