segformer-b0-finetuned-test

This model is a fine-tuned version of nvidia/mit-b0 on the None dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.2053
• eval_mean_iou: 0.5448
• eval_mean_accuracy: 0.6296
• eval_overall_accuracy: 0.9130
• eval_accuracy_Structure (dimensional): nan
• eval_accuracy_Impervious (planiform): 0.9578
• eval_accuracy_Fences: 0.3758
• eval_accuracy_Water Storage/Tank: nan
• eval_accuracy_Pool < 100 sqft: 0.0
• eval_accuracy_Pool > 100 sqft: 0.8208
• eval_accuracy_Irrigated Planiform: 0.8708
• eval_accuracy_Irrigated Dimensional Low: 0.6817
• eval_accuracy_Irrigated Dimensional High: 0.9472
• eval_accuracy_Irrigated Bare: 0.4827
• eval_accuracy_Irrigable Planiform: 0.6668
• eval_accuracy_Irrigable Dimensional Low: 0.6013
• eval_accuracy_Irrigable Dimensional High: 0.7902
• eval_accuracy_Irrigable Bare: 0.5657
• eval_accuracy_Native Planiform: 0.9093
• eval_accuracy_Native Dimensional Low: 0.0
• eval_accuracy_Native Dimensional High: 0.0961
• eval_accuracy_Native Bare: 0.9332
• eval_accuracy_UDL: nan
• eval_accuracy_Open Water: 0.6613
• eval_accuracy_Artificial Turf: 0.9720
• eval_iou_Structure (dimensional): 0.0
• eval_iou_Impervious (planiform): 0.8964
• eval_iou_Fences: 0.3104
• eval_iou_Water Storage/Tank: nan
• eval_iou_Pool < 100 sqft: 0.0
• eval_iou_Pool > 100 sqft: 0.8199
• eval_iou_Irrigated Planiform: 0.7563
• eval_iou_Irrigated Dimensional Low: 0.5480
• eval_iou_Irrigated Dimensional High: 0.8920
• eval_iou_Irrigated Bare: 0.4053
• eval_iou_Irrigable Planiform: 0.6007
• eval_iou_Irrigable Dimensional Low: 0.5083
• eval_iou_Irrigable Dimensional High: 0.7595
• eval_iou_Irrigable Bare: 0.5106
• eval_iou_Native Planiform: 0.8678
• eval_iou_Native Dimensional Low: 0.0
• eval_iou_Native Dimensional High: 0.0961
• eval_iou_Native Bare: 0.8293
• eval_iou_UDL: nan
• eval_iou_Open Water: 0.5929
• eval_iou_Artificial Turf: 0.9584
• eval_runtime: 6.2852
• eval_samples_per_second: 15.91
• eval_steps_per_second: 1.114
• epoch: 10.8
• step: 270

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1