41a80b51aa5104087f070a8dcad8f9af

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

• Loss: 0.9280
• Data Size: 1.0
• Epoch Runtime: 60.4610
• Accuracy: 0.6542
• F1 Macro: 0.5708

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	1.6122	0	4.7762	0.1858	0.0627
No log	1	1973	1.4683	0.0078	5.3407	0.3853	0.1112
0.032	2	3946	1.4560	0.0156	5.8536	0.3853	0.1112
1.4282	3	5919	1.3611	0.0312	6.7241	0.4229	0.1819
1.2476	4	7892	1.1574	0.0625	8.5589	0.5220	0.3782
1.0309	5	9865	0.9788	0.125	12.0983	0.5970	0.4576
0.9513	6	11838	0.9173	0.25	19.2222	0.6185	0.5002
0.9202	7	13811	0.8577	0.5	33.3267	0.6401	0.5643
0.8319	8.0	15784	0.8319	1.0	60.6466	0.6529	0.5819
0.7423	9.0	17757	0.8373	1.0	60.1827	0.6594	0.5808
0.6419	10.0	19730	0.8568	1.0	60.8935	0.6535	0.5830
0.5818	11.0	21703	0.9340	1.0	61.0317	0.6603	0.5737
0.5256	12.0	23676	0.9280	1.0	60.4610	0.6542	0.5708

Framework versions

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