UIT-NO-PREroberta-base-finetuned

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

• Loss: 0.6381
• F1: 0.7484
• Roc Auc: 0.8089
• Accuracy: 0.4819

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: 16
• eval_batch_size: 16
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.5255	1.0	139	0.4719	0.4630	0.6487	0.3177
0.4113	2.0	278	0.3648	0.6966	0.7646	0.4513
0.3089	3.0	417	0.3617	0.6980	0.7689	0.4422
0.1934	4.0	556	0.3952	0.7089	0.7656	0.4458
0.1829	5.0	695	0.3931	0.7257	0.7871	0.4549
0.1347	6.0	834	0.4276	0.6949	0.7681	0.4404
0.1027	7.0	973	0.4205	0.7311	0.7935	0.4621
0.0771	8.0	1112	0.4617	0.7286	0.7917	0.4567
0.0642	9.0	1251	0.4679	0.7355	0.8056	0.4567
0.0493	10.0	1390	0.5254	0.7186	0.7834	0.4549
0.0339	11.0	1529	0.5343	0.7250	0.7909	0.4621
0.0272	12.0	1668	0.5412	0.7245	0.7856	0.4747
0.0225	13.0	1807	0.5775	0.7319	0.7936	0.4621
0.0311	14.0	1946	0.5828	0.7440	0.8056	0.4747
0.0091	15.0	2085	0.5922	0.7351	0.7978	0.4711
0.0077	16.0	2224	0.6233	0.7254	0.7889	0.4711
0.0075	17.0	2363	0.6304	0.7277	0.7909	0.4765
0.0047	18.0	2502	0.6235	0.7335	0.7996	0.4765
0.0041	19.0	2641	0.6322	0.7405	0.8015	0.4747
0.0036	20.0	2780	0.6420	0.7368	0.7982	0.4711
0.0033	21.0	2919	0.6381	0.7484	0.8089	0.4819
0.003	22.0	3058	0.6516	0.7438	0.8056	0.4747
0.003	23.0	3197	0.6617	0.7379	0.7993	0.4675
0.0028	24.0	3336	0.6647	0.7422	0.8032	0.4819
0.0028	25.0	3475	0.6717	0.7414	0.8016	0.4747
0.003	26.0	3614	0.6629	0.7406	0.8026	0.4711
0.003	27.0	3753	0.6657	0.7440	0.8038	0.4747
0.0026	28.0	3892	0.6662	0.7443	0.8039	0.4693
0.0033	29.0	4031	0.6673	0.7437	0.8034	0.4711
0.0026	30.0	4170	0.6673	0.7437	0.8034	0.4711

Framework versions

• Transformers 4.48.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.21.0