022d58726f6384a5f5a95ae41f261f3f

This model is a fine-tuned version of FacebookAI/xlm-roberta-large-finetuned-conll02-spanish on the contemmcm/cls_20newsgroups dataset. It achieves the following results on the evaluation set:

• Loss: 0.5181
• Data Size: 1.0
• Epoch Runtime: 88.9296
• Accuracy: 0.8687
• F1 Macro: 0.8689

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	3.0673	0	6.0318	0.0504	0.0062
No log	1	499	3.0293	0.0078	6.8250	0.0464	0.0077
0.0307	2	998	3.0015	0.0156	8.0503	0.0512	0.0049
0.0556	3	1497	2.7525	0.0312	10.0024	0.1162	0.0531
0.0988	4	1996	1.9302	0.0625	13.2104	0.3251	0.2555
1.7684	5	2495	1.4205	0.125	19.1066	0.4957	0.4238
1.1607	6	2994	1.0011	0.25	29.9067	0.6996	0.6688
0.8565	7	3493	0.9473	0.5	50.4729	0.7404	0.7353
0.6241	8.0	3992	0.6276	1.0	89.7303	0.8183	0.8153
0.5687	9.0	4491	0.6575	1.0	89.0207	0.8160	0.8095
0.429	10.0	4990	0.5353	1.0	89.0179	0.8385	0.8330
0.3451	11.0	5489	0.5077	1.0	90.2628	0.8543	0.8533
0.3589	12.0	5988	0.4468	1.0	89.2016	0.8707	0.8700
0.3712	13.0	6487	0.5230	1.0	88.7512	0.8722	0.8715
0.3405	14.0	6986	0.5255	1.0	88.3086	0.8632	0.8628
0.2699	15.0	7485	0.5308	1.0	87.9859	0.8682	0.8667
0.3075	16.0	7984	0.5181	1.0	88.9296	0.8687	0.8689

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1