Commit
•
2703fc1
1
Parent(s):
6ee0c53
Upload README.md
Browse files
README.md
CHANGED
|
@@ -1,130 +1,196 @@
|
|
|
|
|
| 1 |
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs
|
| 8 |
results: []
|
| 9 |
---
|
| 10 |
|
| 11 |
-
|
| 12 |
-
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
|
| 14 |
-
# drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs
|
| 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: 0.3499
|
| 19 |
-
-
|
| 20 |
-
-
|
|
|
|
|
|
|
|
|
|
| 21 |
- R2: 0.4361
|
| 22 |
-
- Explained Variance: 0.4376
|
| 23 |
-
- Learning Rate: 1e-05
|
| 24 |
|
| 25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
|
|
|
| 30 |
|
| 31 |
-
|
| 32 |
|
| 33 |
-
|
|
|
|
|
|
|
| 34 |
|
| 35 |
-
|
|
|
|
| 36 |
|
| 37 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
|
| 39 |
-
|
|
|
|
|
|
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 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 |
-
| 0.
|
| 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 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
| 130 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
---
|
| 3 |
+
language:
|
| 4 |
+
- eng
|
| 5 |
+
license: cc0-1.0
|
| 6 |
tags:
|
| 7 |
+
- multilabel-image-classification
|
| 8 |
+
- multilabel
|
| 9 |
- generated_from_trainer
|
| 10 |
+
base_model: drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs
|
| 11 |
model-index:
|
| 12 |
- name: drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs
|
| 13 |
results: []
|
| 14 |
---
|
| 15 |
|
| 16 |
+
drone-DinoVdeau-produttoria_binary-probabilities is a fine-tuned version of [drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs](https://huggingface.co/drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs). It achieves the following results on the test set:
|
|
|
|
| 17 |
|
|
|
|
| 18 |
|
|
|
|
|
|
|
| 19 |
- Loss: 0.3499
|
| 20 |
+
- F1 Micro: 0.8642
|
| 21 |
+
- F1 Macro: 0.7488
|
| 22 |
+
- Accuracy: 0.1948
|
| 23 |
+
- RMSE: 0.1848
|
| 24 |
+
- MAE: 0.1248
|
| 25 |
- R2: 0.4361
|
|
|
|
|
|
|
| 26 |
|
| 27 |
+
| Class | F1 per class |
|
| 28 |
+
|----------|-------|
|
| 29 |
+
| Acropore_branched | 0.7745 |
|
| 30 |
+
| Acropore_digitised | 0.5393 |
|
| 31 |
+
| Acropore_tabular | 0.4310 |
|
| 32 |
+
| Algae | 0.9823 |
|
| 33 |
+
| Dead_coral | 0.7850 |
|
| 34 |
+
| Fish | 0.6550 |
|
| 35 |
+
| Millepore | 0.4152 |
|
| 36 |
+
| No_acropore_encrusting | 0.6755 |
|
| 37 |
+
| No_acropore_massive | 0.7954 |
|
| 38 |
+
| No_acropore_sub_massive | 0.7421 |
|
| 39 |
+
| Rock | 0.9888 |
|
| 40 |
+
| Rubble | 0.9645 |
|
| 41 |
+
| Sand | 0.9853 |
|
| 42 |
+
|
| 43 |
|
| 44 |
+
---
|
| 45 |
|
| 46 |
+
# Model description
|
| 47 |
+
drone-DinoVdeau-produttoria_binary-probabilities is a model built on top of drone-DinoVdeau-produttoria_binary-probabilities-large-2024_11_03-batch-size64_freeze_probs model for underwater multilabel image classification.The classification head is a combination of linear, ReLU, batch normalization, and dropout layers.
|
| 48 |
|
| 49 |
+
The source code for training the model can be found in this [Git repository](https://github.com/SeatizenDOI/DinoVdeau).
|
| 50 |
|
| 51 |
+
- **Developed by:** [lombardata](https://huggingface.co/lombardata), credits to [César Leblanc](https://huggingface.co/CesarLeblanc) and [Victor Illien](https://huggingface.co/groderg)
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
|
| 55 |
+
# Intended uses & limitations
|
| 56 |
+
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.
