segformer-b0-scene-parse-150

This model is a fine-tuned version of nvidia/mit-b0 on the Grizzlygg/CrownTest dataset. It achieves the following results on the evaluation set:

• Loss: 0.7268
• Mean Iou: 0.3888
• Mean Accuracy: 0.7776
• Overall Accuracy: 0.7776
• Accuracy Unlabeled: nan
• Accuracy Crown: 0.7776
• Iou Unlabeled: 0.0
• Iou Crown: 0.7776

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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Crown	Iou Unlabeled	Iou Crown
0.3303	10.0	20	0.6046	0.3722	0.7444	0.7444	nan	0.7444	0.0	0.7444
0.2269	20.0	40	0.6817	0.3613	0.7226	0.7226	nan	0.7226	0.0	0.7226
0.1893	30.0	60	0.7231	0.3717	0.7435	0.7435	nan	0.7435	0.0	0.7435
0.185	40.0	80	0.7688	0.3880	0.7760	0.7760	nan	0.7760	0.0	0.7760
0.1704	50.0	100	0.7268	0.3888	0.7776	0.7776	nan	0.7776	0.0	0.7776

Framework versions

• Transformers 4.30.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.13.3