Commit
•
199369e
1
Parent(s):
b4d6f8b
Upload README.md
Browse files
README.md
CHANGED
|
@@ -1,147 +1,206 @@
|
|
|
|
|
| 1 |
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: drone-DinoVdeau-large-2024_09_17-batch-size64_epochs100_freeze
|
| 8 |
results: []
|
| 9 |
---
|
| 10 |
|
| 11 |
-
|
| 12 |
-
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the None dataset.
|
| 17 |
-
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 0.3578
|
| 19 |
-
-
|
| 20 |
-
-
|
| 21 |
-
- Mae: 0.1288
|
| 22 |
- R2: 0.4008
|
| 23 |
-
-
|
| 24 |
-
|
|
|
|
| 25 |
|
| 26 |
-
|
|
|
|
| 27 |
|
| 28 |
-
|
| 29 |
|
| 30 |
-
|
| 31 |
|
| 32 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
|
| 34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
-
|
| 37 |
|
| 38 |
-
|
| 39 |
|
| 40 |
-
|
| 41 |
|
| 42 |
The following hyperparameters were used during training:
|
| 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 |
-
| 0.
|
| 77 |
-
| 0.
|
| 78 |
-
| 0.
|
| 79 |
-
| 0.
|
| 80 |
-
| 0.
|
| 81 |
-
| 0.
|
| 82 |
-
| 0.
|
| 83 |
-
| 0.
|
| 84 |
-
| 0.
|
| 85 |
-
| 0.
|
| 86 |
-
| 0.
|
| 87 |
-
| 0.
|
| 88 |
-
| 0.
|
| 89 |
-
| 0.
|
| 90 |
-
| 0.
|
| 91 |
-
| 0.
|
| 92 |
-
| 0.
|
| 93 |
-
| 0.
|
| 94 |
-
| 0.
|
| 95 |
-
| 0.
|
| 96 |
-
| 0.
|
| 97 |
-
| 0.
|
| 98 |
-
| 0.
|
| 99 |
-
| 0.
|
| 100 |
-
| 0.
|
| 101 |
-
| 0.
|
| 102 |
-
| 0.
|
| 103 |
-
| 0.
|
| 104 |
-
| 0.
|
| 105 |
-
| 0.
|
| 106 |
-
| 0.
|
| 107 |
-
| 0.
|
| 108 |
-
| 0.
|
| 109 |
-
| 0.
|
| 110 |
-
| 0.
|
| 111 |
-
| 0.
|
| 112 |
-
| 0.
|
| 113 |
-
| 0.
|
| 114 |
-
| 0.
|
| 115 |
-
| 0.
|
| 116 |
-
| 0.
|
| 117 |
-
| 0.
|
| 118 |
-
| 0.
|
| 119 |
-
| 0.
|
| 120 |
-
| 0.
|
| 121 |
-
| 0.
|
| 122 |
-
| 0.
|
| 123 |
-
| 0.
|
| 124 |
-
| 0.
|
| 125 |
-
| 0.
|
| 126 |
-
| 0.
|
| 127 |
-
| 0.
|
| 128 |
-
| 0.
|
| 129 |
-
| 0.
|
| 130 |
-
| 0.
|
| 131 |
-
| 0.
|
| 132 |
-
| 0.
|
| 133 |
-
| 0.
|
| 134 |
-
| 0.
|
| 135 |
-
| 0.
|
| 136 |
-
| 0.
|
| 137 |
-
| 0.
|
| 138 |
-
| 0.
|
| 139 |
-
| 0.
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
|
| 144 |
-
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
---
|
| 3 |
+
language:
|
| 4 |
+
- eng
|
| 5 |
+
license: wtfpl
|
| 6 |
tags:
|
| 7 |
+
- multilabel-image-classification
|
| 8 |
+
- multilabel
|
| 9 |
- generated_from_trainer
|
| 10 |
+
base_model: facebook/dinov2-large
|
| 11 |
model-index:
|
| 12 |
- name: drone-DinoVdeau-large-2024_09_17-batch-size64_epochs100_freeze
|
| 13 |
results: []
|
| 14 |
---
|
| 15 |
|
| 16 |
+
DinoVd'eau is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large). It achieves the following results on the test set:
|
|
|
|
| 17 |
|
| 18 |
+
- Explained variance: 0.4014
|
|
|
|
|
|
|
|
|
|
| 19 |
- Loss: 0.3578
|
| 20 |
+
- MAE: 0.1288
|
| 21 |
+
- MSE: 0.0378
|
|
|
|
| 22 |
- R2: 0.4008
|
| 23 |
+
- RMSE: 0.1943
|
| 24 |
+
|
| 25 |
+
---
|
| 26 |
|
| 27 |
+
# Model description
|
| 28 |
+
DinoVd'eau is a model built on top of dinov2 model for underwater multilabel image classification.The classification head is a combination of linear, ReLU, batch normalization, and dropout layers.
