mit-b0_whitefly

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

• Loss: 0.2644
• Mean Iou: 0.4948
• Mean Accuracy: 0.4968
• Overall Accuracy: 0.9893
• Accuracy Background: 0.9907
• Accuracy Whitefly: 0.0029
• Iou Background: 0.9893
• Iou Whitefly: 0.0004

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
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Whitefly	Iou Background	Iou Whitefly
0.6275	0.4	20	0.6849	0.3529	0.4550	0.7049	0.7056	0.2044	0.7048	0.0010
0.4423	0.8	40	0.5826	0.4745	0.5157	0.9468	0.9480	0.0834	0.9468	0.0022
0.3793	1.2	60	0.4444	0.4868	0.4927	0.9731	0.9744	0.0110	0.9731	0.0006
0.3102	1.6	80	0.3347	0.4976	0.4986	0.9949	0.9963	0.0009	0.9949	0.0002
0.272	2.0	100	0.3100	0.4983	0.4991	0.9963	0.9977	0.0004	0.9963	0.0002
0.3003	2.4	120	0.2579	0.4983	0.4991	0.9965	0.9979	0.0003	0.9965	0.0001
0.2558	2.8	140	0.2644	0.4948	0.4968	0.9893	0.9907	0.0029	0.9893	0.0004

Framework versions

• Transformers 4.44.1
• Pytorch 2.6.0+cpu
• Datasets 2.21.0
• Tokenizers 0.19.1