File size: 3,417 Bytes
8bd2630
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
58f931d
8bd2630
65e7c6e
 
 
 
 
 
8bd2630
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
58f931d
 
65e7c6e
 
 
 
 
8bd2630
 
 
 
 
65e7c6e
 
8bd2630
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
---
license: apache-2.0
tags:
- vision
- image-segmentation
- generated_from_trainer
model-index:
- name: segformer-b0-pavement
  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. -->

# segformer-b0-pavement

This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the reannayang/FL_pavement dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4165
- Mean Iou: 0.6318
- Mean Accuracy: 0.9700
- Overall Accuracy: 0.9738
- Per Category Iou: [0.0, 0.964166382973358, 0.9809231860559384, 0.0, 0.9295139919583345, 0.9164463823409184]
- Per Category Accuracy: [nan, 0.9643001261034048, 0.9983497924348297, nan, 0.995031342981772, 0.9223532638507954]

## 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: 6e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50

### Training results

| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Per Category Iou                                                                           | Per Category Accuracy                                                                      |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------:|
| 1.0651        | 10.0  | 20   | 1.3005          | 0.5967   | 0.9512        | 0.9534           | [0.0, 0.9462421185372005, 0.9681701711239586, 0.0, 0.7994398965962947, 0.8662896799897185] | [nan, 0.9462421185372005, 0.9693809143181291, nan, 0.9648149753011526, 0.9243828853538124] |
| 0.5732        | 20.0  | 40   | 0.6626          | 0.6287   | 0.9702        | 0.9760           | [0.0, 0.975246652572234, 0.985446932366533, 0.0, 0.9010974339804011, 0.9103918683964157]   | [nan, 0.9772635561160151, 0.9952040842637238, nan, 0.9748678395008233, 0.9334887547997806] |
| 0.6987        | 30.0  | 60   | 0.4319          | 0.6317   | 0.9705        | 0.9758           | [0.0, 0.9709705045212967, 0.9798115236227942, 0.0, 0.9255918522130127, 0.9139245313729214] | [nan, 0.9722194199243379, 0.9986205296134905, nan, 0.9871161568015715, 0.924026330224904]  |
| 0.6915        | 40.0  | 80   | 0.4382          | 0.6237   | 0.9634        | 0.9692           | [0.0, 0.9611727616645649, 0.9725125142706595, 0.0, 0.9147983251179308, 0.8937433316006894] | [nan, 0.9611727616645649, 0.9993811721630611, nan, 0.9971690210012422, 0.896023038946791]  |
| 0.4373        | 50.0  | 100  | 0.4165          | 0.6318   | 0.9700        | 0.9738           | [0.0, 0.964166382973358, 0.9809231860559384, 0.0, 0.9295139919583345, 0.9164463823409184]  | [nan, 0.9643001261034048, 0.9983497924348297, nan, 0.995031342981772, 0.9223532638507954]  |


### Framework versions

- Transformers 4.19.2
- Pytorch 1.7.1
- Datasets 2.2.1
- Tokenizers 0.12.1