global_mbv4_hybrid_large

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

• Loss: 0.0857
• Precision: 0.9806
• Recall: 0.9794
• Accuracy: 0.9834
• F1: 0.9800
• Roc Auc: 0.9986

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: 0.0001
• train_batch_size: 24
• eval_batch_size: 24
• 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
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	Accuracy	F1	Roc Auc
0.3639	0.2577	200	12.4134	0.8978	0.9088	0.9128	0.8996	0.9862
0.2433	0.5155	400	1.5055	0.7737	0.6830	0.6291	0.6382	0.9657
0.2974	0.7732	600	0.1301	0.9570	0.9435	0.9585	0.9491	0.9944
0.0867	1.0309	800	0.1688	0.9438	0.9530	0.9552	0.9472	0.9959
0.0392	1.2887	1000	0.2777	0.9549	0.9195	0.9448	0.9313	0.9970
0.2445	1.5464	1200	0.1142	0.9770	0.9689	0.9772	0.9726	0.9971
0.1727	1.8041	1400	0.2502	0.9683	0.9680	0.9737	0.9682	0.9964
0.1100	2.0619	1600	0.1452	0.9652	0.9707	0.9730	0.9676	0.9976
0.0381	2.3196	1800	0.1217	0.9779	0.9708	0.9788	0.9741	0.9974
0.0441	2.5773	2000	0.1124	0.9762	0.9764	0.9804	0.9763	0.9979
0.0792	2.8351	2200	0.1406	0.9821	0.9772	0.9832	0.9795	0.9983
0.0010	3.0928	2400	0.0857	0.9806	0.9794	0.9834	0.9800	0.9986
0.0003	3.3505	2600	0.1904	0.9821	0.9776	0.9832	0.9797	0.9980
0.0011	3.6082	2800	0.0997	0.9815	0.9790	0.9837	0.9803	0.9984
0.0053	3.8660	3000	0.1172	0.9797	0.9769	0.9821	0.9783	0.9980

Framework versions

• Transformers 5.3.0
• Pytorch 2.10.0+cu128
• Datasets 4.7.0
• Tokenizers 0.22.2