|
| 29 |
|
| 30 |
+
The source code for training the model can be found in this [Git repository](https://github.com/SeatizenDOI/DinoVdeau).
|
| 31 |
|
| 32 |
+
- **Developed by:** [lombardata](https://huggingface.co/lombardata), credits to [César Leblanc](https://huggingface.co/CesarLeblanc) and [Victor Illien](https://huggingface.co/groderg)
|
| 33 |
|
| 34 |
+
---
|
| 35 |
+
|
| 36 |
+
# Intended uses & limitations
|
| 37 |
+
You can use the raw model for classify diverse marine species, encompassing coral morphotypes classes taken from the Global Coral Reef Monitoring Network (GCRMN), habitats classes and seagrass species.
|
| 38 |
+
|
| 39 |
+
---
|
| 40 |
|
| 41 |
+
# Training and evaluation data
|
| 42 |
+
Details on the number of images for each class are given in the following table:
|
| 43 |
+
| Class | train | val | test | Total |
|
| 44 |
+
|:------------------------|--------:|------:|-------:|--------:|
|
| 45 |
+
| Acropore_branched | 1956 | 651 | 652 | 3259 |
|
| 46 |
+
| Acropore_digitised | 1717 | 576 | 576 | 2869 |
|
| 47 |
+
| Acropore_tabular | 1105 | 384 | 379 | 1868 |
|
| 48 |
+
| Algae | 11092 | 3677 | 3674 | 18443 |
|
| 49 |
+
| Dead_coral | 5888 | 1952 | 1959 | 9799 |
|
| 50 |
+
| Fish | 3453 | 1157 | 1157 | 5767 |
|
| 51 |
+
| Millepore | 1760 | 690 | 693 | 3143 |
|
| 52 |
+
| No_acropore_encrusting | 2707 | 974 | 999 | 4680 |
|
| 53 |
+
| No_acropore_massive | 6487 | 2158 | 2167 | 10812 |
|
| 54 |
+
| No_acropore_sub_massive | 5015 | 1776 | 1776 | 8567 |
|
| 55 |
+
| Rock | 11176 | 3725 | 3725 | 18626 |
|
| 56 |
+
| Rubble | 10689 | 3563 | 3563 | 17815 |
|
| 57 |
+
| Sand | 11168 | 3723 | 3723 | 18614 |
|
| 58 |
|
| 59 |
+
---
|
| 60 |
|
| 61 |
+
# Training procedure
|
| 62 |
|
| 63 |
+
## Training hyperparameters
|
| 64 |
|
| 65 |
The following hyperparameters were used during training:
|
| 66 |
+
|
| 67 |
+
- **Number of Epochs**: 100
|
| 68 |
+
- **Learning Rate**: 0.001
|
| 69 |
+
- **Train Batch Size**: 64
|
| 70 |
+
- **Eval Batch Size**: 64
|
| 71 |
+
- **Optimizer**: Adam
|
| 72 |
+
- **LR Scheduler Type**: ReduceLROnPlateau with a patience of 5 epochs and a factor of 0.1
|
| 73 |
+
- **Freeze Encoder**: Yes
|
| 74 |
+
- **Data Augmentation**: Yes
|
| 75 |
+
|
| 76 |
+
|
| 77 |
+
## Data Augmentation
|
| 78 |
+
Data were augmented using the following transformations :
|
| 79 |
+
|
| 80 |
+
Train Transforms
|
| 81 |
+
- **PreProcess**: No additional parameters
|
| 82 |
+
- **Resize**: probability=1.00
|
| 83 |
+
