classifier-de2

This model is a fine-tuned version of bert-base-german-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3294
• Accuracy: 0.8826
• Precision: 0.5399
• Recall: 0.3576
• F1: 0.4302

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: 1.5e-05
• train_batch_size: 256
• eval_batch_size: 256
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.2757	0.0923	900	0.3526	0.8732	0.4426	0.0879	0.1466
0.2537	0.1845	1800	0.3498	0.8739	0.4782	0.1823	0.2640
0.2242	0.2768	2700	0.3381	0.8815	0.5739	0.1712	0.2637
0.2061	0.3690	3600	0.3430	0.8763	0.5022	0.2519	0.3355
0.1914	0.4613	4500	0.3435	0.8784	0.5202	0.2482	0.3360
0.1798	0.5535	5400	0.3240	0.8817	0.5554	0.2291	0.3243
0.1899	0.6458	6300	0.3206	0.8768	0.5052	0.3153	0.3883
0.1761	0.7380	7200	0.3340	0.8846	0.5955	0.2170	0.3181
0.189	0.8303	8100	0.3241	0.8860	0.6141	0.2160	0.3196
0.1644	0.9225	9000	0.3278	0.8861	0.6105	0.2251	0.3289
0.1582	1.0148	9900	0.3437	0.8847	0.5773	0.2633	0.3616
0.1511	1.1070	10800	0.3187	0.8836	0.5556	0.3076	0.3960
0.1602	1.1993	11700	0.3198	0.8860	0.5858	0.2764	0.3756
0.149	1.2915	12600	0.3244	0.8842	0.5635	0.2945	0.3868
0.1512	1.3838	13500	0.3281	0.8863	0.5792	0.3040	0.3987
0.1463	1.4760	14400	0.3228	0.8869	0.5947	0.2753	0.3763
0.1372	1.5683	15300	0.3300	0.8872	0.5869	0.3048	0.4012
0.1545	1.6605	16200	0.3229	0.8866	0.5807	0.3086	0.4030
0.1755	1.7528	17100	0.3070	0.8854	0.5652	0.3280	0.4151
0.1403	1.8450	18000	0.3212	0.8877	0.5995	0.2836	0.3851
0.1425	1.9373	18900	0.3179	0.8861	0.5722	0.3235	0.4133
0.1271	2.0295	19800	0.3483	0.8843	0.5545	0.3411	0.4224
0.1235	2.1218	20700	0.3362	0.8858	0.5685	0.3294	0.4171
0.1324	2.2140	21600	0.3294	0.8826	0.5399	0.3576	0.4302
0.1236	2.3063	22500	0.3345	0.8859	0.5712	0.3214	0.4113
0.1264	2.3985	23400	0.3575	0.8876	0.5879	0.3141	0.4094
0.1157	2.4908	24300	0.3405	0.8872	0.5863	0.3058	0.4020
0.1261	2.5830	25200	0.3372	0.8874	0.5853	0.3165	0.4109
0.1346	2.6753	26100	0.3398	0.8863	0.5747	0.3205	0.4115
0.1099	2.7675	27000	0.3492	0.8872	0.5843	0.3122	0.4070
0.1295	2.8598	27900	0.3374	0.8871	0.5813	0.3191	0.4120
0.1259	2.9520	28800	0.3410	0.8875	0.5863	0.3152	0.4100

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu126
• Datasets 3.5.0
• Tokenizers 0.21.1