learning_rate: 0.001
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: 3.0

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
20.2563	0.0820	5	21.3957	0.6098	0.8081	0.2887	0.0455	0.4462	0.4070	0.2101
10.1877	0.1639	10	10.5430	0.6259	0.9990	0.3944	0.0909	0.5161	0.8344	0.7934
19.7171	0.2459	15	15.0145	0.6640	0.6460	0.4366	0.1364	0.5376	0.4769	0.2386
12.7113	0.3279	20	9.1551	0.6994	0.9698	0.5986	0.2045	0.6452	0.7857	0.6288
2.9301	0.4098	25	12.7105	0.7198	0.0940	0.6690	0.2955	0.6774	0.4496	0.1640
12.105	0.4918	30	6.1946	0.7462	0.8701	0.6197	0.2045	0.6613	0.7646	0.4504
22.5626	0.5738	35	12.1968	0.6490	0.1597	0.7042	0.4318	0.6720	0.5121	0.3108
26.3391	0.6557	40	10.2366	0.4358	0.9904	0.1690	0.1136	0.3387	0.6834	0.7566
11.7123	0.7377	45	9.4350	0.6479	0.2422	0.6972	0.4545	0.6613	0.5246	0.3528
17.0171	0.8197	50	7.6977	0.6729	0.8069	0.5141	0.1818	0.5860	0.6565	0.4523
19.4829	0.9016	55	5.4818	0.7531	0.9349	0.6479	0.1591	0.6935	0.8135	0.5483
6.3104	0.9836	60	8.7774	0.6732	0.9904	0.7254	0.4091	0.6935	0.9464	0.9026
2.7738	1.0656	65	6.0502	0.6817	0.7432	0.5070	0.2045	0.5753	0.6498	0.4424
2.8422	1.1475	70	4.5455	0.7391	0.8403	0.6901	0.2955	0.6935	0.8344	0.6000
3.8587	1.2295	75	6.1371	0.7571	0.4321	0.7606	0.2955	0.7473	0.6200	0.3220
2.7705	1.3115	80	17.4541	0.7652	1.0000	0.7746	0.3636	0.7419	0.9949	0.9421
0.1934	1.3934	85	15.7842	0.7855	0.9999	0.8380	0.3409	0.7957	0.9751	0.8777
0.0025	1.4754	90	21.1862	0.7720	1.0000	0.7535	0.2955	0.7419	0.9638	0.8749
9.3142	1.5574	95	34.8107	0.5773	1.0	0.8803	0.7273	0.7366	0.9655	0.8864
1.5052	1.6393	100	50.0028	0.5371	0.9999	0.9718	0.8864	0.7688	0.9849	0.9177
0.0006	1.7213	105	48.1031	0.5365	1.0000	0.9648	0.8864	0.7634	0.9818	0.9278
11.6875	1.8033	110	32.3815	0.6556	1.0	0.8380	0.5227	0.7527	0.9709	0.8929
0.0	1.8852	115	22.3083	0.6992	1.0000	0.7676	0.3864	0.7312	0.9544	0.8861
2.77	1.9672	120	29.0116	0.6342	1.0000	0.8873	0.6136	0.7688	0.9926	0.9525
0.0004	2.0492	125	32.2736	0.6388	1.0000	0.9155	0.6364	0.7849	0.9930	0.9577
0.0393	2.1311	130	17.3054	0.7246	1.0000	0.6620	0.2955	0.6720	0.9736	0.8861
8.1721	2.2131	135	13.1604	0.7092	0.9998	0.6620	0.2727	0.6774	0.9499	0.8413
0.0004	2.2951	140	19.4123	0.7173	1.0000	0.7535	0.3409	0.7312	0.9641	0.8849
0.0003	2.3770	145	22.6310	0.7087	1.0000	0.7817	0.3864	0.7419	0.9646	0.8879
0.0093	2.4590	150	23.4070	0.7050	1.0000	0.7817	0.4091	0.7366	0.9625	0.8878
0.0002	2.5410	155	23.9414	0.7025	1.0000	0.7254	0.3409	0.7097	0.9627	0.8881
5.0789	2.6230	160	25.0047	0.6994	1.0000	0.7958	0.4091	0.7473	0.9654	0.8890
0.0038	2.7049	165	26.9008	0.7073	1.0000	0.7535	0.3636	0.7258	0.9706	0.8909
0.0001	2.7869	170	27.2674	0.7022	1.0	0.6972	0.3409	0.6882	0.9717	0.8909
0.0147	2.8689	175	27.2968	0.7085	1.0	0.7254	0.3409	0.7097	0.9723	0.8903
0.0	2.9508	180	27.1671	0.7177	1.0	0.7113	0.3182	0.7043	0.9724	0.8912

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_epo3_lr1e-3_r8_2602050904

bosch_RM_Qwen_S12345_LLM_false_STRUCT_true_epo3_lr1e-3_r8_2602050904

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_epo3_lr1e-3_r8_2602050904