acb1a7ea85ac9779d2da7d5853fe0f82

This model is a fine-tuned version of google-bert/bert-base-german-dbmdz-cased on the contemmcm/amazon_reviews_2013 [cell-phone] dataset. It achieves the following results on the evaluation set:

• Loss: 0.9231
• Data Size: 1.0
• Epoch Runtime: 107.2634
• Accuracy: 0.6672
• F1 Macro: 0.5926

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	1.6855	0	7.3280	0.1815	0.1099
No log	1	1973	1.5024	0.0078	8.6292	0.3036	0.1516
0.0318	2	3946	1.3125	0.0156	8.7919	0.4701	0.2408
1.192	3	5919	1.1183	0.0312	10.4265	0.5381	0.3511
1.0552	4	7892	1.0130	0.0625	13.7176	0.5757	0.4726
0.9694	5	9865	0.9311	0.125	19.9050	0.6147	0.4945
0.9042	6	11838	0.8656	0.25	32.9112	0.6429	0.5302
0.8763	7	13811	0.8349	0.5	57.9569	0.6461	0.5643
0.7884	8.0	15784	0.8012	1.0	108.7645	0.6646	0.5911
0.7013	9.0	17757	0.8179	1.0	107.0642	0.6765	0.5908
0.6005	10.0	19730	0.8713	1.0	107.7805	0.6511	0.5953
0.5411	11.0	21703	0.9613	1.0	109.1855	0.6551	0.5817
0.4842	12.0	23676	0.9231	1.0	107.2634	0.6672	0.5926

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1