- **RandomHorizontalFlip**: probability=0.25
|
| 84 |
+
- **RandomVerticalFlip**: probability=0.25
|
| 85 |
+
- **ColorJiggle**: probability=0.25
|
| 86 |
+
- **RandomPerspective**: probability=0.25
|
| 87 |
+
- **Normalize**: probability=1.00
|
| 88 |
+
|
| 89 |
+
Val Transforms
|
| 90 |
+
- **PreProcess**: No additional parameters
|
| 91 |
+
- **Resize**: probability=1.00
|
| 92 |
+
- **Normalize**: probability=1.00
|
| 93 |
+
|
| 94 |
+
|
| 95 |
+
|
| 96 |
+
## Training results
|
| 97 |
+
Epoch | Explained Variance | Validation Loss | MAE | MSE | R2 | RMSE | Learning Rate
|
| 98 |
+
--- | --- | --- | --- | --- | --- | --- | ---
|
| 99 |
+
1 | 0.28 | 0.386 | 0.157 | 0.046 | 0.262 | 0.215 | 0.001
|
| 100 |
+
2 | 0.321 | 0.376 | 0.147 | 0.044 | 0.312 | 0.21 | 0.001
|
| 101 |
+
3 | 0.339 | 0.372 | 0.145 | 0.043 | 0.332 | 0.206 | 0.001
|
| 102 |
+
4 | 0.357 | 0.367 | 0.14 | 0.041 | 0.355 | 0.202 | 0.001
|
| 103 |
+
5 | 0.349 | 0.369 | 0.139 | 0.042 | 0.343 | 0.205 | 0.001
|
| 104 |
+
6 | 0.359 | 0.367 | 0.141 | 0.041 | 0.355 | 0.202 | 0.001
|
| 105 |
+
7 | 0.35 | 0.368 | 0.141 | 0.042 | 0.346 | 0.204 | 0.001
|
| 106 |
+
8 | 0.364 | 0.366 | 0.139 | 0.041 | 0.36 | 0.201 | 0.001
|
| 107 |
+
9 | 0.361 | 0.366 | 0.134 | 0.041 | 0.355 | 0.202 | 0.001
|
| 108 |
+
10 | 0.356 | 0.367 | 0.138 | 0.041 | 0.353 | 0.202 | 0.001
|
| 109 |
+
11 | 0.357 | 0.367 | 0.137 | 0.041 | 0.355 | 0.202 | 0.001
|
| 110 |
+
12 | 0.36 | 0.366 | 0.14 | 0.041 | 0.359 | 0.202 | 0.001
|
| 111 |
+
13 | 0.37 | 0.363 | 0.136 | 0.04 | 0.37 | 0.199 | 0.001
|
| 112 |
+
14 | 0.363 | 0.367 | 0.142 | 0.041 | 0.356 | 0.202 | 0.001
|
| 113 |
+
15 | 0.364 | 0.364 | 0.14 | 0.04 | 0.362 | 0.201 | 0.001
|
| 114 |
+
16 | 0.372 | 0.364 | 0.136 | 0.04 | 0.369 | 0.2 | 0.001
|
| 115 |
+
17 | 0.373 | 0.367 | 0.141 | 0.041 | 0.362 | 0.202 | 0.001
|
| 116 |
+
18 | 0.371 | 0.363 | 0.137 | 0.04 | 0.37 | 0.2 | 0.001
|
| 117 |
+
19 | 0.373 | 0.363 | 0.135 | 0.04 | 0.372 | 0.199 | 0.001
|
| 118 |
+
20 | 0.362 | 0.365 | 0.135 | 0.041 | 0.359 | 0.201 | 0.001
|
| 119 |
+
21 | 0.363 | 0.367 | 0.136 | 0.041 | 0.358 | 0.202 | 0.001
|
| 120 |
+
22 | 0.37 | 0.365 | 0.137 | 0.04 | 0.368 | 0.2 | 0.001
|
| 121 |
+
23 | 0.374 | 0.363 | 0.136 | 0.04 | 0.37 | 0.2 | 0.001
|
| 122 |
+
24 | 0.376 | 0.363 | 0.139 | 0.04 | 0.373 | 0.199 | 0.001
|
| 123 |
+
25 | 0.373 | 0.364 | 0.138 | 0.04 | 0.37 | 0.2 | 0.001
|
| 124 |
+
26 | 0.384 | 0.361 | 0.133 | 0.039 | 0.382 | 0.198 | 0.0001
|
| 125 |
+
27 | 0.388 | 0.36 | 0.135 | 0.039 | 0.386 | 0.197 | 0.0001
