radar-encoder-freeze-pan26

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

• Loss: 0.7864
• Roc-auc: 0.716
• Brier: 0.731
• C@1: 0.65
• F1: 0.659
• F05u: 0.649
• Mean: 0.681

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 0.03
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Roc-auc	Brier	C@1	F1	F05u	Mean
0.1171	0.6748	500	0.1264	0.984	0.956	0.942	0.955	0.955	0.958
0.0899	1.3495	1000	0.0864	0.992	0.97	0.963	0.971	0.971	0.973
0.0498	2.0243	1500	0.0732	0.994	0.976	0.968	0.975	0.971	0.977
0.0716	2.6991	2000	0.0638	0.995	0.979	0.972	0.979	0.977	0.98
0.0672	3.3738	2500	0.0583	0.996	0.98	0.972	0.979	0.98	0.981
0.0453	4.0486	3000	0.0618	0.996	0.98	0.974	0.98	0.977	0.981
0.0785	4.7233	3500	0.0555	0.997	0.979	0.97	0.977	0.983	0.981
0.0467	5.3981	4000	0.0522	0.997	0.982	0.977	0.982	0.981	0.984
0.0672	6.0729	4500	0.0534	0.997	0.983	0.978	0.983	0.981	0.984
0.0249	6.7476	5000	0.0694	0.997	0.979	0.973	0.98	0.972	0.98
0.0227	7.4224	5500	0.0488	0.997	0.983	0.977	0.982	0.983	0.984
0.0261	8.0972	6000	0.0507	0.997	0.983	0.979	0.984	0.982	0.985
0.0654	8.7719	6500	0.0478	0.997	0.983	0.978	0.983	0.984	0.985
0.0291	9.4467	7000	0.0494	0.997	0.984	0.978	0.983	0.982	0.985

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2