trueparagraph.ai-xlnet

This model is a fine-tuned version of xlnet-base-cased on the None dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.8951
• F1: 0.8984
• Precision: 0.8674
• Recall: 0.9316
• Mcc: 0.7924
• Roc Auc: 0.8952
• Pr Auc: 0.8421
• Log Loss: 1.8813
• Loss: 0.2913

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Accuracy	F1	Precision	Recall	Mcc	Roc Auc	Pr Auc	Log Loss	Validation Loss
0.649	0.6297	500	0.8006	0.8195	0.7457	0.9095	0.6164	0.8010	0.7233	4.0119	0.4063
0.4104	1.2594	1000	0.8409	0.8294	0.8892	0.7772	0.6870	0.8406	0.8020	2.4398	0.4054
0.4101	1.8892	1500	0.8100	0.8359	0.7332	0.9722	0.6560	0.8107	0.7266	3.4982	0.4405
0.4046	2.5189	2000	0.7754	0.8120	0.6959	0.9747	0.6012	0.7762	0.6909	3.0282	0.5111
0.3992	3.1486	2500	0.8664	0.8625	0.8843	0.8418	0.7336	0.8663	0.8232	2.7164	0.3871
0.3691	3.7783	3000	0.8774	0.8850	0.8303	0.9475	0.7626	0.8777	0.8128	1.8936	0.3413
0.2581	4.4081	3500	0.8951	0.8984	0.8674	0.9316	0.7924	0.8952	0.8421	1.8813	0.2913

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1