|
| 126 |
+
28 | 0.39 | 0.359 | 0.134 | 0.038 | 0.389 | 0.196 | 0.0001
|
| 127 |
+
29 | 0.391 | 0.36 | 0.135 | 0.038 | 0.389 | 0.196 | 0.0001
|
| 128 |
+
30 | 0.389 | 0.36 | 0.135 | 0.039 | 0.388 | 0.197 | 0.0001
|
| 129 |
+
31 | 0.392 | 0.359 | 0.132 | 0.038 | 0.391 | 0.196 | 0.0001
|
| 130 |
+
32 | 0.393 | 0.358 | 0.133 | 0.038 | 0.393 | 0.196 | 0.0001
|
| 131 |
+
33 | 0.395 | 0.358 | 0.131 | 0.038 | 0.395 | 0.195 | 0.0001
|
| 132 |
+
34 | 0.397 | 0.358 | 0.132 | 0.038 | 0.395 | 0.195 | 0.0001
|
| 133 |
+
35 | 0.395 | 0.358 | 0.132 | 0.038 | 0.395 | 0.195 | 0.0001
|
| 134 |
+
36 | 0.39 | 0.359 | 0.135 | 0.039 | 0.39 | 0.196 | 0.0001
|
| 135 |
+
37 | 0.397 | 0.358 | 0.131 | 0.038 | 0.397 | 0.195 | 0.0001
|
| 136 |
+
38 | 0.394 | 0.358 | 0.133 | 0.038 | 0.392 | 0.196 | 0.0001
|
| 137 |
+
39 | 0.397 | 0.358 | 0.131 | 0.038 | 0.396 | 0.195 | 0.0001
|
| 138 |
+
40 | 0.4 | 0.357 | 0.133 | 0.038 | 0.398 | 0.195 | 0.0001
|
| 139 |
+
41 | 0.399 | 0.358 | 0.132 | 0.038 | 0.396 | 0.195 | 0.0001
|
| 140 |
+
42 | 0.399 | 0.357 | 0.133 | 0.038 | 0.397 | 0.195 | 0.0001
|
| 141 |
+
43 | 0.402 | 0.357 | 0.133 | 0.038 | 0.401 | 0.194 | 0.0001
|
| 142 |
+
44 | 0.403 | 0.357 | 0.131 | 0.038 | 0.401 | 0.194 | 0.0001
|
| 143 |
+
45 | 0.403 | 0.357 | 0.132 | 0.038 | 0.402 | 0.194 | 0.0001
|
| 144 |
+
46 | 0.401 | 0.357 | 0.13 | 0.038 | 0.4 | 0.194 | 0.0001
|
| 145 |
+
47 | 0.4 | 0.357 | 0.129 | 0.038 | 0.397 | 0.195 | 0.0001
|
| 146 |
+
48 | 0.404 | 0.356 | 0.13 | 0.038 | 0.402 | 0.194 | 0.0001
|
| 147 |
+
49 | 0.402 | 0.357 | 0.131 | 0.038 | 0.401 | 0.194 | 0.0001
|
| 148 |
+
50 | 0.401 | 0.357 | 0.132 | 0.038 | 0.4 | 0.194 | 0.0001
|
| 149 |
+
51 | 0.402 | 0.358 | 0.134 | 0.038 | 0.396 | 0.195 | 0.0001
|
| 150 |
+
52 | 0.405 | 0.356 | 0.131 | 0.037 | 0.404 | 0.194 | 0.0001
|
| 151 |
+
53 | 0.405 | 0.357 | 0.131 | 0.038 | 0.403 | 0.194 | 0.0001
|
| 152 |
+
54 | 0.402 | 0.357 | 0.132 | 0.038 | 0.401 | 0.194 | 0.0001
|
| 153 |
+
55 | 0.405 | 0.356 | 0.129 | 0.038 | 0.403 | 0.194 | 0.0001
|
| 154 |
+
56 | 0.405 | 0.357 | 0.128 | 0.038 | 0.402 | 0.194 | 0.0001
|
| 155 |
+
57 | 0.405 | 0.356 | 0.129 | 0.038 | 0.403 | 0.194 | 0.0001
|
| 156 |
+
58 | 0.406 | 0.356 | 0.13 | 0.038 | 0.404 | 0.194 | 0.0001
|
| 157 |
+
59 | 0.406 | 0.356 | 0.129 | 0.037 | 0.405 | 0.194 | 1e-05
|
| 158 |
+
60 | 0.408 | 0.356 | 0.13 | 0.037 | 0.406 | 0.193 | 1e-05
|
| 159 |
+
61 | 0.407 | 0.355 | 0.13 | 0.037 | 0.407 | 0.193 | 1e-05
|
| 160 |
+
62 | 0.406 | 0.356 | 0.132 | 0.038 | 0.404 | 0.194 | 1e-05
