0b3e2c86ac7ec4fd20b451d1710dfd27

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

• Loss: 0.9243
• Data Size: 1.0
• Epoch Runtime: 16.2356
• Accuracy: 0.6688
• F1 Macro: 0.6381
• Rouge1: 0.6688
• Rouge2: 0.0
• Rougel: 0.6685
• Rougelsum: 0.6685

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	0.6688	0	2.0146	0.6198	0.3872	0.6198	0.0	0.6195	0.6198
No log	1	294	0.6746	0.0078	2.4462	0.6164	0.3980	0.6161	0.0	0.6158	0.6164
No log	2	588	0.6738	0.0156	2.5972	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
No log	3	882	0.6887	0.0312	2.8015	0.5184	0.5075	0.5184	0.0	0.5185	0.5184
0.0279	4	1176	0.6695	0.0625	3.1927	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0557	5	1470	0.6649	0.125	4.1918	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0972	6	1764	0.6659	0.25	5.9000	0.6268	0.4362	0.6268	0.0	0.6265	0.6268
0.6639	7	2058	0.6532	0.5	9.3830	0.6320	0.4609	0.6320	0.0	0.6311	0.6320
0.6322	8.0	2352	0.6489	1.0	16.6337	0.6265	0.5733	0.6268	0.0	0.6265	0.6268
0.5987	9.0	2646	0.6842	1.0	16.1903	0.6645	0.5890	0.6648	0.0	0.6642	0.6645
0.5089	10.0	2940	0.7818	1.0	16.8641	0.6342	0.6260	0.6345	0.0	0.6342	0.6342
0.3893	11.0	3234	0.7896	1.0	16.2934	0.6703	0.6377	0.6706	0.0	0.6703	0.6703
0.284	12.0	3528	0.9243	1.0	16.2356	0.6688	0.6381	0.6688	0.0	0.6685	0.6685

Framework versions

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