8554aa066133e9b4f3f165f6738b29d0

This model is a fine-tuned version of google-bert/bert-base-german-dbmdz-uncased on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 0.9226
• Data Size: 1.0
• Epoch Runtime: 15.3599
• Accuracy: 0.6884
• F1 Macro: 0.6635
• Rouge1: 0.6887
• Rouge2: 0.0
• Rougel: 0.6881
• Rougelsum: 0.6881

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	0.7818	0	2.0160	0.4602	0.4552	0.4608	0.0	0.4602	0.4605
No log	1	294	0.7006	0.0078	2.3780	0.6201	0.3873	0.6201	0.0	0.6195	0.6201
No log	2	588	0.6709	0.0156	2.4774	0.5956	0.4693	0.5956	0.0	0.5953	0.5956
No log	3	882	0.6684	0.0312	2.7330	0.6173	0.4271	0.6173	0.0	0.6167	0.6173
0.0277	4	1176	0.6683	0.0625	3.1122	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0555	5	1470	0.6734	0.125	4.0629	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0949	6	1764	0.6636	0.25	6.0339	0.6320	0.4502	0.6320	0.0	0.6314	0.6320
0.6385	7	2058	0.6472	0.5	9.2371	0.6308	0.6131	0.6311	0.0	0.6305	0.6310
0.6051	8.0	2352	0.6039	1.0	16.0590	0.6829	0.6493	0.6832	0.0	0.6826	0.6832
0.4912	9.0	2646	0.6524	1.0	15.6518	0.6927	0.6707	0.6927	0.0	0.6921	0.6930
0.3608	10.0	2940	0.7361	1.0	16.4633	0.6991	0.6710	0.6991	0.0	0.6988	0.6988
0.2591	11.0	3234	0.8189	1.0	15.7975	0.6915	0.6740	0.6915	0.0	0.6915	0.6912
0.1806	12.0	3528	0.9226	1.0	15.3599	0.6884	0.6635	0.6887	0.0	0.6881	0.6881

Framework versions

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