|
| 57 |
|
| 58 |
+
---
|
| 59 |
+
|
| 60 |
+
# Training and evaluation data
|
| 61 |
+
Details on the estimated number of images for each class are given in the following table:
|
| 62 |
+
| Class | train | test | val | Total |
|
| 63 |
+
|:------------------------|--------:|-------:|------:|--------:|
|
| 64 |
+
| Acropore_branched | 1282 | 458 | 480 | 2220 |
|
| 65 |
+
| Acropore_digitised | 800 | 276 | 253 | 1329 |
|
| 66 |
+
| Acropore_tabular | 377 | 133 | 135 | 645 |
|
| 67 |
+
| Algae | 9284 | 3133 | 3167 | 15584 |
|
| 68 |
+
| Dead_coral | 3207 | 1097 | 1103 | 5407 |
|
| 69 |
+
| Fish | 1360 | 481 | 487 | 2328 |
|
| 70 |
+
| Millepore | 258 | 110 | 94 | 462 |
|
| 71 |
+
| No_acropore_encrusting | 981 | 403 | 407 | 1791 |
|
| 72 |
+
| No_acropore_massive | 3261 | 1225 | 1291 | 5777 |
|
| 73 |
+
| No_acropore_sub_massive | 2371 | 857 | 889 | 4117 |
|
| 74 |
+
| Rock | 10077 | 3388 | 3402 | 16867 |
|
| 75 |
+
| Rubble | 8590 | 2880 | 2878 | 14348 |
|
| 76 |
+
| Sand | 9880 | 3283 | 3311 | 16474 |
|
| 77 |
+
|
| 78 |
+
---
|
| 79 |
|
| 80 |
+
# Training procedure
|
| 81 |
+
|
| 82 |
+
## Training hyperparameters
|
| 83 |
|
| 84 |
The following hyperparameters were used during training:
|
| 85 |
+
|
| 86 |
+
- **Number of Epochs**: 68.0
|
| 87 |
+
- **Learning Rate**: 0.001
|
| 88 |
+
- **Train Batch Size**: 64
|
| 89 |
+
- **Eval Batch Size**: 64
|
| 90 |
+
- **Optimizer**: Adam
|
| 91 |
+
- **LR Scheduler Type**: ReduceLROnPlateau with a patience of 5 epochs and a factor of 0.1
|
| 92 |
+
- **Freeze Encoder**: Yes
|
| 93 |
+
- **Data Augmentation**: Yes
|
| 94 |
+
|
| 95 |
+
|
| 96 |
+
## Data Augmentation
|
| 97 |
+
Data were augmented using the following transformations :
|
| 98 |
+
|
| 99 |
+
Train Transforms
|
| 100 |
+
- **PreProcess**: No additional parameters
|
| 101 |
+
- **Resize**: probability=1.00
|
| 102 |
+
- **RandomHorizontalFlip**: probability=0.25
|
| 103 |
+
- **RandomVerticalFlip**: probability=0.25
|
| 104 |
+
- **ColorJiggle**: probability=0.25
|
| 105 |
+
- **RandomPerspective**: probability=0.25
|
| 106 |
+
- **Normalize**: probability=1.00
|
| 107 |
+
|
| 108 |
+
Val Transforms
|
| 109 |
+
- **PreProcess**: No additional parameters
|
| 110 |
+
- **Resize**: probability=1.00
|
| 111 |
+
- **Normalize**: probability=1.00
|
| 112 |
+
|
| 113 |
+
|
| 114 |
+
|
| 115 |
+
## Training results
|
| 116 |
+
Epoch | Validation Loss | MAE | RMSE | R2 | Learning Rate
|
| 117 |
+
--- | --- | --- | --- | --- | ---
|
| 118 |
+
1 | 0.3794757127761841 | 0.14892756938934326 | 0.20668388903141022 | 0.2894136607646942 | 0.001
|
| 119 |
+
2 | 0.3673744201660156 | 0.1374034583568573 | 0.198309525847435 | 0.35173478722572327 | 0.001
|
| 120 |
+
3 | 0.3671453297138214 | 0.1413687765598297 | 0.198079913854599 | 0.35207054018974304 | 0.001
|
| 121 |
+
4 | 0.36317145824432373 | 0.1391323208808899 | 0.19521364569664001 | 0.3708474040031433 | 0.001
|
| 122 |
+
5 | 0.3678734302520752 | 0.141828715801239 | 0.19933271408081055 | 0.3453221321105957 | 0.001
|
| 123 |
+
6 | 0.36250030994415283 | 0.13803647458553314 | 0.19508354365825653 | 0.37177884578704834 | 0.001
|
| 124 |
+
7 | 0.36188462376594543 | 0.1347939670085907 | 0.19410446286201477 | 0.3771490454673767 | 0.001
|
| 125 |
+
8 | 0.36125126481056213 | 0.13681310415267944 | 0.19354337453842163 | 0.37883251905441284 | 0.001
|
| 126 |
+
9 | 0.3603876233100891 | 0.135352224111557 | 0.19339486956596375 | 0.3812035918235779 | 0.001
