slm-segformer-080823-b1

This model is a fine-tuned version of nvidia/mit-b1 on patches from sketched boundaries during the SmartLandMaps project. It achieves the following results on the evaluation set:

• Train Loss: 0.0257
• Validation Loss: 0.0271
• Validation Mean Iou: 0.8583
• Validation Mean Accuracy: 0.9196
• Validation Overall Accuracy: 0.9888
• Validation Per Category Iou: [0.98849374 0.72800628]
• Validation Per Category Accuracy: [0.99410982 0.84499592]
• Epoch: 9

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:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 6e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Validation Loss	Validation Mean Iou	Validation Mean Accuracy	Validation Overall Accuracy	Validation Per Category Iou	Validation Per Category Accuracy	Epoch
0.1099	0.0958	0.8134	0.9461	0.9824	[0.98176563 0.64502586]	[0.98511039 0.90705184]	0
0.0478	0.0523	0.8440	0.9449	0.9865	[0.9860406 0.7019479]	[0.98964683 0.90022015]	1
0.0379	0.0409	0.8476	0.9325	0.9873	[0.98687303 0.70826431]	[0.99144882 0.87350047]	2
0.0336	0.0331	0.8551	0.9394	0.9880	[0.98757544 0.72269531]	[0.99166337 0.88706795]	3
0.0310	0.0329	0.8541	0.9426	0.9878	[0.98735586 0.72081351]	[0.99118853 0.89409615]	4
0.0292	0.0317	0.8516	0.9348	0.9877	[0.98729336 0.71599644]	[0.99171219 0.87789181]	5
0.0282	0.0296	0.8572	0.9336	0.9884	[0.98798391 0.72647977]	[0.99252058 0.87472321]	6
0.0271	0.0319	0.8476	0.9253	0.9875	[0.98709267 0.70802268]	[0.99221744 0.85835533]	7
0.0267	0.0295	0.8549	0.9298	0.9882	[0.98782022 0.72192985]	[0.9926388 0.86691624]	8
0.0257	0.0271	0.8583	0.9196	0.9888	[0.98849374 0.72800628]	[0.99410982 0.84499592]	9

Framework versions

• Transformers 4.31.0
• TensorFlow 2.12.0
• Tokenizers 0.13.3