BantuBERTa-vmw-finetuned

This model is a fine-tuned version of dsfsi/BantuBERTa on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3269
• F1: 0.1633
• Roc Auc: 0.5507
• Accuracy: 0.4729

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.4095	1.0	66	0.3112	0.0	0.5	0.4612
0.2908	2.0	132	0.2903	0.0	0.5	0.4612
0.284	3.0	198	0.2882	0.0	0.5	0.4612
0.2958	4.0	264	0.2879	0.0	0.5	0.4612
0.2749	5.0	330	0.2851	0.0139	0.5033	0.4651
0.265	6.0	396	0.2868	0.0574	0.5173	0.4729
0.2281	7.0	462	0.2802	0.1073	0.5338	0.4845
0.1963	8.0	528	0.2876	0.1109	0.5353	0.4651
0.1749	9.0	594	0.3040	0.0966	0.5282	0.4496
0.1534	10.0	660	0.3046	0.1322	0.5396	0.4612
0.1336	11.0	726	0.3129	0.1602	0.5473	0.4690
0.1194	12.0	792	0.3222	0.1287	0.5372	0.4612
0.1106	13.0	858	0.3269	0.1633	0.5507	0.4729
0.1004	14.0	924	0.3296	0.1269	0.5360	0.4496
0.091	15.0	990	0.3364	0.1612	0.5513	0.4690
0.0855	16.0	1056	0.3347	0.1570	0.5472	0.4690
0.079	17.0	1122	0.3399	0.1513	0.5440	0.4574

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0