segformer_cracks

This model is a fine-tuned version of nvidia/mit-b3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0499
• Mean Iou: 0.7718
• Mean Accuracy: 0.8317
• Overall Accuracy: 0.9798
• Per Category Iou: [0.9792869895386617, 0.564265846038068]
• Per Category Accuracy: [0.9923313345080351, 0.671108360646227]

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Per Category Iou	Per Category Accuracy
0.0686	1.0	1541	0.0557	0.7541	0.8082	0.9785	[0.9779708221514636, 0.5303006858963294]	[0.9928845967047768, 0.6234677160845897]
0.049	2.0	3082	0.0527	0.7633	0.8239	0.9790	[0.9784073819168878, 0.5481400031368636]	[0.9920481107810992, 0.6557623260692017]
0.0468	3.0	4623	0.0547	0.7526	0.7996	0.9788	[0.9783360187606548, 0.5269084757862701]	[0.993975994702418, 0.6052805015161615]
0.0456	4.0	6164	0.0509	0.7677	0.8276	0.9794	[0.9788937969015667, 0.556522438909845]	[0.9922581622702671, 0.6629042271896711]
0.044	5.0	7705	0.0505	0.7678	0.8265	0.9795	[0.9789809420595871, 0.5566804258721124]	[0.9924358981457169, 0.6606494246283242]
0.0436	6.0	9246	0.0502	0.7696	0.8265	0.9798	[0.9792607857315766, 0.5598563478221208]	[0.9927329763880554, 0.6603118480646505]
0.0431	7.0	10787	0.0499	0.7718	0.8317	0.9798	[0.9792869895386617, 0.564265846038068]	[0.9923313345080351, 0.671108360646227]

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1
• Datasets 2.14.5
• Tokenizers 0.13.3