nb-bert-base_FGN

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

• Loss: 1.0904
• F1-score: 0.8074

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	120	0.6228	0.7307
No log	2.0	240	0.7442	0.7474
No log	3.0	360	0.7118	0.7785
No log	4.0	480	1.2081	0.7137
0.5388	5.0	600	1.1968	0.7628
0.5388	6.0	720	1.0904	0.8074
0.5388	7.0	840	1.2685	0.8007
0.5388	8.0	960	1.4070	0.7783
0.123	9.0	1080	1.6120	0.7608
0.123	10.0	1200	1.5899	0.7695
0.123	11.0	1320	1.4975	0.7705
0.123	12.0	1440	1.4624	0.7983
0.0475	13.0	1560	1.5148	0.7711
0.0475	14.0	1680	1.4680	0.7926
0.0475	15.0	1800	1.4216	0.8006
0.0475	16.0	1920	1.4962	0.8006
0.0201	17.0	2040	1.4150	0.7883
0.0201	18.0	2160	1.4259	0.7755
0.0201	19.0	2280	1.5040	0.7799
0.0201	20.0	2400	1.5045	0.7808

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1