nb-bert-edu-scorer-swe

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: 0.9269
• Precision: 0.3790
• Recall: 0.3418
• F1 Macro: 0.3416
• Accuracy: 0.4269

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: 3e-05
• train_batch_size: 512
• eval_batch_size: 256
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1 Macro	Accuracy
No log	0	0	5.1598	0.0513	0.1667	0.0785	0.3080
1.0761	5.4348	1000	1.0157	0.3984	0.3359	0.3328	0.4098
1.0442	10.8696	2000	1.0227	0.4107	0.3519	0.3521	0.4076
1.0356	16.3043	3000	1.0053	0.4106	0.3551	0.3552	0.4171

Framework versions

• Transformers 4.55.0
• Pytorch 2.5.1+cu121
• Datasets 4.0.0
• Tokenizers 0.21.4