MeMo_BERT-WSD-DanBERT

This model is a fine-tuned version of alexanderfalk/danbert-small-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.1673
• F1-score: 0.3866

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	61	1.5325	0.1229
No log	2.0	122	1.5469	0.1229
No log	3.0	183	1.5848	0.1916
No log	4.0	244	1.7138	0.2595
No log	5.0	305	2.5576	0.1820
No log	6.0	366	2.5028	0.3250
No log	7.0	427	2.8110	0.2125
No log	8.0	488	3.2862	0.3449
0.7249	9.0	549	3.1673	0.3866
0.7249	10.0	610	4.1707	0.2961
0.7249	11.0	671	4.2567	0.3072
0.7249	12.0	732	4.0008	0.3608
0.7249	13.0	793	4.0726	0.3239
0.7249	14.0	854	4.2054	0.3091
0.7249	15.0	915	4.2703	0.3107
0.7249	16.0	976	4.2723	0.3014
0.0021	17.0	1037	4.2581	0.3014
0.0021	18.0	1098	4.2739	0.3014
0.0021	19.0	1159	4.2988	0.3014
0.0021	20.0	1220	4.3953	0.3107

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2