xlm-roberta-base-hin-finetuned-augmentation-LUNAR

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

• Loss: 0.1980
• F1: 0.8479
• Roc Auc: 0.9212
• Accuracy: 0.7626

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.4151	1.0	207	0.4164	0.0	0.5	0.2142
0.2692	2.0	414	0.2547	0.6165	0.7649	0.5962
0.1741	3.0	621	0.1773	0.7668	0.8547	0.6999
0.1411	4.0	828	0.1631	0.8063	0.8795	0.7394
0.092	5.0	1035	0.1541	0.8079	0.8705	0.7531
0.0756	6.0	1242	0.1488	0.8303	0.9012	0.7640
0.0559	7.0	1449	0.1637	0.8352	0.9097	0.7476
0.0509	8.0	1656	0.1528	0.8254	0.8952	0.7531
0.0383	9.0	1863	0.1796	0.8258	0.9086	0.7544
0.0305	10.0	2070	0.1691	0.8373	0.9073	0.7653
0.0253	11.0	2277	0.1713	0.8361	0.9089	0.7640
0.0219	12.0	2484	0.1906	0.8353	0.9144	0.7558
0.0169	13.0	2691	0.1902	0.8348	0.9138	0.7517
0.0173	14.0	2898	0.1868	0.8422	0.9117	0.7667
0.0191	15.0	3105	0.1974	0.8423	0.9193	0.7572
0.0085	16.0	3312	0.1962	0.8434	0.9211	0.7626
0.0072	17.0	3519	0.1953	0.8452	0.9187	0.7626
0.0091	18.0	3726	0.1961	0.8466	0.9226	0.7626
0.0066	19.0	3933	0.1980	0.8473	0.9211	0.7626
0.0075	20.0	4140	0.1980	0.8479	0.9212	0.7626

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0