f4fd5dda367dfd06ec1c996fb8ed06f6

This model is a fine-tuned version of distilbert/distilbert-base-german-cased on the dair-ai/emotion [split] dataset. It achieves the following results on the evaluation set:

• Loss: 0.2399
• Data Size: 1.0
• Epoch Runtime: 15.2043
• Accuracy: 0.9148
• F1 Macro: 0.8725

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	1.7970	0	1.1572	0.0398	0.0199
No log	1	500	1.6278	0.0078	1.4528	0.2893	0.0911
No log	2	1000	1.5845	0.0156	1.5784	0.3488	0.0862
No log	3	1500	1.5594	0.0312	1.8853	0.3488	0.0862
No log	4	2000	1.5592	0.0625	2.2634	0.3740	0.1336
0.0851	5	2500	1.4394	0.125	2.9597	0.4577	0.1895
1.1987	6	3000	0.8622	0.25	4.6902	0.6930	0.4681
0.0663	7	3500	0.3319	0.5	8.2457	0.8831	0.8339
0.2041	8.0	4000	0.2181	1.0	14.8055	0.9118	0.8681
0.1628	9.0	4500	0.2189	1.0	14.7299	0.9153	0.8735
0.1501	10.0	5000	0.2255	1.0	14.9658	0.9209	0.8811
0.1014	11.0	5500	0.2331	1.0	14.8122	0.9199	0.8748
0.1257	12.0	6000	0.2399	1.0	15.2043	0.9148	0.8725

Framework versions

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