<!-- 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. -->
segformer-b0-finetuned-segments-thin-sections-oct2
This model is a fine-tuned version of nvidia/mit-b0 on the mmcint/ara-sections-test dataset. It achieves the following results on the evaluation set:
- Loss: 0.5032
- Mean Iou: 0.0229
- Mean Accuracy: 0.0343
- Overall Accuracy: 0.0268
- Accuracy No-label: nan
- Accuracy Liptinite: 0.0686
- Accuracy Vitrinite: 0.0
- Accuracy Inertinite: nan
- Iou No-label: 0.0
- Iou Liptinite: 0.0686
- Iou Vitrinite: 0.0
- Iou Inertinite: nan
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: 6e-05
- train_batch_size: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy No-label | Accuracy Liptinite | Accuracy Vitrinite | Accuracy Inertinite | Iou No-label | Iou Liptinite | Iou Vitrinite | Iou Inertinite |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.1412 | 5.0 | 20 | 1.3339 | 0.1628 | 0.4001 | 0.3155 | nan | 0.7892 | 0.0110 | nan | 0.0 | 0.6403 | 0.0110 | 0.0 |
| 0.8928 | 10.0 | 40 | 1.0736 | 0.0046 | 0.0092 | 0.0072 | nan | 0.0183 | 0.0 | nan | 0.0 | 0.0183 | 0.0 | 0.0 |
| 0.785 | 15.0 | 60 | 0.7552 | 0.0027 | 0.0055 | 0.0043 | nan | 0.0109 | 0.0 | nan | 0.0 | 0.0109 | 0.0 | 0.0 |
| 0.6991 | 20.0 | 80 | 0.6564 | 0.0022 | 0.0033 | 0.0026 | nan | 0.0067 | 0.0 | nan | 0.0 | 0.0067 | 0.0 | nan |
| 0.6379 | 25.0 | 100 | 0.5975 | 0.0041 | 0.0062 | 0.0049 | nan | 0.0124 | 0.0 | nan | 0.0 | 0.0124 | 0.0 | nan |
| 0.7257 | 30.0 | 120 | 0.5259 | 0.0016 | 0.0023 | 0.0018 | nan | 0.0047 | 0.0 | nan | 0.0 | 0.0047 | 0.0 | nan |
| 0.613 | 35.0 | 140 | 0.5320 | 0.0131 | 0.0197 | 0.0154 | nan | 0.0394 | 0.0 | nan | 0.0 | 0.0394 | 0.0 | nan |
| 0.5534 | 40.0 | 160 | 0.5038 | 0.0135 | 0.0202 | 0.0158 | nan | 0.0405 | 0.0 | nan | 0.0 | 0.0405 | 0.0 | nan |
| 0.5424 | 45.0 | 180 | 0.5016 | 0.0278 | 0.0417 | 0.0327 | nan | 0.0835 | 0.0 | nan | 0.0 | 0.0835 | 0.0 | nan |
| 0.533 | 50.0 | 200 | 0.5032 | 0.0229 | 0.0343 | 0.0268 | nan | 0.0686 | 0.0 | nan | 0.0 | 0.0686 | 0.0 | nan |
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1
- Datasets 2.14.5
- Tokenizers 0.13.3
