File size: 4,991 Bytes
a9e388f
ab0e089
 
 
 
 
 
 
 
 
 
 
a9e388f
 
ab0e089
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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: other
base_model: nvidia/segformer-b0-finetuned-ade-512-512
tags:
- vision
- image-segmentation
- generated_from_trainer
metrics:
- precision
model-index:
- name: segformer-b0-finetuned-segments-pv_v1_3x_normalized_p100_4batch
  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. -->

[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/mouadn773/huggingface/runs/yy31wgdz)
# segformer-b0-finetuned-segments-pv_v1_3x_normalized_p100_4batch

This model is a fine-tuned version of [nvidia/segformer-b0-finetuned-ade-512-512](https://huggingface.co/nvidia/segformer-b0-finetuned-ade-512-512) on the mouadenna/satellite_PV_dataset_train_test_v1 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0067
- Mean Iou: 0.8641
- Precision: 0.9173

## 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: 0.0004
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.001
- num_epochs: 40
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step  | Validation Loss | Mean Iou | Precision |
|:-------------:|:-------:|:-----:|:---------------:|:--------:|:---------:|
| 0.0077        | 0.9993  | 687   | 0.0077          | 0.7897   | 0.8235    |
| 0.0056        | 2.0     | 1375  | 0.0059          | 0.8193   | 0.8760    |
| 0.0065        | 2.9993  | 2062  | 0.0064          | 0.8222   | 0.9068    |
| 0.0047        | 4.0     | 2750  | 0.0061          | 0.8195   | 0.9299    |
| 0.0039        | 4.9993  | 3437  | 0.0055          | 0.8440   | 0.9075    |
| 0.0044        | 6.0     | 4125  | 0.0063          | 0.8208   | 0.8479    |
| 0.0034        | 6.9993  | 4812  | 0.0080          | 0.7750   | 0.8153    |
| 0.0037        | 8.0     | 5500  | 0.0053          | 0.8475   | 0.9084    |
| 0.004         | 8.9993  | 6187  | 0.0073          | 0.8013   | 0.8237    |
| 0.003         | 10.0    | 6875  | 0.0056          | 0.8476   | 0.8955    |
| 0.0038        | 10.9993 | 7562  | 0.0058          | 0.8273   | 0.9144    |
| 0.0028        | 12.0    | 8250  | 0.0065          | 0.8143   | 0.8888    |
| 0.0031        | 12.9993 | 8937  | 0.0064          | 0.8175   | 0.9188    |
| 0.003         | 14.0    | 9625  | 0.0051          | 0.8491   | 0.9027    |
| 0.0025        | 14.9993 | 10312 | 0.0059          | 0.8558   | 0.9085    |
| 0.0029        | 16.0    | 11000 | 0.0057          | 0.8454   | 0.9029    |
| 0.0026        | 16.9993 | 11687 | 0.0057          | 0.8547   | 0.9230    |
| 0.0024        | 18.0    | 12375 | 0.0059          | 0.8579   | 0.9045    |
| 0.0025        | 18.9993 | 13062 | 0.0059          | 0.8645   | 0.9094    |
| 0.0025        | 20.0    | 13750 | 0.0059          | 0.8498   | 0.9174    |
| 0.0024        | 20.9993 | 14437 | 0.0056          | 0.8576   | 0.8970    |
| 0.0022        | 22.0    | 15125 | 0.0063          | 0.8541   | 0.8952    |
| 0.0031        | 22.9993 | 15812 | 0.0054          | 0.8508   | 0.9154    |
| 0.0021        | 24.0    | 16500 | 0.0057          | 0.8545   | 0.9119    |
| 0.0022        | 24.9993 | 17187 | 0.0058          | 0.8474   | 0.9149    |
| 0.0022        | 26.0    | 17875 | 0.0066          | 0.8325   | 0.8879    |
| 0.0021        | 26.9993 | 18562 | 0.0062          | 0.8522   | 0.9156    |
| 0.0021        | 28.0    | 19250 | 0.0063          | 0.8488   | 0.8932    |
| 0.002         | 28.9993 | 19937 | 0.0061          | 0.8579   | 0.9200    |
| 0.002         | 30.0    | 20625 | 0.0059          | 0.8624   | 0.9182    |
| 0.0021        | 30.9993 | 21312 | 0.0061          | 0.8564   | 0.9013    |
| 0.0019        | 32.0    | 22000 | 0.0060          | 0.8601   | 0.9091    |
| 0.0018        | 32.9993 | 22687 | 0.0059          | 0.8640   | 0.9163    |
| 0.0017        | 34.0    | 23375 | 0.0062          | 0.8622   | 0.9187    |
| 0.0017        | 34.9993 | 24062 | 0.0062          | 0.8634   | 0.9245    |
| 0.0017        | 36.0    | 24750 | 0.0064          | 0.8655   | 0.9196    |
| 0.0017        | 36.9993 | 25437 | 0.0063          | 0.8642   | 0.9197    |
| 0.0016        | 38.0    | 26125 | 0.0065          | 0.8634   | 0.9166    |
| 0.0016        | 38.9993 | 26812 | 0.0067          | 0.8639   | 0.9186    |
| 0.0016        | 39.9709 | 27480 | 0.0067          | 0.8641   | 0.9173    |


### Framework versions

- Transformers 4.42.3
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1