Revision_11_03_XLM_Lexical_MetaXLM_Q1-52K

This model is a fine-tuned version of phunganhsang/Revision_XLMRoBERTa_Lexical_Dataset_52k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6261
• Accuracy: 0.6702
• F1: 0.6700

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 OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.2033	100	0.5391	0.8042	0.8011
No log	0.4065	200	0.6216	0.7874	0.7850
No log	0.6098	300	0.7804	0.7932	0.7910
No log	0.8130	400	0.9558	0.7129	0.7128
0.2247	1.0163	500	1.9012	0.5537	0.5443
0.2247	1.2195	600	2.3194	0.5565	0.5471
0.2247	1.4228	700	2.2556	0.5536	0.5439
0.2247	1.6260	800	2.2496	0.5725	0.5656
0.2247	1.8293	900	2.7606	0.5464	0.5355
0.0731	2.0325	1000	2.9351	0.5569	0.5477
0.0731	2.2358	1100	2.4703	0.6671	0.6668
0.0731	2.4390	1200	2.3544	0.6629	0.6625
0.0731	2.6423	1300	2.2919	0.6682	0.6679
0.0731	2.8455	1400	2.0367	0.7121	0.7120
0.0361	3.0488	1500	2.5626	0.6488	0.6480
0.0361	3.2520	1600	2.7892	0.6435	0.6424
0.0361	3.4553	1700	2.5784	0.6623	0.6619
0.0361	3.6585	1800	2.7657	0.6436	0.6425
0.0361	3.8618	1900	2.3433	0.6884	0.6884
0.0232	4.0650	2000	2.4573	0.6932	0.6932
0.0232	4.2683	2100	2.8929	0.6499	0.6492
0.0232	4.4715	2200	2.6676	0.6660	0.6657
0.0232	4.6748	2300	2.6564	0.6681	0.6678
0.0232	4.8780	2400	2.6261	0.6702	0.6700

Framework versions

• Transformers 5.3.0
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2