yemen2016/danskbert_ED2

danskbert_ED1

This model is a fine-tuned version of vesteinn/DanskBERT on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5087
• F1-score: 0.9098

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1-score
No log	1.0	69	0.4066	0.8196
No log	2.0	138	0.3546	0.8846
No log	3.0	207	0.4711	0.9013
No log	4.0	276	0.4296	0.8932
No log	5.0	345	0.7490	0.8513
No log	6.0	414	0.5087	0.9098
No log	7.0	483	0.8659	0.8669
0.2398	8.0	552	0.8049	0.8601
0.2398	9.0	621	0.9197	0.8678
0.2398	10.0	690	0.9867	0.8678
0.2398	11.0	759	0.9999	0.8681
0.2398	12.0	828	0.8314	0.8852
0.2398	13.0	897	0.8388	0.8934
0.2398	14.0	966	0.8358	0.9016
0.0092	15.0	1035	0.8468	0.9016
0.0092	16.0	1104	0.8554	0.9016
0.0092	17.0	1173	0.8654	0.8933
0.0092	18.0	1242	0.8707	0.8933
0.0092	19.0	1311	0.8731	0.8933
0.0092	20.0	1380	0.8740	0.8933

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1