groderg/SegForCoral-b2-2025_06_30_19127-bs16_refine

SegForCoral-b2-2025_06_30_19127-bs16_refine is a fine-tuned version of nvidia/mit-b2.

---

Model description

SegForCoral-b2-2025_06_30_19127-bs16_refine is a model built on top of nvidia/mit-b2 model for underwater multilabel image classification.The classification head is a combination of linear, ReLU, batch normalization, and dropout layers.

The source code for training the model can be found in this Git repository.

• Developed by: lombardata, credits to César Leblanc and Victor Illien

---

Intended uses & limitations

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 procedure

Training hyperparameters

The following hyperparameters were used during training:

• Number of Epochs: 20.0
• Learning Rate: 1e-05
• Train Batch Size: 16
• Eval Batch Size: 16
• Optimizer: Adam
• LR Scheduler Type: ReduceLROnPlateau with a patience of 5 epochs and a factor of 0.1
• Freeze Encoder: No
• Data Augmentation: No

Training results

Epoch	Validation Loss	Learning Rate
1	0.6623146533966064	1e-05
2	0.6218862533569336	1e-05
3	0.5973241329193115	1e-05
4	0.5866864919662476	1e-05
5	0.5728989243507385	1e-05
6	0.5666205286979675	1e-05
7	0.5585390329360962	1e-05
8	0.5711775422096252	1e-05
9	0.550676703453064	1e-05
10	0.5479335188865662	1e-05
11	0.5629982352256775	1e-05
12	0.5543109178543091	1e-05
13	0.5586446523666382	1e-05
14	0.5647701621055603	1e-05
15	0.5843802690505981	1e-05
16	0.5760069489479065	1e-05
17	0.5521435141563416	1.0000000000000002e-06
18	0.5510115027427673	1.0000000000000002e-06
19	0.5491860508918762	1.0000000000000002e-06
20	0.5507251620292664	1.0000000000000002e-06

---

Framework Versions

• Transformers: 4.51.3
• Pytorch: 2.6.0+cu124
• Datasets: 3.6.0
• Tokenizers: 0.21.1