MeMo_BERT-WSD-DanskBERT

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

• Loss: 1.7755
• F1-score: 0.5209

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	1.4766	0.1229
No log	2.0	122	1.4366	0.1229
No log	3.0	183	1.3636	0.2462
No log	4.0	244	1.2889	0.3692
No log	5.0	305	1.4150	0.3786
No log	6.0	366	1.5581	0.3409
No log	7.0	427	1.6512	0.4664
No log	8.0	488	1.7405	0.4661
0.9424	9.0	549	1.7755	0.5209
0.9424	10.0	610	2.4738	0.4351
0.9424	11.0	671	2.4721	0.4858
0.9424	12.0	732	2.9449	0.4491
0.9424	13.0	793	2.8346	0.4528
0.9424	14.0	854	3.0715	0.4845
0.9424	15.0	915	3.1416	0.4520
0.9424	16.0	976	3.0893	0.5197
0.1197	17.0	1037	3.1668	0.4764
0.1197	18.0	1098	3.2142	0.4656
0.1197	19.0	1159	3.2174	0.5087
0.1197	20.0	1220	3.2239	0.5087

Framework versions

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