segcrack9k_conglomerate_segformer

This model is a fine-tuned version of nvidia/mit-b5 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0333
• Mean Iou: 0.3608
• Mean Accuracy: 0.7217
• Overall Accuracy: 0.7217
• Accuracy Background: nan
• Accuracy Crack: 0.7217
• Iou Background: 0.0
• Iou Crack: 0.7217

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 Crack	Iou Background	Iou Crack
0.0259	0.14	1000	0.0404	0.3267	0.6534	0.6534	nan	0.6534	0.0	0.6534
0.0186	0.27	2000	0.0378	0.3586	0.7172	0.7172	nan	0.7172	0.0	0.7172
0.0348	0.41	3000	0.0375	0.3209	0.6418	0.6418	nan	0.6418	0.0	0.6418
0.011	0.54	4000	0.0356	0.3496	0.6991	0.6991	nan	0.6991	0.0	0.6991
0.0132	0.68	5000	0.0350	0.3459	0.6918	0.6918	nan	0.6918	0.0	0.6918
0.0573	0.81	6000	0.0339	0.3575	0.7149	0.7149	nan	0.7149	0.0	0.7149
0.1466	0.95	7000	0.0333	0.3608	0.7217	0.7217	nan	0.7217	0.0	0.7217

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3