You can deploy this model and start earning money today!

leobianco/
bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622

Name: bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622
Availability: InStock

mistralai/Mistral-7B-Instruct-v0.3

apache-2.0

Views

No views yet

bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 20.6371
Roc Auc: 0.7638
Best Threshold: 0.9969
Tpr At Best Threshold: 0.7746
Fpr At Best Threshold: 0.3182
Accuracy At Best Threshold: 0.7527
Avg Score True Positives: 0.8507
Avg Score True Negatives: 0.4702

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: 1
eval_batch_size: 1
seed: 130798
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	79.1626	0.5855	0.0000	0.7887	0.6136	0.6935	0.2545	0.1752
72.2571	0.0820	5	65.5011	0.5946	0.0000	0.7746	0.5909	0.6882	0.3428	0.2329
16.0313	0.1639	10	29.7849	0.6248	0.6442	0.8239	0.5909	0.7258	0.8432	0.6325
80.6602	0.2459	15	30.8574	0.6408	0.9954	0.8521	0.5455	0.7581	0.9126	0.6668
50.8594	0.3279	20	32.1196	0.6572	0.0005	0.8662	0.5227	0.7742	0.6737	0.4699
98.4743	0.4098	25	45.5972	0.6573	0.0000	0.8451	0.5227	0.7581	0.5048	0.3263
67.25	0.4918	30	31.2721	0.6817	0.0015	0.8239	0.5227	0.7419	0.6572	0.4026
53.5156	0.5738	35	26.6427	0.7370	1.0000	0.6338	0.2273	0.6667	0.9028	0.5657
14.6237	0.6557	40	47.5671	0.7090	0.0072	0.6479	0.2955	0.6613	0.4693	0.1919
29.7305	0.7377	45	35.5632	0.7179	0.8706	0.5986	0.2273	0.6398	0.6292	0.2982
16.4668	0.8197	50	24.1822	0.7462	0.9874	0.7394	0.3409	0.7204	0.8192	0.4850
43.6588	0.9016	55	31.9505	0.7222	0.0000	0.8592	0.5	0.7742	0.6501	0.3239
21.0022	0.9836	60	46.1210	0.7088	0.0815	0.5634	0.2045	0.6183	0.5036	0.1984
11.6331	1.0656	65	25.1979	0.7406	1.0000	0.6127	0.25	0.6452	0.7878	0.4813
46.3125	1.1475	70	25.2101	0.7303	1.0000	0.6620	0.2727	0.6774	0.8302	0.4996
9.4047	1.2295	75	32.7500	0.7297	0.0000	0.8873	0.5227	0.7903	0.6274	0.3043
63.75	1.3115	80	48.6296	0.7136	0.0000	0.6831	0.3409	0.6774	0.4234	0.1882
0.1065	1.3934	85	24.5158	0.7152	0.9951	0.7254	0.3409	0.7097	0.7991	0.5129
17.5223	1.4754	90	33.7389	0.6969	1.0	0.7817	0.4091	0.7366	0.9448	0.7661
6.1267	1.5574	95	26.9548	0.6949	0.9999	0.6690	0.3409	0.6667	0.8537	0.6065
0.9258	1.6393	100	32.0598	0.6945	0.0010	0.8380	0.5227	0.7527	0.6609	0.3934
0.0016	1.7213	105	25.8364	0.7223	0.9987	0.6479	0.2955	0.6613	0.7626	0.4472
1.2126	1.8033	110	23.1846	0.7446	0.9943	0.7676	0.3636	0.7366	0.8494	0.5490
1.2505	1.8852	115	22.6262	0.7502	1.0000	0.7535	0.3182	0.7366	0.9041	0.5740
4.875	1.9672	120	22.7600	0.7566	1.0000	0.7606	0.3182	0.7419	0.9210	0.5600
0.9001	2.0492	125	20.7627	0.7683	0.9990	0.7746	0.3182	0.7527	0.8780	0.5033
0.0003	2.1311	130	21.1997	0.7642	0.9995	0.7958	0.3182	0.7688	0.8873	0.5248
11.5625	2.2131	135	21.5238	0.7631	0.9997	0.7887	0.3182	0.7634	0.8851	0.5311
6.0942	2.2951	140	21.6662	0.7502	0.9993	0.8028	0.3182	0.7742	0.8771	0.5224
0.0001	2.3770	145	21.5443	0.7564	0.9984	0.7535	0.3182	0.7366	0.8455	0.4743
0.0	2.4590	150	21.4491	0.7606	0.8267	0.8169	0.3864	0.7688	0.8307	0.4397
10.6875	2.5410	155	21.4279	0.7606	0.9654	0.8028	0.3636	0.7634	0.8306	0.4467
3.9062	2.6230	160	21.2492	0.7578	0.9883	0.7887	0.3409	0.7581	0.8418	0.4602
0.0005	2.7049	165	21.0866	0.7575	0.9833	0.8099	0.3636	0.7688	0.8468	0.4685
0.1292	2.7869	170	21.1093	0.7592	0.9702	0.8099	0.3636	0.7688	0.8475	0.4628
17.2188	2.8689	175	20.8946	0.7638	0.9922	0.8099	0.3636	0.7688	0.8519	0.4709
0.0	2.9508	180	20.6371	0.7638	0.9969	0.7746	0.3182	0.7527	0.8507	0.4702

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_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622

bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

leobianco/
bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr1e-4_r8_2602051622