917fdbf5d00514c6d528865bcd93a5b6

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

• Loss: 0.0744
• Data Size: 1.0
• Epoch Runtime: 34.9024
• Accuracy: 0.9869
• F1 Macro: 0.9862

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
No log	0	0	0.5977	0	3.0801	0.7832	0.7724
No log	1	650	0.3101	0.0078	3.7643	0.9352	0.9309
No log	2	1300	0.1451	0.0156	3.7222	0.9537	0.9518
No log	3	1950	0.0853	0.0312	4.4347	0.9713	0.9698
No log	4	2600	0.2011	0.0625	5.5514	0.9516	0.9479
0.0077	5	3250	0.1233	0.125	7.4980	0.9608	0.9579
0.0599	6	3900	0.0473	0.25	11.4473	0.9850	0.9842
0.0481	7	4550	0.0566	0.5	19.1742	0.9830	0.9820
0.0303	8.0	5200	0.0627	1.0	35.7040	0.9882	0.9876
0.0205	9.0	5850	0.0506	1.0	34.6314	0.9898	0.9892
0.0153	10.0	6500	0.0744	1.0	34.9024	0.9869	0.9862

Framework versions

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