learning_rate: 0.0005
train_batch_size: 1
eval_batch_size: 1
seed: 12345
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 8
total_eval_batch_size: 8
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: 5.0

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	40.8364	0.6060	0.0100	0.4225	0.1364	0.5269	0.1403	0.0493
28.4303	0.0820	5	37.2045	0.6061	0.0178	0.4225	0.1591	0.5215	0.1723	0.0586
2.4595	0.1639	10	17.3171	0.6252	0.7337	0.4437	0.1591	0.5376	0.4992	0.2985
21.8362	0.2459	15	13.1581	0.6402	0.9268	0.4718	0.1364	0.5645	0.6271	0.4886
17.4392	0.3279	20	17.1869	0.6330	0.5789	0.4296	0.1364	0.5323	0.4593	0.2454
6.61	0.4098	25	13.5830	0.6543	0.7941	0.4648	0.1364	0.5591	0.5562	0.3255
2.8632	0.4918	30	20.3218	0.6508	0.0351	0.6268	0.3182	0.6398	0.3088	0.1698
18.7251	0.5738	35	9.4912	0.7066	0.7886	0.5986	0.2045	0.6452	0.6831	0.4018
10.6638	0.6557	40	14.5011	0.7310	0.1312	0.6408	0.1591	0.6882	0.4086	0.1126
16.9208	0.7377	45	5.1152	0.7588	0.8597	0.6620	0.2045	0.6935	0.7940	0.4915
10.5139	0.8197	50	10.9615	0.7330	0.1871	0.6831	0.25	0.6989	0.4694	0.1836
22.8182	0.9016	55	7.8489	0.7277	0.9529	0.8521	0.4773	0.7742	0.9442	0.8424
3.8305	0.9836	60	6.1365	0.6926	0.9258	0.7113	0.3636	0.6935	0.8900	0.7313
3.3955	1.0656	65	5.3038	0.7042	0.7860	0.7042	0.3182	0.6989	0.8117	0.5467
5.919	1.1475	70	4.3711	0.7678	0.7387	0.8099	0.3864	0.7634	0.8312	0.5278
12.4359	1.2295	75	4.2587	0.7984	0.9317	0.5563	0.0909	0.6398	0.8135	0.4768
0.7468	1.3115	80	5.8436	0.8170	0.6907	0.5704	0.0682	0.6559	0.6335	0.2486
4.0849	1.3934	85	4.0666	0.8301	0.6783	0.8732	0.3864	0.8118	0.8714	0.4930
2.6468	1.4754	90	6.6973	0.8385	0.9977	0.7113	0.1818	0.7366	0.9653	0.7259
15.4974	1.5574	95	5.5296	0.8409	0.9515	0.7254	0.1818	0.7473	0.8724	0.4492
12.9018	1.6393	100	15.6907	0.8256	1.0000	0.8380	0.3182	0.8011	0.9725	0.8632
0.0327	1.7213	105	18.1021	0.7775	1.0000	0.5915	0.1136	0.6613	0.9726	0.8857
5.6293	1.8033	110	18.3712	0.7764	1.0000	0.6479	0.1818	0.6882	0.9809	0.8961
0.332	1.8852	115	15.2280	0.7768	1.0000	0.6268	0.1364	0.6828	0.9731	0.8731
4.1369	1.9672	120	17.5538	0.7668	1.0000	0.8169	0.3636	0.7742	0.9824	0.9020
0.0015	2.0492	125	15.5503	0.7614	0.9999	0.7676	0.3182	0.7473	0.9748	0.8720
0.0001	2.1311	130	10.2795	0.7864	0.9995	0.6831	0.1591	0.7204	0.9537	0.7530
3.168	2.2131	135	17.9315	0.7556	1.0000	0.8310	0.3409	0.7903	0.9697	0.8612
1.7896	2.2951	140	25.6388	0.7411	1.0000	0.8451	0.3636	0.7957	0.9922	0.9052
0.0	2.3770	145	18.7833	0.7653	1.0000	0.7887	0.3182	0.7634	0.9683	0.8811
0.0018	2.4590	150	18.3909	0.7511	1.0000	0.6761	0.2273	0.6989	0.9563	0.8601