|
| 127 |
+
10 | 0.3612792491912842 | 0.13375206291675568 | 0.19321060180664062 | 0.38122493028640747 | 0.001
|
| 128 |
+
11 | 0.3603772521018982 | 0.13226205110549927 | 0.19312407076358795 | 0.3844555616378784 | 0.001
|
| 129 |
+
12 | 0.361823707818985 | 0.13859649002552032 | 0.1941623091697693 | 0.3774065375328064 | 0.001
|
| 130 |
+
13 | 0.35931822657585144 | 0.13433586061000824 | 0.1924724578857422 | 0.3875495195388794 | 0.001
|
| 131 |
+
14 | 0.3604746460914612 | 0.13521355390548706 | 0.19319292902946472 | 0.38306838274002075 | 0.001
|
| 132 |
+
15 | 0.36050480604171753 | 0.13660094141960144 | 0.19354429841041565 | 0.3816676735877991 | 0.001
|
| 133 |
+
16 | 0.3599933683872223 | 0.13121920824050903 | 0.1921611875295639 | 0.388213574886322 | 0.001
|
| 134 |
+
17 | 0.3628774583339691 | 0.13783428072929382 | 0.1932491511106491 | 0.38425371050834656 | 0.001
|
| 135 |
+
18 | 0.36154037714004517 | 0.1323489546775818 | 0.19430074095726013 | 0.37684857845306396 | 0.001
|
| 136 |
+
19 | 0.3594801127910614 | 0.13296250998973846 | 0.19218452274799347 | 0.38950252532958984 | 0.001
|
| 137 |
+
20 | 0.3565874397754669 | 0.13302744925022125 | 0.19017010927200317 | 0.40064936876296997 | 0.0001
|
| 138 |
+
21 | 0.35486647486686707 | 0.13062793016433716 | 0.18895885348320007 | 0.40758493542671204 | 0.0001
|
| 139 |
+
22 | 0.35447388887405396 | 0.13081100583076477 | 0.18863680958747864 | 0.4096067547798157 | 0.0001
|
| 140 |
+
23 | 0.3544616997241974 | 0.13033078610897064 | 0.18823565542697906 | 0.411631315946579 | 0.0001
|
| 141 |
+
24 | 0.3539991080760956 | 0.1316699981689453 | 0.1881898045539856 | 0.4120980501174927 | 0.0001
|
| 142 |
+
25 | 0.3545873463153839 | 0.12844440340995789 | 0.18831981718540192 | 0.41126883029937744 | 0.0001
|
| 143 |
+
26 | 0.3529074192047119 | 0.1263934224843979 | 0.18757320940494537 | 0.4154190421104431 | 0.0001
|
| 144 |
+
27 | 0.3532767593860626 | 0.129387766122818 | 0.187411367893219 | 0.41658732295036316 | 0.0001
|
| 145 |
+
28 | 0.3532498776912689 | 0.12938687205314636 | 0.18755248188972473 | 0.41600102186203003 | 0.0001
|
| 146 |
+
29 | 0.35306474566459656 | 0.1302015781402588 | 0.1871432662010193 | 0.41835859417915344 | 0.0001
|
| 147 |
+
30 | 0.3536038398742676 | 0.1291646808385849 | 0.18775980174541473 | 0.414754718542099 | 0.0001
|
| 148 |
+
31 | 0.3530591130256653 | 0.1267225444316864 | 0.1876552253961563 | 0.415239542722702 | 0.0001
|
| 149 |
+
32 | 0.3528367877006531 | 0.12877780199050903 | 0.18764065206050873 | 0.4161965548992157 | 0.0001
|
| 150 |
+
33 | 0.35152381658554077 | 0.12729588150978088 | 0.18640562891960144 | 0.4225224256515503 | 0.0001
|
| 151 |
+
34 | 0.35195404291152954 | 0.12629321217536926 | 0.18677598237991333 | 0.4202421009540558 | 0.0001
|
| 152 |
+
35 | 0.35178276896476746 | 0.12782610952854156 | 0.18657900393009186 | 0.42147499322891235 | 0.0001
|
| 153 |
+
36 | 0.35231974720954895 | 0.12849368155002594 | 0.18713095784187317 | 0.4192589223384857 | 0.0001
|
| 154 |
+
37 | 0.3515876829624176 | 0.12726719677448273 | 0.18659605085849762 | 0.4216739237308502 | 0.0001
|
| 155 |
+
38 | 0.35274896025657654 | 0.12742024660110474 | 0.1878250390291214 | 0.41570571064949036 | 0.0001
|
| 156 |
+
39 | 0.35124146938323975 | 0.12662582099437714 | 0.18624022603034973 | 0.4241558611392975 | 0.0001
|
| 157 |
+
