Upload README.md with huggingface_hub

README.md

@@ -2,13 +2,13 @@
 language:
 - eng
 license: apache-2.0
-base_model: facebook/dinov2-large
 tags:
 - multilabel-image-classification
 - multilabel
 - generated_from_trainer
 metrics:
 - accuracy
+base_model: facebook/dinov2-large
 model-index:
 - name: DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze
   results: []
@@ -19,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # DinoVdeau-large-2024_04_03-with_data_aug_batch-size32_epochs150_freeze
 
-This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the multilabel_complete_dataset dataset.
+DinoVd'eau is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the multilabel_complete_dataset dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.1181
 - F1 Micro: 0.8219
@@ -30,18 +30,71 @@ It achieves the following results on the evaluation set:
 
 ## Model description
 
-More information needed
+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.
+- **Developed by:** [lombardata](https://huggingface.co/lombardata), credits to [César Leblanc](https://huggingface.co/CesarLeblanc)
 
 ## Intended uses & limitations
 
-More information needed
+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.
 
 ## Training and evaluation data
 
-More information needed
+Details on the number of images for each class are given in the following table:
+ |  |train |val |test |Total |
+ |--- | --- | --- | --- | --- | 
+ | Acropore_branched | 1504 | 445 | 430 | 2379 | 
+ | Acropore_digitised | 593 | 151 | 144 | 888 | 
+ | Acropore_sub_massive | 148 | 54 | 41 | 243 | 
+ | Acropore_tabular | 1012 | 290 | 287 | 1589 | 
+ | Algae_assembly | 2545 | 858 | 835 | 4238 | 
+ | Algae_drawn_up | 376 | 123 | 121 | 620 | 
+ | Algae_limestone | 1652 | 561 | 559 | 2772 | 
+ | Algae_sodding | 3094 | 1011 | 1012 | 5117 | 
+ | Atra/Leucospilota | 1081 | 352 | 359 | 1792 | 
+ | Bleached_coral | 220 | 70 | 70 | 360 | 
+ | Blurred | 192 | 62 | 66 | 320 | 
+ | Dead_coral | 2001 | 637 | 626 | 3264 | 
+ | Fish | 2068 | 611 | 642 | 3321 | 
+ | Homo_sapiens | 162 | 60 | 60 | 282 | 
+ | Human_object | 157 | 60 | 53 | 270 | 
+ | Living_coral | 147 | 56 | 47 | 250 | 
+ | Millepore | 378 | 131 | 128 | 637 | 
+ | No_acropore_encrusting | 422 | 152 | 151 | 725 | 
+ | No_acropore_foliaceous | 200 | 46 | 40 | 286 | 
+ | No_acropore_massive | 1033 | 337 | 335 | 1705 | 
+ | No_acropore_solitary | 193 | 56 | 54 | 303 | 
+ | No_acropore_sub_massive | 1412 | 418 | 426 | 2256 | 
+ | Rock | 4487 | 1481 | 1489 | 7457 | 
+ | Sand | 5806 | 1959 | 1954 | 9719 | 
+ | Scrap | 3063 | 1030 | 1030 | 5123 | 
+ | Sea_cucumber | 1396 | 453 | 445 | 2294 | 
+ | Sea_urchins | 319 | 122 | 104 | 545 | 
+ | Sponge | 273 | 107 | 90 | 470 | 
+ | Syringodium_isoetifolium | 1198 | 399 | 398 | 1995 | 
+ | Thalassodendron_ciliatum | 781 | 260 | 262 | 1303 | 
+ | Useless | 579 | 193 | 193 | 965 | 
+
 
 ## Training procedure
 
+### Data Augmentation
+
+Data were augmented using the following transformations :
+- training transformations : Sequential(
+  (0): PreProcess()
+  (1): Resize(output_size=(518, 518), p=1.0, p_batch=1.0, same_on_batch=True, size=(518, 518), side=short, resample=bilinear, align_corners=True, antialias=False)
+  (2): RandomHorizontalFlip(p=0.25, p_batch=1.0, same_on_batch=False)
+  (3): RandomVerticalFlip(p=0.25, p_batch=1.0, same_on_batch=False)
+  (4): ColorJiggle(brightness=0.0, contrast=0.0, saturation=0.0, hue=0.0, p=0.25, p_batch=1.0, same_on_batch=False)
+  (5): RandomPerspective(distortion_scale=0.5, p=0.25, p_batch=1.0, same_on_batch=False, align_corners=False, resample=bilinear)
+  (6): Normalize(p=1.0, p_batch=1.0, same_on_batch=True, mean=tensor([0.4850, 0.4560, 0.4060]), std=tensor([0.2290, 0.2240, 0.2250]))
+) 
+- validation transformations : Sequential(
+  (0): PreProcess()
+  (1): Resize(output_size=(518, 518), p=1.0, p_batch=1.0, same_on_batch=True, size=(518, 518), side=short, resample=bilinear, align_corners=True, antialias=False)
+  (2): Normalize(p=1.0, p_batch=1.0, same_on_batch=True, mean=tensor([0.4850, 0.4560, 0.4060]), std=tensor([0.2290, 0.2240, 0.2250]))
+)
+
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
@@ -50,7 +103,8 @@ The following hyperparameters were used during training:
 - eval_batch_size: 32
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
+- lr_scheduler_type: ReduceLROnPlateau with a patience of 5 epochs and a factor of 0.1
+- freeze_encoder: True
 - num_epochs: 150
 
 ### Training results