0.0	2.5410	155	25.9910	0.7547	1.0	0.7465	0.2955	0.7366	0.9815	0.9009
0.0019	2.6230	160	34.1670	0.6602	1.0	0.9366	0.6136	0.8065	0.9936	0.9545
1.5701	2.7049	165	22.7788	0.7576	1.0000	0.8310	0.3864	0.7796	0.9887	0.9090
0.0476	2.7869	170	11.7353	0.7877	0.9998	0.7465	0.2727	0.7419	0.9405	0.6857
0.0	2.8689	175	20.2683	0.7806	1.0000	0.7465	0.25	0.7473	0.9702	0.8812
0.0023	2.9508	180	26.1243	0.7346	1.0	0.8169	0.3636	0.7742	0.9755	0.8918
0.0	3.0328	185	29.7478	0.6829	1.0	0.8592	0.5	0.7742	0.9791	0.8963
0.0	3.1148	190	30.3112	0.6662	1.0	0.8732	0.5455	0.7742	0.9779	0.8930
3.5625	3.1967	195	36.5642	0.5975	1.0	0.9225	0.7273	0.7688	0.9927	0.9145
0.0004	3.2787	200	41.5135	0.5587	1.0	0.9577	0.8409	0.7688	0.9930	0.9505
0.0377	3.3607	205	40.3929	0.5637	1.0	0.9437	0.8182	0.7634	0.9929	0.9384
0.0002	3.4426	210	37.0087	0.5878	1.0	0.9225	0.75	0.7634	0.9910	0.9106
0.0	3.5246	215	34.9349	0.6218	1.0	0.9014	0.6591	0.7688	0.9860	0.9048
0.0025	3.6066	220	33.4669	0.6400	1.0	0.8944	0.6136	0.7742	0.9828	0.8984
1.0156	3.6885	225	33.2069	0.6396	1.0	0.8944	0.6136	0.7742	0.9807	0.8936
0.0015	3.7705	230	34.4314	0.6328	1.0	0.9014	0.6364	0.7742	0.9826	0.8972
0.0	3.8525	235	35.1037	0.6144	1.0000	0.9225	0.6818	0.7796	0.9850	0.9025
0.0	3.9344	240	35.1884	0.6251	1.0	0.9085	0.6591	0.7742	0.9849	0.9038
0.0001	4.0164	245	35.2080	0.6226	1.0	0.9014	0.6591	0.7688	0.9849	0.9005
0.0001	4.0984	250	35.3930	0.6126	1.0000	0.9155	0.6818	0.7742	0.9837	0.8964
0.0	4.1803	255	35.5661	0.6120	1.0000	0.9155	0.6818	0.7742	0.9841	0.8982
0.0	4.2623	260	35.4792	0.6126	1.0000	0.9155	0.6818	0.7742	0.9838	0.8998
0.0	4.3443	265	35.4589	0.6120	1.0000	0.9155	0.6818	0.7742	0.9834	0.8981
0.0	4.4262	270	35.9668	0.6145	1.0	0.9085	0.6818	0.7688	0.9839	0.9000
0.0	4.5082	275	36.3038	0.6173	1.0	0.9155	0.6818	0.7742	0.9856	0.9011
0.0	4.5902	280	36.5773	0.6133	1.0	0.9085	0.6818	0.7688	0.9856	0.9035
0.0	4.6721	285	36.4383	0.6200	1.0	0.9225	0.6818	0.7796	0.9858	0.9026
0.0	4.7541	290	36.5334	0.6164	1.0	0.9155	0.6818	0.7742	0.9854	0.9020
0.0	4.8361	295	36.4867	0.6139	1.0	0.9085	0.6818	0.7688	0.9856	0.9018
0.0	4.9180	300	36.5142	0.6169	1.0	0.9155	0.6818	0.7742	0.9854	0.9010
0.0001	5.0	305	36.5186	0.6172	1.0	0.9155	0.6818	0.7742	0.9851	0.9027

Framework versions

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

leobianco/bosch_RM_Qwen_S12345_LLM_false_STRUCT_true_epo5_lr5e-4_r8_2602041705

bosch_RM_Qwen_S12345_LLM_false_STRUCT_true_epo5_lr5e-4_r8_2602041705

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

leobianco/
bosch_RM_Qwen_S12345_LLM_false_STRUCT_true_epo5_lr5e-4_r8_2602041705