UIT-NO-PRExlnet-large-cased-finetuned

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

• Loss: 0.6311
• F1: 0.7534
• Roc Auc: 0.8047
• Accuracy: 0.5018

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.6008	1.0	139	0.5907	0.0994	0.5061	0.1227
0.5594	2.0	278	0.5834	0.1435	0.5	0.1300
0.4458	3.0	417	0.3967	0.6474	0.7336	0.4007
0.3153	4.0	556	0.3647	0.7128	0.7775	0.4495
0.2474	5.0	695	0.3392	0.7382	0.7952	0.4693
0.1915	6.0	834	0.3702	0.7346	0.7980	0.5054
0.1194	7.0	973	0.4083	0.7340	0.7994	0.4982
0.0953	8.0	1112	0.4656	0.7507	0.8101	0.4910
0.0482	9.0	1251	0.5682	0.7438	0.7934	0.4838
0.0504	10.0	1390	0.5374	0.7419	0.8069	0.4729
0.0265	11.0	1529	0.6019	0.7408	0.8011	0.4838
0.0082	12.0	1668	0.6136	0.7429	0.8015	0.4874
0.0077	13.0	1807	0.6212	0.7461	0.8020	0.4982
0.0117	14.0	1946	0.6089	0.7519	0.8086	0.4928
0.0044	15.0	2085	0.6246	0.7508	0.8050	0.5
0.0041	16.0	2224	0.6382	0.7460	0.8005	0.4946
0.0024	17.0	2363	0.6333	0.7467	0.8011	0.5
0.0053	18.0	2502	0.6311	0.7534	0.8047	0.5018
0.0027	19.0	2641	0.6311	0.7508	0.8032	0.5036
0.0029	20.0	2780	0.6318	0.7520	0.8037	0.5054

Framework versions

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