windowz_test-020525

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Model Preparation Time: 0.0011
• Accuracy: 0.9920
• F1: 0.9921
• Iou: 0.9850
• Contour Dice: 0.9925
• Loss: 0.0221

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 1000
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Model Preparation Time		Dice	Validation Loss
0.4601	5.0	25630	0.0011	0.9713	0.9889	0.0642
0.4405	10.0	51260	0.0011	0.9760	0.9904	0.0372
0.3981	15.0	76890	0.0011	0.9793	0.9932	0.0316
0.361	20.0	102520	0.0011	0.9830	0.9925	0.0340
0.3978	25.0	128150	0.0011	0.9814	0.9903	0.0309
0.4293	30.0	153780	0.0011	0.9835	0.9923	0.0259
0.3522	35.0	179410	0.0011	0.9855	0.9927	0.0269
0.3718	40.0	205040	0.0011	0.9842	0.9931	0.0265
0.37	45.0	230670	0.0011	0.9851	0.9931	0.0242
0.3866	50.0	256300	0.0011	0.984	0.9920	0.0255
0.3416	55.0	281930	0.0011	0.9848	0.9930	0.0249
0.3461	60.0	307560	0.0011	0.9850	0.9925	0.0221
0.3509	65.0	333190	0.0011	0.9842	0.9919	0.0240
0.3616	70.0	358820	0.0011	0.9847	0.9931	0.0226
0.3406	75.0	384450	0.0011	0.9840	0.9926	0.0237

Framework versions

• Transformers 4.45.0
• Pytorch 2.5.1+cu124
• Datasets 2.21.0
• Tokenizers 0.20.3