MeMo_BERT-SA_botxo

This model is a fine-tuned version of Maltehb/danish-bert-botxo on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6385
• F1-score: 0.7374

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	297	0.7785	0.6992
0.7119	2.0	594	0.9146	0.6732
0.7119	3.0	891	1.4686	0.7087
0.2712	4.0	1188	1.5626	0.7091
0.2712	5.0	1485	1.9710	0.6548
0.09	6.0	1782	2.0442	0.7169
0.0273	7.0	2079	2.1153	0.7326
0.0273	8.0	2376	2.3741	0.7005
0.0071	9.0	2673	2.6151	0.6977
0.0071	10.0	2970	2.5560	0.7107
0.0112	11.0	3267	2.5697	0.7220
0.0063	12.0	3564	2.5956	0.7185
0.0063	13.0	3861	2.6383	0.7189
0.0001	14.0	4158	2.9775	0.6755
0.0001	15.0	4455	2.6557	0.7247
0.0044	16.0	4752	2.6788	0.7305
0.0023	17.0	5049	2.6385	0.7374
0.0023	18.0	5346	2.6750	0.7186
0.0025	19.0	5643	2.6821	0.7186
0.0025	20.0	5940	2.6820	0.7186

Framework versions

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