test

This model is a fine-tuned version of nielsr/lilt-xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6605
• Precision: 0.7460
• Recall: 0.7692
• F1: 0.7575
• Accuracy: 0.7526

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.3333	100	1.0907	0.4971	0.6210	0.5522	0.5889
No log	2.6667	200	0.7374	0.6135	0.6857	0.6476	0.7475
No log	4.0	300	0.8119	0.6292	0.7193	0.6713	0.7490
No log	5.3333	400	0.8152	0.6930	0.7555	0.7229	0.7616
0.6197	6.6667	500	0.9915	0.6824	0.7682	0.7227	0.7458
0.6197	8.0	600	1.0589	0.6952	0.7809	0.7356	0.7680
0.6197	9.3333	700	1.1514	0.7072	0.7285	0.7177	0.7456
0.6197	10.6667	800	1.1828	0.7190	0.7652	0.7414	0.7625
0.6197	12.0	900	1.2011	0.7301	0.7606	0.7450	0.7679
0.0998	13.3333	1000	1.2323	0.7347	0.7662	0.7501	0.7622
0.0998	14.6667	1100	1.3060	0.7413	0.7881	0.7640	0.7688
0.0998	16.0	1200	1.3649	0.7337	0.7636	0.7484	0.7647
0.0998	17.3333	1300	1.3661	0.7319	0.7789	0.7547	0.7685
0.0998	18.6667	1400	1.4831	0.7386	0.7672	0.7526	0.7635
0.0226	20.0	1500	1.4216	0.7299	0.7682	0.7486	0.7654
0.0226	21.3333	1600	1.5146	0.7295	0.7733	0.7507	0.7539
0.0226	22.6667	1700	1.6595	0.7398	0.7748	0.7569	0.7476
0.0226	24.0	1800	1.5785	0.7609	0.7702	0.7656	0.7677
0.0226	25.3333	1900	1.5824	0.7544	0.7886	0.7711	0.7587
0.0057	26.6667	2000	1.6605	0.7460	0.7692	0.7575	0.7526
0.0057	28.0	2100	1.6459	0.7396	0.7697	0.7544	0.7520
0.0057	29.3333	2200	1.6605	0.7467	0.7748	0.7605	0.7541

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cpu
• Datasets 3.0.0
• Tokenizers 0.19.1