40 | 0.35209622979164124 | 0.13019172847270966 | 0.18663105368614197 | 0.4224165081977844 | 0.0001
|
| 158 |
+
41 | 0.35067644715309143 | 0.1265629082918167 | 0.18584123253822327 | 0.42641735076904297 | 0.0001
|
| 159 |
+
42 | 0.3512935936450958 | 0.12775851786136627 | 0.18596960604190826 | 0.4262687563896179 | 0.0001
|
| 160 |
+
43 | 0.3510710895061493 | 0.12741515040397644 | 0.1859511435031891 | 0.42624664306640625 | 0.0001
|
| 161 |
+
44 | 0.35139599442481995 | 0.12441141903400421 | 0.18587811291217804 | 0.4265681505203247 | 0.0001
|
| 162 |
+
45 | 0.35247302055358887 | 0.127328023314476 | 0.18626871705055237 | 0.42492759227752686 | 0.0001
|
| 163 |
+
46 | 0.3505423069000244 | 0.12581512331962585 | 0.18559609353542328 | 0.4274958372116089 | 0.0001
|
| 164 |
+
47 | 0.3517468571662903 | 0.1250177025794983 | 0.18658187985420227 | 0.4231443405151367 | 0.0001
|
| 165 |
+
48 | 0.35043978691101074 | 0.12591718137264252 | 0.18564504384994507 | 0.42857199907302856 | 0.0001
|
| 166 |
+
49 | 0.35074281692504883 | 0.12717720866203308 | 0.1857146918773651 | 0.4274061322212219 | 0.0001
|
| 167 |
+
50 | 0.3515849709510803 | 0.1283276230096817 | 0.1857057511806488 | 0.42797213792800903 | 0.0001
|
| 168 |
+
51 | 0.35289809107780457 | 0.1288221776485443 | 0.1866857409477234 | 0.42265036702156067 | 0.0001
|
| 169 |
+
52 | 0.3505743443965912 | 0.12677451968193054 | 0.18569740653038025 | 0.4281761944293976 | 0.0001
|
| 170 |
+
53 | 0.35052910447120667 | 0.1273086667060852 | 0.18561594188213348 | 0.4285990595817566 | 0.0001
|
| 171 |
+
54 | 0.35016006231307983 | 0.12655657529830933 | 0.1853920817375183 | 0.4299810826778412 | 0.0001
|
| 172 |
+
55 | 0.35006165504455566 | 0.12509843707084656 | 0.18541744351387024 | 0.42986157536506653 | 0.0001
|
| 173 |
+
56 | 0.35072585940361023 | 0.12430255115032196 | 0.18582786619663239 | 0.4273306131362915 | 0.0001
|
| 174 |
+
57 | 0.3508891463279724 | 0.12534378468990326 | 0.18598994612693787 | 0.4273567497730255 | 0.0001
|
| 175 |
+
58 | 0.3492669463157654 | 0.12510134279727936 | 0.1846422404050827 | 0.4338167607784271 | 0.0001
|
| 176 |
+
59 | 0.3500733971595764 | 0.12414979934692383 | 0.18548892438411713 | 0.42818644642829895 | 0.0001
|
| 177 |
+
60 | 0.350059449672699 | 0.12591439485549927 | 0.18521927297115326 | 0.43032628297805786 | 0.0001
|
| 178 |
+
61 | 0.34978389739990234 | 0.126389279961586 | 0.18503333628177643 | 0.43050628900527954 | 0.0001
|
| 179 |
+
62 | 0.34984564781188965 | 0.1265084147453308 | 0.18499605357646942 | 0.4322562515735626 | 0.0001
|
| 180 |
+
63 | 0.35018646717071533 | 0.1270289421081543 | 0.18513011932373047 | 0.4321424067020416 | 0.0001
|
| 181 |
+
64 | 0.34996479749679565 | 0.12558279931545258 | 0.1853456199169159 | 0.43004974722862244 | 0.0001
|
| 182 |
+
65 | 0.3501463234424591 | 0.1280103474855423 | 0.1853969395160675 | 0.42989540100097656 | 1e-05
|
| 183 |
+
66 | 0.34930846095085144 | 0.1253172904253006 | 0.1847212016582489 | 0.43362313508987427 | 1e-05
|
| 184 |
+
67 | 0.3493542969226837 | 0.12613731622695923 | 0.18472003936767578 | 0.43344247341156006 | 1e-05
|
| 185 |
+
68 | 0.3500206172466278 | 0.12607118487358093 | 0.18558326363563538 | 0.4291488826274872 | 1e-05
|
| 186 |
+
|
| 187 |
+
|
| 188 |
+
---
|
| 189 |
+
|
| 190 |
+
# Framework Versions
|
| 191 |
+
|
| 192 |
+
- **Transformers**: 4.41.0
|
| 193 |
+
- **Pytorch**: 2.5.0+cu124
|
| 194 |
+
- **Datasets**: 3.0.2
|
| 195 |
+
- **Tokenizers**: 0.19.1
|
| 196 |
+
|