xlm-roberta-base-swa-finetuned-RC

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

• Loss: 0.1652
• F1: 0.6196
• Roc Auc: 0.7843
• Accuracy: 0.7657

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
0.3064	1.0	166	0.3109	0.0	0.5	0.4535
0.3091	2.0	332	0.2913	0.0055	0.5013	0.4550
0.2823	3.0	498	0.2741	0.0566	0.5171	0.4701
0.2575	4.0	664	0.2694	0.1859	0.5657	0.5034
0.2513	5.0	830	0.2551	0.2475	0.5989	0.5283
0.2104	6.0	996	0.2473	0.2775	0.6160	0.5465
0.1827	7.0	1162	0.2317	0.3347	0.6367	0.5873
0.1572	8.0	1328	0.2249	0.4273	0.6836	0.6100
0.1431	9.0	1494	0.2194	0.4431	0.6948	0.6311
0.1326	10.0	1660	0.1990	0.4635	0.6931	0.6599
0.0972	11.0	1826	0.1994	0.4851	0.6984	0.6697
0.1022	12.0	1992	0.1880	0.5475	0.7437	0.7014
0.0886	13.0	2158	0.1845	0.5665	0.7634	0.7060
0.0794	14.0	2324	0.1738	0.5891	0.7635	0.7347
0.0689	15.0	2490	0.1690	0.6007	0.7687	0.7513
0.0658	16.0	2656	0.1688	0.6105	0.7782	0.7536
0.0551	17.0	2822	0.1674	0.6168	0.7806	0.7581
0.0585	18.0	2988	0.1643	0.6153	0.7800	0.7604
0.0519	19.0	3154	0.1652	0.6196	0.7843	0.7657
0.0546	20.0	3320	0.1646	0.6146	0.7799	0.7611

Framework versions

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