learning_rate: 0.001
train_batch_size: 4
eval_batch_size: 4
seed: 12345
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 32
total_eval_batch_size: 32
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_ratio: 0.1
num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Roc Auc	Best Threshold	Tpr At Best Threshold	Fpr At Best Threshold	Accuracy At Best Threshold	Avg Score True Positives	Avg Score True Negatives
No log	0	0	52.0496	0.6024	0.0005	0.3761	0.1	0.4830	0.0006	0.0002
5.5898	0.3571	5	18.8454	0.5830	1.0000	0.4359	0.2333	0.5034	1.0000	1.0000
9.444	0.7143	10	4.3848	0.6064	0.7957	0.4444	0.1667	0.5238	0.7334	0.6831
28.3906	1.0714	15	18.9426	0.5954	0.0496	0.5812	0.3333	0.5986	0.1507	0.0848
5.373	1.4286	20	6.4637	0.7140	0.9901	0.5043	0.1	0.5850	0.9549	0.9537
9.345	1.7857	25	5.4524	0.7563	0.9784	0.6410	0.1667	0.6803	0.9569	0.9291
5.3418	2.1429	30	3.4594	0.8333	0.9168	0.6667	0.1	0.7143	0.9014	0.7557
1.5391	2.5	35	3.7429	0.8692	0.9649	0.6410	0.0	0.7143	0.9489	0.8100
2.7979	2.8571	40	3.0348	0.8940	0.9442	0.7094	0.0667	0.7551	0.9347	0.6661
2.4222	3.2143	45	2.9182	0.9372	0.9540	0.8120	0.0667	0.8367	0.9633	0.6779
2.078	3.5714	50	2.2918	0.9296	0.5234	0.9145	0.2	0.8912	0.8245	0.2734
7.0792	3.9286	55	9.4923	0.8838	1.0000	0.6752	0.0	0.7415	0.9857	0.8457
1.0662	4.2857	60	2.0239	0.9383	0.9226	0.8120	0.0667	0.8367	0.9291	0.4229
1.7245	4.6429	65	6.1088	0.8872	0.1441	0.8205	0.1667	0.8231	0.6341	0.1319
4.0828	5.0	70	2.9309	0.9127	0.9599	0.8462	0.2	0.8367	0.9618	0.4915
0.0753	5.3571	75	2.4783	0.9655	0.9828	0.9060	0.0667	0.9116	0.9824	0.4380
0.1425	5.7143	80	1.8943	0.9644	0.9707	0.8632	0.0333	0.8844	0.9566	0.3225
0.2458	6.0714	85	2.0320	0.9615	0.8557	0.7949	0.0	0.8367	0.8681	0.1457
0.0487	6.4286	90	3.5018	0.9507	0.9993	0.8120	0.0667	0.8367	0.9839	0.4357
0.7497	6.7857	95	5.7445	0.9611	0.0183	0.9231	0.1333	0.9116	0.7639	0.0453
0.0312	7.1429	100	3.3379	0.9664	0.9730	0.9744	0.1667	0.9456	0.9868	0.3942
0.0018	7.5	105	2.0328	0.9658	0.9067	0.9402	0.1333	0.9252	0.9640	0.2284
0.001	7.8571	110	3.2039	0.9652	0.3961	0.8803	0.1	0.8844	0.8497	0.1036
0.0024	8.2143	115	2.3889	0.9655	0.9843	0.7778	0.0	0.8231	0.9006	0.1537
0.0	8.5714	120	2.4719	0.9664	0.9736	0.9316	0.1333	0.9184	0.9702	0.2444
0.0001	8.9286	125	2.7159	0.9682	0.9841	0.9487	0.1333	0.9320	0.9815	0.2714
0.0	9.2857	130	2.6452	0.9684	0.9806	0.9487	0.1333	0.9320	0.9796	0.2612
0.0	9.6429	135	2.6742	0.9675	0.9947	0.9231	0.1	0.9184	0.9800	0.2645
0.0005	10.0	140	2.7010	0.9672	0.9951	0.9145	0.1	0.9116	0.9787	0.2746

Framework versions

PEFT 0.18.1
Transformers 4.57.6
Pytorch 2.9.1+cu128
Datasets 4.5.0
Tokenizers 0.22.2

leobianco/npov_RM_Qwen_S12345_LLM_false_STRUCT_false_epo10_lr1e-3_r8_2602261748

npov_RM_Qwen_S12345_LLM_false_STRUCT_false_epo10_lr1e-3_r8_2602261748

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

leobianco/
npov_RM_Qwen_S12345_LLM_false_STRUCT_false_epo10_lr1e-3_r8_2602261748