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: 1.6527
• Precision: 0.7393
• Recall: 0.7759
• F1: 0.7571
• Accuracy: 0.7614

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	0.9776	0.4894	0.6113	0.5436	0.6166
No log	2.6667	200	0.8649	0.6249	0.6296	0.6273	0.7231
No log	4.0	300	0.8745	0.6449	0.7392	0.6888	0.7326
No log	5.3333	400	0.9419	0.6292	0.7168	0.6702	0.7367
0.6362	6.6667	500	0.9902	0.7090	0.7458	0.7269	0.7700
0.6362	8.0	600	1.0048	0.7050	0.7315	0.7180	0.7614
0.6362	9.3333	700	1.1327	0.6918	0.7305	0.7106	0.7568
0.6362	10.6667	800	1.3954	0.6952	0.7366	0.7153	0.7333
0.6362	12.0	900	1.2721	0.7002	0.7509	0.7247	0.7491
0.1105	13.3333	1000	1.3422	0.7166	0.7356	0.7260	0.7521
0.1105	14.6667	1100	1.3957	0.72	0.7427	0.7312	0.7605
0.1105	16.0	1200	1.4581	0.7250	0.7519	0.7382	0.7608
0.1105	17.3333	1300	1.4598	0.7459	0.7371	0.7415	0.7577
0.1105	18.6667	1400	1.5542	0.7182	0.7504	0.7339	0.7497
0.0271	20.0	1500	1.5411	0.7176	0.7779	0.7465	0.7549
0.0271	21.3333	1600	1.6468	0.7251	0.7539	0.7393	0.7452
0.0271	22.6667	1700	1.6821	0.7311	0.7550	0.7429	0.7487
0.0271	24.0	1800	1.6220	0.7364	0.7499	0.7431	0.7585
0.0271	25.3333	1900	1.6220	0.7403	0.7667	0.7533	0.7599
0.0055	26.6667	2000	1.6161	0.7370	0.7682	0.7523	0.7687
0.0055	28.0	2100	1.6683	0.7396	0.7713	0.7551	0.7599
0.0055	29.3333	2200	1.6527	0.7393	0.7759	0.7571	0.7614

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1