README.md

metadata

license: other
tags:
  - generated_from_trainer
model-index:
  - name: segformer-b0-DeepCrack
    results: []

segformer-b0-DeepCrack

This model is a fine-tuned version of nvidia/mit-b4 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0145
• Mean Iou: 0.0585
• Mean Accuracy: 0.1170
• Overall Accuracy: 0.1170
• Accuracy Background: nan
• Accuracy Cracked: 0.1170
• Iou Background: 0.0
• Iou Cracked: 0.1170

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Cracked	Iou Background	Iou Cracked
0.3566	0.13	20	0.2932	0.0200	0.0401	0.0401	nan	0.0401	0.0	0.0401
0.1639	0.27	40	0.1326	0.1353	0.2706	0.2706	nan	0.2706	0.0	0.2706
0.0562	0.4	60	0.0548	0.0681	0.1363	0.1363	nan	0.1363	0.0	0.1363
0.0283	0.53	80	0.0294	0.0815	0.1630	0.1630	nan	0.1630	0.0	0.1630
0.0187	0.67	100	0.0194	0.0720	0.1439	0.1439	nan	0.1439	0.0	0.1439
0.0144	0.8	120	0.0160	0.0656	0.1313	0.1313	nan	0.1313	0.0	0.1313
0.0152	0.93	140	0.0145	0.0585	0.1170	0.1170	nan	0.1170	0.0	0.1170

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3