b8c7c4b659ea4fcdb67eaf9175def7bb

This model is a fine-tuned version of google-bert/bert-base-german-dbmdz-cased on the nyu-mll/glue [mrpc] dataset. It achieves the following results on the evaluation set:

• Loss: 0.8612
• Data Size: 1.0
• Epoch Runtime: 6.8322
• Accuracy: 0.7783
• F1 Macro: 0.7466
• Rouge1: 0.7783
• Rouge2: 0.0
• Rougel: 0.7783
• Rougelsum: 0.7783

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6784	0	1.4683	0.5843	0.5072	0.5843	0.0	0.5843	0.5837
No log	1	114	0.6534	0.0078	1.6914	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	2	228	0.6605	0.0156	1.5900	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	3	342	0.6862	0.0312	2.0467	0.4463	0.4445	0.4469	0.0	0.4458	0.4463
0.0223	4	456	0.6167	0.0625	2.3077	0.6840	0.4875	0.6840	0.0	0.6834	0.6840
0.0223	5	570	0.5570	0.125	2.8103	0.7134	0.6414	0.7134	0.0	0.7134	0.7134
0.0223	6	684	0.4778	0.25	3.4063	0.7571	0.7285	0.7571	0.0	0.7571	0.7577
0.1346	7	798	0.4368	0.5	4.4745	0.7824	0.7537	0.7824	0.0	0.7827	0.7824
0.4096	8.0	912	0.4504	1.0	7.2289	0.7978	0.7602	0.7978	0.0	0.7978	0.7978
0.2573	9.0	1026	0.5179	1.0	7.0306	0.7907	0.7683	0.7907	0.0	0.7913	0.7907
0.1592	10.0	1140	0.7360	1.0	6.8460	0.7836	0.7381	0.7842	0.0	0.7836	0.7836
0.1818	11.0	1254	0.8612	1.0	6.8322	0.7783	0.7466	0.7783	0.0	0.7783	0.7783

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1