You can deploy this model and start earning money today!

leobianco/
bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr5e-4_r8_2602051922

Name: bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr5e-4_r8_2602051922
Availability: InStock

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

apache-2.0

Views

No views yet

bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr5e-4_r8_2602051922

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: 45.9846
Roc Auc: 0.5331
Best Threshold: 1.0000
Tpr At Best Threshold: 0.9718
Fpr At Best Threshold: 0.8864
Accuracy At Best Threshold: 0.7688
Avg Score True Positives: 0.9916
Avg Score True Negatives: 0.9679

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: 0.0005
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
57.3731	0.0820	5	29.4257	0.6147	0.9964	0.7465	0.5	0.6882	0.8456	0.6452
1.7539	0.1639	10	27.3688	0.6593	0.2628	0.8944	0.5682	0.7849	0.8649	0.5807
49.2344	0.2459	15	42.2568	0.7065	0.0496	0.6268	0.25	0.6559	0.5394	0.2132
10.7992	0.3279	20	22.5685	0.7871	1.0000	0.8028	0.2727	0.7849	0.8911	0.5055
104.2969	0.4098	25	44.7564	0.7964	0.0000	0.7817	0.25	0.7742	0.6029	0.1887
40.5057	0.4918	30	43.4222	0.7165	1.0000	0.9225	0.4773	0.8280	0.9556	0.7450
25.5625	0.5738	35	36.3869	0.7449	0.9284	0.7113	0.2727	0.7151	0.7522	0.3699
29.1538	0.6557	40	81.5410	0.6647	0.0000	0.6620	0.3409	0.6613	0.3386	0.1329
13.3752	0.7377	45	63.1843	0.6367	0.0023	0.5	0.2045	0.5699	0.2802	0.1056
463.5	0.8197	50	50.7853	0.5138	1.0000	0.9648	0.9318	0.7527	0.9999	1.0000
42.5	0.9016	55	54.6338	0.6184	0.0114	0.4225	0.2273	0.5054	0.2395	0.1134
12.8289	0.9836	60	19.5652	0.6101	1.0000	0.3662	0.1591	0.4785	0.9037	0.7798
0.1935	1.0656	65	30.2404	0.6681	0.0007	0.7535	0.4773	0.6989	0.4138	0.2378
4.4435	1.1475	70	13.6321	0.7592	0.8366	0.8028	0.3636	0.7634	0.8256	0.4446
7.6102	1.2295	75	13.2956	0.7647	0.0953	0.8521	0.3864	0.7957	0.7824	0.3570
15.7013	1.3115	80	23.0096	0.7222	1.0000	0.8380	0.4318	0.7742	0.9736	0.7859
0.0023	1.3934	85	25.6853	0.6942	1.0	0.6761	0.3409	0.6720	0.9491	0.7718
24.9375	1.4754	90	80.7625	0.5354	1.0	0.9789	0.9091	0.7688	0.9920	0.9318
0.0	1.5574	95	34.5263	0.6680	1.0000	0.7113	0.3864	0.6882	0.8553	0.6467
0.0	1.6393	100	35.8591	0.6363	1.0000	0.9155	0.6591	0.7796	0.9722	0.8453
0.5899	1.7213	105	31.2971	0.6339	1.0000	0.9155	0.6364	0.7849	0.9697	0.8148
0.0004	1.8033	110	37.4110	0.5983	1.0000	0.9155	0.7045	0.7688	0.9852	0.8992
0.0	1.8852	115	24.3760	0.6756	1.0000	0.8451	0.5227	0.7581	0.9484	0.7774
6.9062	1.9672	120	24.2837	0.6849	1.0000	0.8662	0.5227	0.7742	0.9534	0.8091
0.0	2.0492	125	37.4777	0.5973	1.0	0.9014	0.7045	0.7581	0.9896	0.9294
0.0	2.1311	130	35.9309	0.5984	1.0	0.8732	0.6818	0.7419	0.9708	0.9091
0.0972	2.2131	135	34.5928	0.6206	1.0	0.8732	0.6364	0.7527	0.9609	0.8865
0.0009	2.2951	140	39.1306	0.5689	1.0	0.9085	0.7727	0.7473	0.9819	0.9243
0.0062	2.3770	145	43.4752	0.5407	1.0000	0.9577	0.8636	0.7634	0.9869	0.9762
0.2646	2.4590	150	41.5047	0.5479	1.0000	0.9437	0.8409	0.7581	0.9782	0.9538
0.0	2.5410	155	41.3486	0.5591	1.0	0.9366	0.8182	0.7581	0.9719	0.9251
0.0	2.6230	160	42.7021	0.5514	1.0	0.9437	0.8409	0.7581	0.9789	0.9517
8.5	2.7049	165	44.1056	0.5405	1.0000	0.9507	0.8636	0.7581	0.9828	0.9553
0.0	2.7869	170	45.4856	0.5437	1.0	0.9507	0.8636	0.7581	0.9892	0.9610
0.0493	2.8689	175	46.0514	0.5329	1.0000	0.9648	0.8864	0.7634	0.9909	0.9650
0.0029	2.9508	180	45.9846	0.5331	1.0000	0.9718	0.8864	0.7688	0.9916	0.9679

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_lr5e-4_r8_2602051922

bosch_RM_mistralai_S130798_LLM_false_STRUCT_true_epo3_lr5e-4_r8_2602051922

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_lr5e-4_r8_2602051922