windowz_test

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

• Model Preparation Time: 0.0013
• Accuracy: 0.9922
• F1: 0.9924
• Iou: 0.9856
• Loss: 0.0214

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		Validation Loss
0.4994	5.0	25630	0.0013	0.9812	0.0328
0.4423	10.0	51260	0.0013	0.9806	0.0356
0.4365	15.0	76890	0.0013	0.9836	0.0309
0.3913	20.0	102520	0.0013	0.1924	0.9382
0.3415	25.0	128150	0.0013	0.9851	0.0250
0.3952	30.0	153780	0.0013	0.9860	0.0236
0.3746	35.0	179410	0.0013	0.9818	0.0288
0.3569	40.0	205040	0.0013	0.1850	0.8815
0.3612	45.0	230670	0.0013	0.9856	0.0214
0.3501	50.0	256300	0.0013	0.9848	0.0241
0.3374	55.0	281930	0.0013	0.9847	0.0232
0.3742	60.0	307560	0.0013	0.1883	0.9705

Framework versions

• Transformers 4.45.0
• Pytorch 2.5.1
• Datasets 3.2.0
• Tokenizers 0.20.3