xlm-roberta-base-ukr-finetuned

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: 0.2892
• F1: 0.5354
• Roc Auc: 0.7561
• Accuracy: 0.6225

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
0.3156	1.0	121	0.2864	0.0	0.5	0.4940
0.3014	2.0	242	0.2773	0.0078	0.5016	0.4940
0.279	3.0	363	0.2487	0.0958	0.5384	0.5261
0.2225	4.0	484	0.2313	0.2855	0.6115	0.5622
0.193	5.0	605	0.2181	0.3816	0.6592	0.6386
0.1321	6.0	726	0.2235	0.3669	0.6665	0.6386
0.1209	7.0	847	0.2270	0.4572	0.7031	0.6386
0.1023	8.0	968	0.2300	0.4030	0.6884	0.6506
0.0924	9.0	1089	0.2392	0.4278	0.6947	0.6667
0.0691	10.0	1210	0.2506	0.4498	0.7062	0.6426
0.0635	11.0	1331	0.2579	0.4643	0.7134	0.6225
0.0537	12.0	1452	0.2558	0.4755	0.7312	0.6345
0.044	13.0	1573	0.2784	0.4893	0.7246	0.6145
0.042	14.0	1694	0.2744	0.4892	0.7298	0.6305
0.0363	15.0	1815	0.2824	0.4976	0.7341	0.6225
0.0356	16.0	1936	0.2892	0.5354	0.7561	0.6225
0.0281	17.0	2057	0.2904	0.5271	0.7479	0.6305
0.0283	18.0	2178	0.2916	0.5071	0.7348	0.6345
0.0261	19.0	2299	0.2935	0.4995	0.7341	0.6305
0.0243	20.0	2420	0.2932	0.4995	0.7341	0.6305

Framework versions

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