segformer-b5-finetuned-ce-head-batch3

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: 0.0522
• Mean Iou: 0.7770
• Mean Accuracy: 0.8416
• Overall Accuracy: 0.9816
• Accuracy Bg: 0.9924
• Accuracy Head: 0.6907
• Iou Bg: 0.9812
• Iou Head: 0.5729

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Bg	Accuracy Head	Iou Bg	Iou Head
0.089	2.9412	100	0.0360	0.7934	0.8932	0.9859	0.9911	0.7952	0.9856	0.6012
0.0178	5.8824	200	0.0358	0.7851	0.8727	0.9859	0.9922	0.7532	0.9856	0.5845
0.0047	8.8235	300	0.0324	0.8094	0.8977	0.9875	0.9926	0.8028	0.9872	0.6316

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1
• Datasets 3.1.0
• Tokenizers 0.20.3