xlm-roberta-base-pcm-noaug

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: 0.5261
• F1: 0.4912
• Roc Auc: 0.6688
• Accuracy: 0.3065

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
0.4984	1.0	117	0.4862	0.1155	0.5131	0.2081
0.456	2.0	234	0.4323	0.2110	0.5565	0.2145
0.4207	3.0	351	0.4142	0.2856	0.5909	0.2903
0.3932	4.0	468	0.4189	0.3215	0.6145	0.2984
0.3576	5.0	585	0.4417	0.3400	0.6254	0.3048
0.324	6.0	702	0.4158	0.4024	0.6332	0.3129
0.3011	7.0	819	0.4176	0.4393	0.6516	0.3016
0.2675	8.0	936	0.4433	0.4546	0.6663	0.3290
0.2436	9.0	1053	0.4513	0.4435	0.6547	0.3258
0.2169	10.0	1170	0.4674	0.4624	0.6641	0.3177
0.2241	11.0	1287	0.4843	0.4706	0.6649	0.2984
0.1853	12.0	1404	0.4866	0.4646	0.6601	0.3323
0.1751	13.0	1521	0.5068	0.4555	0.6557	0.3081
0.1596	14.0	1638	0.4991	0.4640	0.6571	0.3145
0.1458	15.0	1755	0.5174	0.4784	0.6667	0.3210
0.1446	16.0	1872	0.5261	0.4912	0.6688	0.3065
0.1502	17.0	1989	0.5211	0.4876	0.6669	0.3242
0.1309	18.0	2106	0.5219	0.4823	0.6646	0.3242
0.1419	19.0	2223	0.5247	0.4767	0.6626	0.3258
0.138	20.0	2340	0.5241	0.4793	0.6631	0.3242

Framework versions

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