|
| 161 |
+
63 | 0.409 | 0.356 | 0.129 | 0.037 | 0.408 | 0.193 | 1e-05
|
| 162 |
+
64 | 0.409 | 0.355 | 0.13 | 0.037 | 0.408 | 0.193 | 1e-05
|
| 163 |
+
65 | 0.406 | 0.356 | 0.131 | 0.038 | 0.405 | 0.194 | 1e-05
|
| 164 |
+
66 | 0.409 | 0.355 | 0.13 | 0.037 | 0.408 | 0.193 | 1e-05
|
| 165 |
+
67 | 0.408 | 0.355 | 0.13 | 0.037 | 0.408 | 0.193 | 1e-05
|
| 166 |
+
68 | 0.407 | 0.356 | 0.13 | 0.037 | 0.406 | 0.193 | 1e-05
|
| 167 |
+
69 | 0.409 | 0.355 | 0.13 | 0.037 | 0.408 | 0.193 | 1e-05
|
| 168 |
+
70 | 0.409 | 0.356 | 0.131 | 0.037 | 0.407 | 0.193 | 1e-05
|
| 169 |
+
71 | 0.407 | 0.356 | 0.13 | 0.037 | 0.407 | 0.193 | 1e-05
|
| 170 |
+
72 | 0.408 | 0.356 | 0.13 | 0.037 | 0.407 | 0.193 | 1e-05
|
| 171 |
+
73 | 0.409 | 0.355 | 0.13 | 0.037 | 0.408 | 0.193 | 1.0000000000000002e-06
|
| 172 |
+
74 | 0.409 | 0.355 | 0.128 | 0.037 | 0.409 | 0.193 | 1.0000000000000002e-06
|
| 173 |
+
75 | 0.406 | 0.356 | 0.13 | 0.037 | 0.405 | 0.194 | 1.0000000000000002e-06
|
| 174 |
+
76 | 0.408 | 0.356 | 0.128 | 0.037 | 0.406 | 0.193 | 1.0000000000000002e-06
|
| 175 |
+
77 | 0.405 | 0.356 | 0.132 | 0.038 | 0.404 | 0.194 | 1.0000000000000002e-06
|
| 176 |
+
78 | 0.409 | 0.355 | 0.131 | 0.037 | 0.409 | 0.193 | 1.0000000000000002e-06
|
| 177 |
+
79 | 0.402 | 0.357 | 0.131 | 0.038 | 0.4 | 0.195 | 1.0000000000000002e-06
|
| 178 |
+
80 | 0.406 | 0.356 | 0.131 | 0.037 | 0.405 | 0.194 | 1.0000000000000002e-06
|
| 179 |
+
81 | 0.409 | 0.356 | 0.131 | 0.037 | 0.408 | 0.193 | 1.0000000000000002e-07
|
| 180 |
+
82 | 0.409 | 0.356 | 0.131 | 0.037 | 0.407 | 0.193 | 1.0000000000000002e-07
|
| 181 |
+
83 | 0.41 | 0.356 | 0.13 | 0.037 | 0.407 | 0.193 | 1.0000000000000002e-07
|
| 182 |
+
84 | 0.408 | 0.356 | 0.131 | 0.037 | 0.406 | 0.193 | 1.0000000000000002e-07
|
| 183 |
+
|
| 184 |
+
|
| 185 |
+
---
|
| 186 |
+
|
| 187 |
+
# CO2 Emissions
|
| 188 |
+
|
| 189 |
+
The estimated CO2 emissions for training this model are documented below:
|
| 190 |
+
|
| 191 |
+
- **Emissions**: 0.22861184690098074 grams of CO2
|
| 192 |
+
- **Source**: Code Carbon
|
| 193 |
+
- **Training Type**: fine-tuning
|
| 194 |
+
- **Geographical Location**: Brest, France
|
| 195 |
+
- **Hardware Used**: NVIDIA Tesla V100 PCIe 32 Go
|
| 196 |
+
|
| 197 |
+
|
| 198 |
+
---
|
| 199 |
+
|
| 200 |
+
# Framework Versions
|
| 201 |
+
|
| 202 |
+
- **Transformers**: 4.41.1
|
| 203 |
+
- **Pytorch**: 2.3.0+cu121
|
| 204 |
+
- **Datasets**: 2.19.1
|
| 205 |
+
- **Tokenizers**: 0.19.1
|
| 206 |
+
|