radar-encoder-freeze-raid

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

• Loss: 0.1972
• Roc-auc: 0.974
• Brier: 0.941
• C@1: 0.92
• F1: 0.918
• F05u: 0.935
• Mean: 0.938

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
0.2243	1.0776	500	0.3152	0.946	0.898	0.85	0.83	0.912	0.887
0.2362	2.1552	1000	0.2601	0.958	0.919	0.887	0.881	0.923	0.914
0.1790	3.2328	1500	0.2396	0.963	0.926	0.9	0.895	0.929	0.923
0.2652	4.3103	2000	0.2677	0.965	0.916	0.885	0.875	0.934	0.915
0.1927	5.3879	2500	0.2230	0.968	0.932	0.906	0.908	0.908	0.925
0.1476	6.4655	3000	0.2172	0.971	0.933	0.908	0.905	0.936	0.931
0.2706	7.5431	3500	0.2093	0.971	0.936	0.913	0.913	0.928	0.932
0.1720	8.6207	4000	0.2072	0.972	0.937	0.914	0.913	0.929	0.933
0.1574	9.6983	4500	0.2077	0.972	0.937	0.914	0.913	0.931	0.933

Framework versions

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